Maintaining Legionella control in building water systems

Reprinted from Journal AWWA, Vol. 106 (10), by permission. Copyright © 2014, American Water Works Association. Permission to reproduce this
document is granted for informational purposes only and does not represent or imply approval or endorsement by AWWA of any particular product or service.
Maintaining Legionella
control in building
water systems
egionella and other waterborne pathogens can present a risk to
consumers of potable water. In particular, building hot water
systems have been established as the primary reservoir for bac­
teria linked to cases of Legionnaires’ disease (LD). These systems
provide ideal conditions for Legionella proliferation because of
their elevated temperature and lack of disinfection residual. Control of
Legionella in potable water systems has become a focus for health care
facilities because they serve a population that is particularly susceptible to LD
from underlying health conditions, such as suppressed immune systems. In
addition, the potential for disease transmission, associated liability concerns,
and anticipated new standards from professional organizations are raising
awareness of the importance of maintaining building water quality in non–
health care facilities.
The US Environmental Protection Agency’s (USEPA’s) National Primary
Drinking Water Regulations establish maximum contaminant levels (MCLs)
for several potentially harmful constituents, including microorganisms. The
MCLs include using treatment techniques to achieve removal of Giardia
lamblia, Cryptosporidium, enteric viruses, total coliforms, and Legionella. The
regulations make the assumption that if Giardia and viruses are removed
according to the treatment techniques in the Surface Water Treatment Rule,
then Legionella will also be controlled in the potable water supply. The key
word in this regulation is “controlled,” which suggests that Legionella may
remain present but at quantities that do not present negative health effects
from exposure. Legionella may be controlled through municipal water treat­
ment processes, not because treatment techniques are successful in removal
of Legionella, but because the physicochemical conditions of the treated cold
2014 © American Water Works Association
Four technologies providing residual disinfectant to control Legionella
Supplemental Chlorination
Chlorine Dioxide
2–4 mg/L free residual chlorine
0.1–0.8 mg/L ClO2
2.0–4.0 mg/L as Cl2
Copper 0.2–0.8 ppm
Silver 0.02–0.08 ppm
system residual
pH > 8 affects efficacy
No impact in 6.0–10.0 range
No impact in 7.0–9.0 range
pH > 8.5 may affect efficacy
Elevated temperatures
accelerates decay
Elevated temperatures
accelerate decay
Minimal impact by elevated
No impact by temperature
THM and HAA5
Reduced THM and HAA5
compared with chlorine
No chemical reaction to
form by-products
Drinking water
Chlorine < 4.0 mg/L
Chlorine dioxide < 0.8 mg/L
Chloramine < 4.0 mg/L
as Cl2
Copper < 1.3 mg/L
Silver < 0.10 mg/L
Cl2—chlorine, ClO2—chlorine dioxide, HAA5—haloacetic acids, THM—trihalomethane
water are not conducive to amplifi­
cation of Legionella. Water tempera­
tures exceeding 77°F (25°C) have
been observed as necessary to sup­
port the growth of Legionella (Yee
& Wadowsky, 1982). Typical munic­
ipal treated water temperatures are
below 77°F (25°C), allowing Legionella to persist in the water but not
to amplify to detectable quantities.
To maintain microbiological water
quality in the distribution system,
water suppliers provide a disinfec­
tant residual, most commonly in the
form of free chlorine. Drinking water
regulations require water suppliers
that use surface water or groundwa­
ter under the influence of surface
water to maintain a residual disinfec­
tant concentration in the water
entering the distribution system of
not less than 0.2 mg/L for more than
4 hours (40 CFR 141.72(a)(3) and
(b)(2)). The residual disinfectant in
the distribution system cannot be
undetectable in more than 5% of
the samples from the distribution
system each month for any two con­
secutive months (40 CFR 141.72(a)
(4) and (b)(3)). Although the au­­
thors believe most public water sup­
pliers make their best efforts to
maintain a detectable disinfectant
residual in all delivered drinking
water, maintenance of a detectable
disinfectant residual throughout an
entire distribution system may not
always be possible.
