United States
Environmental Protection
Agency
Water Engineering
Research Laboratory
Cincinnati OH 45268
Research and Development
EPA/600/S2-85/092 Sept. 1985
&EHV Project Summary
Ultraviolet Disinfection of
Water for Small Water
Supplies
Dale A. Carlson, Robert W. Seabloom, Foppe B. DeWalle, Theodore F.
Wetzler, Jogeir Engeset, Richard Butler, Somboon Wangsuphachart, and
Sinclair Wang
Ultraviolet (UV) radiation was con-
sidered as an alternative means of dis-
infecting small drinking water supplies.
A major impetus for this study was the
U.S. increase in reported waterborne
disease outbreaks caused by Giardia
lamblia, an organism that is highly re-
sistant to conventional chlorination.
Both field and laboratory studies
were used to evaluate the effectiveness
of UV radiation in reducing the viability
of Escharichia coli, Yersinia sp., and Gi-
ardia sp. UV sources included commer-
cial UV reactors and an excimer laser.
6. muris was used as a surrogate for
G. lamblia so that reliable excystation
and a consistent population of infective
organisms could be attained through-
out the seasons and through the proj-
ect study period.
G. muris cysts were significantly
more resistant to UV than £. coli and
more resistant than Yersinia sp. The ef-
fectiveness of disinfection depended on
the amount of UV radiation reaching
the organisms and on any hydraulic
shortcircuiting. The presence of en-
trapped air in the commerical UV reac-
tors decreased the efficiency of the re-
actor.
Natural or added color in the test
waters decreased the effectiveness of
UV disinfection on G. muris. For the
range and type of turbidity examined,
the shielding effect against bacterial
disinfection noted in other studies was
not observed.
Studies on G. muris cysts indicated
that storage time and temperature af-
fected the viability of the cysts and that
the rate of decrease in viability approxi-
mately doubled with each 10°C increase
in temperature above freezing. Below
freezing, however, cyst viability was
shortened to hours rather than to days
for above-freezing conditions.
Physical stress produced by pressure
and alum addition in water treatment
processes appeared to damage and
even destroy cysts.
This Project Summary was devel-
oped by EPA's Water Engineering Re-
search Laboratory, Cincinnati, OH, to
announce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
Introduction
The germicidal effects of sunlight
have long been known, but artificial ul-
traviolet (UV) radiation can obtain much
better microbiocidal action at a wave-
length of about 260 nm. A common
form of artificial UV radiation with a
wavelength of 253.7 nm can be pro-
duced by low-pressure mercury vapor
lamps. The inactivation of microorgan-
isms by UV radiation is based on photo-
chemical reactions in the DNA that re-
sult in coding system errors.
Early attempts to use UV radiation for
public water supply treatment began in
1910. They were not very successful and
were abandoned in favor of chlorina-
tion. However, two factors have re-
newed interest in UV radiation for pub-
-------
lie water supply disinfection: (1) the fact
that chlorination can produce byprod-
ucts that may be carcinogenic to hu-
mans and possibly toxic to aquatic life,
and (2) a large increase in giardiasis in
the United States, a waterborne disease
caused by the chlorine-resistant Giardia
lamblia cyst. The major impetus for this
study was the dramatic increase
in waterborne disease, specifically
giardiasis, in the United States since the
mid-1960's. The present investigation
uses both field and laboratory studies
to evaluate the effectiveness of UV
radiation in reducing the viability of
Escherichia coli, Yersinia sp., and
Giardia sp. UV sources included com-
mercial UV reactors and an excimer
laser.
Germicidal Characteristics of
UV Radiation
Ultraviolet (UV) radiation is defined
as electromagnetic radiation occupying
a small portion of the electromagnetic
spectrum lying beyond visible light. The
wave lengths of UV radiation range
from about 4 to 400 nm, with the narrow
band between 200 and 310 nm having
the greatest injurious and lethal effects
on microorganisms. The maximum mi-
crobiocidal action occurs at about 260
nm for practically all microorganisms
and is essentially congruent with the UV
absorption and photochemical sensitiv-
ity of deoxyribonucleic acid (DMA). The
inactivation of microorganisms is then
essentially based on photochemical re-
actions in the DNA that result in errors
and faults being introduced into the
coding systems. Nature has developed
various means of molecular-biological
error correction for the protection of the
vital DNA, and the selectivity of the re-
actions may be influenced by changes
in the organisms in different phases of
their life cycle. Thus photo repair sys-
tems may resuscitate a seemingly dead
organism by either longer wavelength
photo irradiation or dark incubation. In
addition, spores have been found to be
very resistant to radiation.