Although maintenance of the min­
imum disinfectant level may provide
some control over growth of general
heterotrophic plate count bacteria,
0.2 mg/L of free residual chlorine is
not sufficient to control regrowth of
all bacteria, including Legionella. A
study of the susceptibility of Legionella to chlorine in tap water con­
cluded that Legionella can survive in
the presence of low levels of chlorine
for relatively long periods of time
(Kuchta et al, 1983). Other studies
have shown that Legionella bacteria
can survive in the presence of chlo­
rine concentrations up to 50 mg/L
because of their association with
other organisms in the biofilms that
exist throughout the treatment and
distribution system. (Cooper & Han­
lon, 2010; Kilvington & Price, 1990;
King et al, 1988). Researchers also
conducted studies during several
years and found that the potential
exists for Legionella growth within
municipal drinking water systems
and that public water supplies may
contaminate the plumbing systems
of hospitals and other large buildings
(States et al, 1987).
Once Legionella enter a building
water system, the bacteria can
amplify (particularly in hot water
systems where temperatures are
optimum), nutrients accumulate,
and any disinfectant residual in the
cold supply is consumed. However,
Legionella are not found in all
buildings, with studies reporting
that 12–70% of surveyed water sys­
tems have some level of Legionella
colonization (Lin et al, 1998).
Where Legionella is found in build­
ing water systems, an assessment of
risk should include identifying the
species and Legionella serogroups
present, the extent of colonization,
and the susceptibility of occupants to
acquiring LD. Using this information,
a building operator can determine if
secondary disinfection is warranted
to reduce the risk associated with the
transmission of Legionella from
building water systems.
To reduce the amplification of
Legionella in building water sys­
tems, particularly those serving a
susceptible population such as
health care facilities, secondary dis­
infectants are often necessary. Table
1 shows four technologies that pro­
vide a residual disinfectant that have
been considered for systemic disin­
fection of building water systems to
control Legionella: supplemental
chlorination, chlorine dioxide,
mono­­chloramine, and copper–silver
2014 © American Water Works Association
USEPA guidelines on use and monitoring of chlorine dioxide as the
primary disinfectant
Chlorine dioxide MRDL
0.8 mg/L as chlorine dioxide
Chlorine dioxide MRDLG
0.8 mg/L as chlorine dioxide
Chlorine dioxide residual monitoring
Daily at entrance of distribution system
Chlorite MCL
1.0 mg/L as chlorite
Chlorite MCLG
0.8 mg/L as chlorite
Chlorite monitoring
Daily at entrance of distribution system and
monthly at three locations in distribution system
MCL—maximum contaminate level, MCLG—maximum contaminate level goal, MRDL—maximum residual
disinfectant level, MRDLG—maximum residual disinfectant level goal
*USEPA, 1998.
ionization. Evaluation of disinfec­
tion methods to demonstrate their
efficacy should be evidenced-based
and follow a four-step approach: (1)
demonstrate in vitro efficacy, (2)
anecdotal experience of efficacy in
individual hospitals, (3) peerreviewed controlled studies of pro­
longed duration documenting effi­
cacy and prevention of LD, and (4)
confirmatory reports from multiple
hospitals with a prolonged duration
of follow-up (Stout & Yu, 2003).
There is not a single disinfection
technology that is applicable to all
water systems, with anecdotal and
published reports of underperfor­
mance or failure of disinfectants for
various reasons. The successful
application of a secondary disinfec­
tant is dependent on several factors,
including the ability to maintain a
disinfectant residual, configuration
of the water system, cost of consum­
ables, operation and maintenance,
source water quality, and permitting
requirements. Most important, the
selected secondary disinfectant
should demonstrate efficacy against
Legionella without negatively affect­
ing the water distribution system.
Supplemental chlorination in­­
volves the injection of sodium hypo­
chlorite into the potable cold and/or
hot water system to achieve free
chlorine residuals ≤ 4 mg/L at the
building outlets. A review of 17 hos­
pitals that used supplemental chlori­
nation for Legionella disinfection
was presented in 1990, and it was
reported in 2011 these facilities were
no longer using supplemental chlori­
nation (Lin et al, 2011; Muraca et al,
1990). Concerns with supplemental
chlorination in building water sys­
tems include accelerated corrosion of
the water system, formation of dis­
infection by-products, and poor bio­
film penetration resulting in persis­
tence of Legionella (Giao et al, 2009;
Garcia et al, 2008; Morris et al,
1992; Helms et al, 1988). For these
reasons, along with previously dis­
cussed peer-reviewed studies docu­
menting Legionella resistance to
chlorination, supplemental chlorina­
tion is not considered an effective
permanent disinfection method for
Legionella in building water systems.