The germicidal action of UV radiation
results from its exposure to or direct
contact with the organisms, and it can
only be effective if it is absorbed. The
lethal effect of UV radiation results from
a photochemical reaction initiated by
absorption of a photon by the molecular
structure rather than by formation of a
toxic substance in the medium.
The inactivation of microorganisms
resulting from UV exposure is propor-
tional to the intensity (mW/cm2) multi-
plied by the time of exposure (sec). The
product of the irradiation intensity per
area and the time is called the UV dose
(mW-sec/cm2). Note that the only ad-
verse effect of an excessive dose of UV
radiation is additional cost.
Sources of UV Radiation
The sources of UV radiation are of
two classes—natural and artificial. The
sun is the most important natural
source of UV light. The oxidizing and
germ-killing effects of sunlight con-
tribute considerably to the conservation
of our environment by natural photo-
chemical processes in the atmosphere
and by natural ultraviolet purification of
surface water. UV light can also be gen-
erated artificially by a wide variety of
arcs and incandescent lamps. One com-
mon form of artificial UV radiation can
be generated from special low-
pressure, mercury-vapor lamps that
produce UV radiation as a result of an
electron flow between the electrodes
through ionized mercury vapor. These
artificial UV radiation sources can sup-
ply energy in such relatively high doses
that in fractions of a second they can
accomplish a higher degree of irradia-
tion than the sun can in several hours.
Since the maximum UV sensitivity of
microorganisms and the UV emission
of the low-pressure mercury vapor
lamp are well matched, the nearly
monochromatic low-pressure mercury
lamp has prevailed as the dominant ra-
diation source in research and practical
applications.
Early Experience with UV Dis-
infection
The first recorded attempt to use UV
radiation for public water supply treat-
ment was made in 1910 in France. Sub-
sequently, UV treatment was tried in the
United States with limited success, but
most systems were abandoned before
1930. The main reasons given for aban-
doning the UV method of treatment
were relatively high operating costs, op-
erating and maintenance problems, and
the advent of chlorination, which was
found to be more efficient and reliable.
Procedures
The UV equipment consisted of two
commercially available UV water disin-
fection units, a laboratory batch UV unit,
and an excimer laser UV unit to provide
coherent UV light at several discrete
wavelengths.
The early stage of this study used
only G. lamblia cysts supplied mainly by
hospitals and pathology laboratories
across the State of Washington. How-
ever, the supply of G. lamblia cysts was
closely related to the outdoor recre-
ational activities in the area and was
hence very seasonal. In addition, the
cysts proved difficult to excyst consis-
tently, which severely limited the
amount of information that could be
gathered. Thus we decided to use
G. mur/s cysts, which were indicated by
current information to be at least as
resistant as G. lamblia cysts. G. muris
cysts provided a relatively higher excy-
stability and less fluctuation of the
results. Until we develop a reliable
method to determine the viability of G.
lamblia, the use of G. muris cysts as a
surrogate is warranted. Female Swiss
Webster mice were used for cyst prop-
agation.
Strains of fully virulent Yersinia enter-
ocolitica were derived from human pa-
tients suffering from chronic gas-
trointestinal disease manifested largely
by recurrent and sporadic diarrhea or by
acute episodes of ileitis. Three separate
media were studied for isolating and
enumerating Y. enterocolitica because
of the lack of an acceptable standard-
ized method. The first two media,
MacTween* and mYE, required mem-
brane filtration of the sample, and the
third medium, Tergitol-7 was inoculated
by surface spreading.
Two different strains of E. coli were
evaluated in the batch UV studies: a
nalidixic-acid-resistant (NAR) strain and
a non-NAR strain.