The successful use of copper–silver
ionization for Legionella control in
building hot water systems has been
well-documented in the peerreviewed literature (Lin et al, 2011;
Stout & Yu, 2003; Colville et al,
1993). Target concentrations vary
between vendors but typically fall
within a range of 0.2–0.8 mg/L for
copper and 0.01–0.08 mg/L for sil­
ver. Ionization systems consist of a
flow cell with metallic copper and
silver anodes connected to an elec­
tronic controller. The release of ions
into the water is controlled by
2014 © American Water Works Association
adjusting the amperage of current
that is applied across the anodes.
Monitoring of copper ions can be
performed in the field using a handheld colorimeter, but laboratory
analysis is required for measurement
of silver ion concentration and to
confirm field-measured copper con­
centrations. Water system operators
considering ionization for secondary
disinfection should be aware of regu­
latory agencies’ unfamiliarity with
ionization because it is not consid­
ered a primary disinfectant in the
municipal treatment process and is
regulated under the Safe Drinking
Water Act as a primary metal con­
taminant (copper) and secondary
contaminant (silver).
Monochloramine is used as a dis­
infectant by municipal drinking
water providers because it is more
stable than chlorine and produces
fewer regulated disinfection by-prod­
ucts. In areas where monochlora­
mine is used at the municipal level
for disinfection of drinking water, a
lower incidence of LD has been doc­
umented compared with municipal
water systems that use chlorine for
disinfection. (Heffelfinger et al,
2003; Kool et al, 1999). Using
monochloramine to treat secondary
distribution systems is made possible
through small-scale monochlora­
mine generators. Monochloramine
for injection into building water sys­
tems is produced from a combina­
tion of a stabilized chlorine solution
and a buffered ammonium salt solu­
tion. Target concentrations for
monochloramine are 2–4 mg/L as
total chlorine. Operators should also
monitor free ammonia in the water
systems receiving monochloramine
treatment to ensure the correct pro­
portion of chemical precursors is
being applied. The first evaluation of
the efficacy monochloramine for
Legionella control in a building hot
water system was performed in an
Italian hospital; the results of this
evaluation showed a dramatic reduc­
tion in Legionella colonization
(Marchesi et al, 2012). The first US
evaluation of monochloramine
application in a hospital water sys­
tem has reported similar results
documenting Legionella efficacy
(Stout et al, 2012). Although these
two evaluations show promising
results, continued long-term evalua­
tion of monochloramine should be
performed to understand any limita­
tions that may be experienced in dif­
ferent water systems.
Chlorine dioxide may also be con­
sidered by water system operators
for Legionella disinfection. The case
study that follows provides a review
of chlorine dioxide and a long-term
evaluation of one facility that has
used chlorine dioxide for Legionella
control during the past 13 years.
Chlorine dioxide has been used as
a water treatment chemical in the
United States since 1944, when it
was used as an oxidant to remove
phenol-related compounds from
drinking water. Chlorine dioxide is
now primarily used by municipal
water providers for disinfection and
control of tastes, odor, iron, manga­
nese, hydrogen sulfides, and phenolic
compounds. USEPA reported in
1999 that an estimated 700–900
public water systems in the United
States use chlorine dioxide as a
municipal water treatment technol­
ogy (USEPA, 1999). Chlorine diox­
ide disinfection for the control of
Legionella in building and secondary
distribution systems has been
reported in facilities throughout the
United States and Europe (Zhang et
al, 2009; Sidari et al, 2004).
USEPA has established regulations
governing the use and monitoring of
chlorine dioxide in water systems
using it as a primary disinfectant
(Table 2). Treatment of potable
water with chlorine dioxide results
in the formation of chlorite, which is
a disinfection by-product regulated
by USEPA. Table 2 also provides
details on the maximum contami­
nant levels and monitoring require­
ments for chlorite in municipal
drinking water systems. USEPA lists
the following health concerns
regarding chlorite: “some people
may experience anemia, some infants
and young children who drink water
containing chlorite in excess of
the MCL could experience nervous
system effects, and similar effects
may occur in fetuses of pregnant
women who drink water containing
chlorite in excess of the MCL”
(USEPA, 2014).