Three different media were used to
isolate and enumerate the two different
strains of £. coli. The first two media,
eosin methylene blue (EMB) and m-
Endo, are commonly used media for
enumerating coliforms from water sam-
ples. Both media were substrates for
membrane filters. The third medium,
Tergitol-7 TTC, was used to differentiate
between E. co//-NAR and background
coliforms. Water samples were surface-
spread on Tergitol-7 TTC agar.
During the course of this investiga-
tion, efforts were made to obtain a UV
disinfection unit with a self-contained
dosimeter so that the actual UV dosage
could be read directly. These efforts
were fruitless, and it was necessary to
rely on actinometry to gauge the out-
*Mention of trade names or commercial products
does not constitute endorsement or recommenda-
tion for use.
-------
put of the UV unit. Since the geometry
of the unit, the flow patterns at various
Reynolds numbers, the degree and in-
tensity of reflection, the average dis-
tance between lamps, the penetration
depth, the cleanliness of the liquid and
of the UV lamp, the degree of short-
circuiting, and the degree of light scat-
tering all have potential effects on the
dosages applied, photo actinometry
was used to access the UV light inten-
sity of exposure.
Giardia cysts, Yersinia, and E. coli or-
ganisms were spiked into the UV treat-
ment unit influent stream through a
manifold and then mixed by a static
mixer. The organisms then passed
through the UV contactor with various
detention times and hence different ra-
diation doses. Color and turbidity were
added to determine their influence on
cyst and organism survival. The effect
of storage time on 6. muris cysts at var-
ious temperatures was also investi-
gated, as well as their settling character-
istics in quiescent conditions. Finally,
zeta potentials were determined for
G. lamblia over a wide range of pH
values.
Results and Discussion
With relatively long exposure times of
about 20 min, a 99-percent kill of cysts
was obtained by the Pen Ray Batch Re-
actor (Figure 1). Figure 2 compares data
on Giardia cyst survival versus UV
dosage from another study (Rice, E. W.
and J. C. Hoff. 1982. Inactivation of
Giardia lamblia Cysts by Ultraviolet Irra-
diation. Appl. Environ. Microbiol.
42:546-547) with data obtained in this
study using a commercially available
reactor. Note that the relatively low de-
struction rate of Giardia cysts shown in
the data from Rice and Hoff in Figure 2
was due to the low UV exposure dose,
and that commercial UV disinfection
units used in this study can achieve dis-
infection of the cysts. For all experimen-
tal conditions in which color was added
to the water (regardless of the source),
an increase in absorbance at 254 nm re-
sulted in an increase in the percentage
of cyst survival (Figure 3). On the other
hand, when turbidity was added, the de-
gree of cyst inactivation was not af-
fected with the retention time held con-
stant (Figure 4).
Though the laser-generated UV radia-
tion has a considerably greater intensity
than the mercury-vapor UV lamps, the
detention time for the laser pulse is on
the order of 10 nano-seconds. Thus
equivalent dose ranges can be obtained
from both sources. Data from this study
suggest that the commerical UV units
are much more effective than the ex-
cimer laser unit in inactivation of Giar-
dia cysts.
Comparison of UV inactivation curves
for Y. enterocolitica, E. coli, and G.
muris cysts graphically exhibits the
tremendous resistance cysts have to UV
inactivation. The fact that both Yersinia
and Giardia cysts are more resistant to
UV than E. coli has important implica-
tions where the total coliform proce-
dure is used to monitor disinfection effi-
ciency and indicate microbiological
water quality. Outbreaks of giardiasis in
water supply systems that reported sat-
isfactory total coliform concentrations
are indicative of the problems as-
sociated with using coliforms as indica-
tor organisms for adequate UV
disinfection.
A study was conducted on the effects
of storage time on G. muris cysts at var-
ious temperatures. The 1°C, 5°C, 10°C,
and 20°C experiments showed differ-
ences in die-off rates. No excystable
cysts were observed at 1"C, 5°C, 10°C,
and 20°C after storage periods of 120,
95, 63, and 26 days, respectively.
The settling characteristics were mea-
sured for G. muris cysts in distilled
water under quiescent conditions. They
demonstrated, for example, that after
2 days, approximately 50 percent of the
cysts were removed from the water
column.