Chlorine dioxide possesses several
characteristics that allow it to per­
form well as a disinfectant. Chlorine
dioxide’s oxidation reduction poten­
tial (0.95 V) is lower than that of
chlorine (1.36 V), whereas its oxida­
tion capacity—5—is greater than
chlorine—2. The oxidation reduc­
tion potential (ORP) measures an
oxidizer’s strength or the speed at
which it reacts with an oxidizable
material. Although chlorine dioxide
has a low ORP, it is more selective
about the types of oxidizable materi­
als with which it reacts. Chlorine
dioxide targets specific organic mol­
ecules, including cysteine, tyrosine,
methionyl, DNA, and RNA, unlike
the broad reactions of chlorine and
ozone. The oxidation capacity indi­
cates that, on a molar basis, chlorine
dioxide has a greater disinfection
capacity than chlorine. The selectiv­
ity and oxidation capacity of chlo­
rine dioxide makes it a stronger oxi­
dative disinfectant than chlorine.
Chlorine dioxide has also been
shown to have superior biofilm pen­
etration and biofouling control com­
pared with chlorine (Simpson et al,
1993; Mayack et al, 1984).
Chlorine dioxide disinfection in
building water systems is performed
by injecting a concentrated chlorine
dioxide solution into the potable
cold and/or hot water systems. Chlo­
rine dioxide can be supplied for
small-scale applications as a stabi­
lized solution or generated onsite by
means of direct oxidation (electro­
chemical) or a chemical blending
process, typically using a strong acid
solution and sodium chlorite. The
various generation methods produce
differing levels of solution purity and
production yields depending on
water system disinfection require­
ments and operating conditions.
In January 2004, the results of
the first large-scale evaluation of
chlorine dioxide efficacy for Legionella disinfection in a US secondary
water distribution system were
published in Journal AWWA
(Sidari et al, 2004). The study con­
cluded that chlorine dioxide was a
promising alternative disinfectant
for Legionella, and long-term stud­
ies of its efficacy were warranted.
This case study provides a longterm evaluation of the chlorine
dioxide system after 13 years of
operation to determine if chlorine
dioxide has continued to produce
acceptable results in controlling
Legionella in the facility’s water
system and the prevention of LD.
Background. The study hospital in
Pennsylvania identified 13 cases of
LD between 1994 and 1999, with
three cases in an 18-month period
that were confirmed as hospitalacquired. Responding to these cases,
the hospital began environmental
monitoring for Legionella in its
drinking water system in 1998.
Interventions to control Legionella
included maintenance of a free chlo­
rine residual throughout the distri­
bution system and performance of
thermal eradication (heat and flush)
in identified Legionella-positive
areas of the distribution system.
Although temporarily effective, the
hospital did not want to use thermal
eradication as a permanent disinfec­
tion approach because of short-term
efficacy, logistical difficulties, and
cost. Alternative disinfection ap­­
proaches evaluated by the hospital
in 1998 included supplemental chlo­
rination, copper–silver ionization,
and chlorine dioxide. Supplemental
chlorination was ruled out because
of corrosion and efficacy concerns.
Copper–silver ionization was ini­
tially considered as an option
because of its effective performance
2014 © American Water Works Association
Campus water distribution system and sampling locations at study hospital
Distribution loop
Sampling location
Parking and roads
Pubic buildings
Staff buildings
at other facilities; however, because
each of the hospital’s 23 buildings
would have required installation and
maintenance of a separate ionization
system, this option was determined
to be cost-prohibitive. Monochlora­
mine was not considered during the
initial selection because generation
technology had not been developed
or evaluated at the time. On the basis
of the favorable literature review of
laboratory trials, reports of success­
ful European applications, and cost
considerations, the hospital decided
to install a chlorine dioxide disinfec­
tion system for treatment of its cam­
pus water distribution system. Sys­
tem installation was completed in
June 2000.
Installation and operation. The hos­
pital operates a sizeable secondary
distribution system on its 140-acre
campus (see the map above). The
facility included 23 buildings provid­
ing 437 patient beds during the initial
study; however, after several demoli­
tion and construction projects, the
campus now incorporates 20 build­
ings providing 478 patient beds. The
distribution system consists of a
520,000-gallon covered, above-grade
concrete reservoir and ~ 10,000-foot
2014 © American Water Works Association
distribution loop of 6- and 8-inch pip­
ing. The distribution loop has re­­
mained largely unchanged since
completion of the initial study, with
only minor modifications to accom­
modate new construction. Each
building is supplied water from the
distribution loop. Cold water is dis­
tributed in each building for potable
and utility uses and separate hot
water generation systems supply hot
water within each building.