Conclusions
1. The excystation procedure for Giar-
dia cyst viability was unreliable for
G. lamblia cysts and could not be
used to provide significant data.
2. To determine cyst viability, G.
muris cysts were used as a surro-
gate for G. lamblia because of their
relatively higher percentage of ex-
cystability and more consistent re-
producibility under laboratory con-
ditions.
3. Unlike chlorine and ozone, UV radi-
ation has no problems with mixing
<(mass transfer) in the contactor; it
also produces no residual. How-
ever, like chlorine and ozone con-
tactors, flow dispersion has a signif-
icant impact on UV's biocidal effect.
4. As the UV reactors approached
plug flow, greater degrees of disin-
fection were obtained.
5. Direct measurement of the actual
UV dosage in the two commercially
available UV contactors was not
possible. Instead, it was necessary
to rely on actinometry and inactiva-
I
too
90
80
70
5 60
50
40
30
20
10
0
Pen Ray UV Bench-Scale Study
1-Liter, 4-Inch Diameter Vessel
Giardia muris Cyst Concentration. 350/ml
in Distilled Water (Temp. 22.7°C-24.5°C)
O--O 1st Run 10/2/83)
A---A 2nd Run (9/27/83)
O-O 3rd Run (10/4/83)
Average Survival Rate
The intensity of the Pen Ray UV Unit used
in this experiment was 0.0914 mW/cm"
35
0 5 10 15 20 25
Exposure Time, min
0 27.4 54.8 82.3 109.7 137.1 164.5 191.9
UV Dosage, mW-sec/cm*
Figure 1. Giardia muris cyst survival versus UV exposure time with a Pen Ray UV lamp.
3
-------
Flow-Through
Disinfection Unit
A Giardia lamblia Study (Rice and Hoi'f. 19811
A Giardia muris Study of Pure Water Systems
1-30 Disinfection Unit using Cedar River
Water or Seattle City Water
• Giardia lamblia Study of Pure Water Systems
1-30 Disinfection Unit using Recirculation
of Distilled Water
• Giardia muris Study of Batch UV Pen Ray Unit
Reduction in unit efficiency due to
short-circuiting at low flowrates when
theoretical detention time exceeded
50 seconds in the flow-through disinfection
unit.
f
T
t
I
I
j I
Figure 2.
J ° ' 10- J ° ' 10*
UV Dosage, mW-sec/cm*
Survival of Giardia sp. cysts versus UV dosage.
104
10s
100
80
60
Color Measured as Absorbance
@ 254 nm (Nominal)
CO
40
20
0.100cm'1
0.050 cm"1
0.0/0 cm'1
Pure Water Systems 1-30 UV unit
Cedar River Water Spiked with
Cedar River Color Concentrate
t (0.1161
. (0.060)
1(0.007) i
I
10 15 20 25
Theoretical Detention Time, sec
30
35
Figure 3.
Effect of Cedar River color on G iardia muris cyst survival under UV light at various
detention times for Pure Water Systems unit.
tion of B. subtil is spores to indicate
the dosage.
6. The low destruction rate for Giardia
cysts reported by previous investi-
gators was primarily due to low UV
exposure dose relative to the actual
dose required.
7. Giardia cyst inactivation is a func-
tion of UV energy absorption and
thus depends on the amount of UV
light that reaches the cyst and the
time of exposure. Where commer-
cial UV unit design permits short-
circuiting, the time for 100 percent
inactivation can be expected to be
protracted.
8. The commercial UV disinfection
units tested can achieve disinfec-
tion of Giardia sp. The design of the
units is important in the detention
time required for exposure to UV
light. For the Pure Water Systems
-------
100
80
g? 60
I 4°
20
Pure Water Systems, 1-30
Cedar River Water Spiked With
Turbidity from the White River
Plus Natural Turbidity
Theoretical Detention Time
• 10 sec
A 75 sec
• 30 sec
456
Turbidity, NTU
10
Figure 4.