The hospital receives the majority
of its daily water supply (approxi­
mately 80%) from the local munici­
pality, which provides conventional
treatment, including chlorination of
a surface water source. The remain­
ing daily demand is met by the hos­
pital using onsite well water that is
chlorinated before being blended
with the municipal supply in the res­
ervoir. Average daily water use dur­
ing the initial study was 400,000 gpd
in the summer and 250,000 gpd in
the winter; by 2013, this was reduced
to 232,000 gpd in the summer and
222,000 gpd in the winter. The
reduction in water usage is attrib­
uted to reduced utility water
demands from energy-efficiency
The chlorine dioxide system was
installed in June 2000 and consisted
of three electrolytic generators.1 In
2009, one unit was replaced with a
2000 series generator and is now
used as the primary generation unit.
The generators directly oxidize a
25% active sodium hypochlorite
solution across a membrane system,
producing a concentrated 500-mg/L
chlorine dioxide solution. The initial
operation of the generators was in a
flow-paced lead, ORP-controlled lag
mode, allowing for continual dosing
of chlorine dioxide into the reservoir.
In March 2003, the generator opera­
tion was changed to a time-paced
application with the generators pro­
ducing chlorine dioxide for approxi­
mately 45 minutes per hour based on
measured chlorine dioxide residual
in the reservoir. The application rate
is adjusted manually by operators to
target a chlorine dioxide range in the
reservoir of 0.3–0.5 mg/L.
The chlorine dioxide system dur­
ing the initial study operated con­
tinuously, seven days per week. In
February 2003, the operational
mode changed because of staffing
availability at the hospital and the
need to have a water treatment oper­
ator onsite during chlorine dioxide
application. Currently, the genera­
tors are typically turned off from
23:00 on Friday night until 01:00
Monday morning each week.
The hospital’s water system is per­
mitted as a nontransient, noncom­
munity water system, and the addi­
tion of the chlorine dioxide treatment
system required a permit modifica­
tion. The permit from the regulatory
agency limited chlorine dioxide con­
tractions to 1.0 mg/L at the reservoir
and 0.8 mg/L at the distal outlets.
The hospital has regularly performed
monitoring of the chlorine dioxide
system to comply with permitting
requirements and to ensure that ade­
quate chlorine dioxide residuals for
maintenance of Legionella control
are present within the distribution
system. The hospital has reported no
permit violations associated with the
chlorine dioxide system since opera­
tion began. Chlorine dioxide is mea­
sured at the entry point to the distri­
bution system on mornings when
disinfectant is being applied (typi­
cally Monday–Friday). Chlorine
dioxide concentration in the reser­
voir is also checked three times per
day (once per shift) during operation
to check on appropriate application
rates. Chlorine dioxide is measured
by the hospital using an N,N-diethylp-phenylenediamine method and
field colorimeter. Chlorite measure­
ment using a titrator2 is performed
by the hospital at the entry point to
the distribution system each day that
disinfectant is applied. The hospital
also performs third-party laboratory
testing of chlorite quarterly at the
distribution system entry, point of
medium residence time, and point of
maximum residence time. Figure 1
shows 2013 average daily chlorine
dioxide and chlorite measurements
at the distribution system entry point
(closest distal outlet to the reservoir).
Chlorine dioxide concentrations
decline during the weekend when
chlorine dioxide application is sus­
pended and then recover during the
week (Figures 1 and 2).
System monitoring. System moni­
toring to support this long-term
evaluation of chlorine dioxide has
been performed since 1998. The
monitoring periods are pre-chlorine
dioxide (pre-ClO2; April 1998–May
2000), phase 1 (June 2000–January
2001), phase 2 (February 2001–
April 2002), phase 3 (October
2003–August 2012), and phase 4
(October 2013).
Figure 3 shows a summary of the
hot water distal outlet monitoring
completed during this long-term
evaluation of chlorine dioxide at the
hospital. The results of pre-ClO2,
phase 1, and phase 2 monitoring
were reported in the initial study.