Effect of turbidity from White River on UV disinfection of G. muris cysts at various
theoretical detention times for the Pure Water Systems unit (with trapped gasi.
unit, 1 percent survival was attained
at detention times longer than 60
sec. In contrast, detention times
longer than about 22 sec were re-
quired to reach survival levels of
less than 1 percent in the Ultraviolet
Technology, Inc., unit.
9. Short-circuiting and collecting of air
bubbles in commercial UV disinfec-
tion units can seriously impair the
disinfection capability of the unit.
10. Commercial UV disinfection units
could be modified to make more ef-
ficient use of the generated UV light
and hence increase the feasibility of
UV disinfection.
11. The information presented here in-
dicates that the coherent UV light of
the excimer laser is no more effi-
cient in removing Giardia cyst vi-
ability than is the UV light from
mercury-vapor tubes in terms of the
energy output.
12. Laser equipment available during
the experimental period was unreli-
able and too difficult to operate to
be considered for use in field or
commercial disinfection systems.
13. For all experimental conditions,
color with an absorbance at 254 nm
was found to increase cyst survival.
For example, color caused a de-
crease in UV disinfection effective-
ness.
14. The presence of relatively small
inorganic or organic particulates
(5-fim diameter or less) had no dis-
cernible effect on the UV disinfec-
tion of cysts.
15. Turbidity in the form of larger sus-
pended particulates (>5 nm) may
provide shielding and protection to
the organisms.
16. Virulent Yersinia were less resistant
to UV inactivation than Giardia.
Hence if Giardia are removed by
UV, Yersinia can be expected to
have been destroyed as well.
17. Virulent Yersinia that contained a
plasmid demonstrated significantly
greater resistance to inactivation by
UV than its nonplasmid counter-
part.
18. UV killed E. co//very effectively.
19. The fact that both Yersinia and Giar-
dia cysts were more resistant to UV
than E. coli has important implica-
tions where the total coliform test js
used to monitor microbiological
water quality.
20. Size and morphological characteris-
tics of organisms and particles ap-
peared to be very important factors
in shielding them from UV radia-
tion.
21. Storage time and temperature af-
fect cyst viability. The decrease in
viability approximately doubles
with each 10°C increase in tempera-
ture above freezing. At freezing
temperatures, however, cyst viabil-
ity is drastically shortened from
months and days to hours.
22. Zeta potentials for 6. lamblia cysts
were time dependent, indicating a
change in cyst characteristics with
storage.
23. The physical stress produced by
pressure and alum addition ap-
peared to damage or even destroy
cysts.
Recommendations
1. The effect of turbidity on disinfection
of Giardia sp. needs more study. The
particle size and form and its interfer-
ence with UV disinfection particu-
larly need to be investigated.
2. A more practical method of measur-
ing and recording the actual deliv-
ered UV dose in the contactor is des-
perately needed.
3. Further work is needed to improve
the accuracy and precision of the
G. lamblia excystation procedure.
4. Further studies should be done to
determine the effects of storage,
temperature, and sedimentation on
Giardia sp. in the water environment.
These factors may significantly influ-
ence the operational mode for treat-
ment of surface water supplies.
5. Pressure and coagulant addition
may also have drastic influences on
the survival of Giardia cysts and thus
need to be investigated.
6. The design of UV contactors must
eliminate or minimize short circuit-
ing and air entrainment, and opti-
mize reflected radiation to improve
biocidal effect.
The full report was submitted in fulfill-
ment of Cooperative Agreement No.
809321 by the University of Washing-
ton, Seattle, WA, under the sponsorship
of the U.S. Environmental Protection
Agency.
-------
Dale A. Carlson, Robert W. Seabloom, Foppe B. DeWalle, Theodore F. Wetzler,
JogeirEngeset, Richard Butler, Somboon Wangsuphachart, and Sinclair Wang
are with University of Washington, Seattle, WA 98195.
Donald J. Reasoner is the EPA Project Officer (see below).
The complete report, entitled "Ultraviolet Disinfection of Water for Small Water
Supplies, "(Order No. PB 85-239 960/AS; Cost: $ 14.50, subject to change) will
be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Water Engineering Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use $300
EPA/600/S2-85/092
0063240 WERL
LOU W TILLEY
RESION V
LIBRARIAN
60604
------- |