Phase 3 consisted of sampling from
hot water distal outlets across the
campus and was conducted 27 times
by the hospital between October
2003 and August 2012. The sam­
pling locations and methods were
generally consistent with those from
the initial study, with adjustments
made for patient occupancy and con­
struction. Phase 4 consisted of a
single sampling event conducted in
October 2013 by the authors to
independently evaluate the system
using sampling methods correspond­
ing to the initial study protocol.
During phase 3, the hospital col­
lected 304 samples from hot water
outlets for Legionella culture with
four samples positive for Legionella
(1.3% distal positivity). In July 2004,
three positive samples were identi­
fied, and in a 2011 sampling event,
one positive sample was identified.
Phase 4 sampling found one of 20
(5% distal positivity) hot water out­
lets positive with Legionella pneumophila, serogroup 5. The mainte­
nance of very low Legionella
positivity has been completed in the
presence of average chlorine dioxide
residuals at the hot water distal out­
lets of 0.08 mg/L (phase 3), which is
not different from the average con­
centration observed during the initial
study (phases 1 and 2; Figure 3).
During the pre-ClO2 period, Legionella was detected in 9% (2/22) of
samples from the building cold
source water. After application of
chlorine dioxide, no Legionella was
detected in phase 1 (0/14, 0.26 mg/L
average chlorine dioxide) and phase
2 (0/80, 0.50 mg/L average chlorine
dioxide). Legionella in the building
cold source water was not monitored
by the hospital during phase 3. In
phase 4, water samples were col­
2014 © American Water Works Association
FIGURE 1 2013 average daily chlorine dioxide and chlorite measurements
at the distribution system entry point
Chlorite MCL
Clorine dioxide MRDL
Cold water chlorine dioxide
Hot water chlorine dioxide
Point of Entry Residual
Day of Week Sampled
MCL—maximum contaminant level, MRDL—maximum residual disinfectant level
lected from the cold water supply to
five buildings and no Legionella was
detected (0/5, 0.22 mg/L average
chlorine dioxide residual).
Another indication of successful
application of a secondary disinfec­
tant is if the application results in
reduced rates of infection. The hos­
pital had experienced 13 cases of
LD in the five years before installa­
tion of the chlorine dioxide system,
and the initial study reported that
no cases of LD had been detected
between June 2000 and April 2002.
The hospital has used the urinary
antigen test for clinical screening of
suspected cases of LD since 2003,
and no cases have been detected in
the 13 years since application of
chlorine dioxide started.
Initial study review. The initial
study provided several conclusions
drawn from an evaluation of the
data collected after approximately
two years of chlorine dioxide appli­
cation (phases 1 and 2). To evaluate
and support the long-term efficacy of
chlorine dioxide for Legionella con­
trol, the conclusions from the initial
study can be reviewed in light of the
data collected (phases 3 and 4) dur­
ing the long-term operation of the
chlorine dioxide system.
Conclusion 1. A significant reduc­
tion in Legionella positivity was
observed in cold building source
water and distal hot water outlets
after application of chlorine dioxide,
and chlorine dioxide provided better
Legionella control at distal outlets
than thermal eradication and chlori­
nation. The data collected during
phases 3 and 4 demonstrated that
Legionella positivity in the distal hot
water outlets has decreased further
from the initial study and the preClO2 levels (Figure 3). Also, after
application of chlorine dioxide, Legionella has not been detected in the
cold building source water during
monitoring in phases 1, 2, and 4 (cold
source water monitoring was not per­
formed during phase 3). The results
of the long-term study indicate that
chlorine dioxide can maintain a last­
ing reduction in Legionella positivity
in a secondary distribution system.
Conclusion 2. Complete eradica­
tion of Legionella from hot water
distal sites was not realized after
2014 © American Water Works Association
1.75 years of chlorine dioxide treat­
ment; however, a trend of declining
distal outlet positivity was observed.
The trend of decreasing Legionella
positivity has continued with only
one positive Legionella sample col­
lected during phase 3 and one posi­
tive sample during phase 4. Overall
percent positivity has continued to
decrease since the initial study. The
positive samples indicate complete
eradication of Legionella in the hot
water systems may not be possible,
although maintenance of a very low
level of Legionella positivity can be
achieved with chlorine dioxide. It
may be unreasonable to achieve
complete eradication of Legionella
from a distribution system, and
complete eradication may not be
necessary to prevent disease (see
Conclusion 4).
Conclusion 3. Chlorine dioxide
concentrations of 0.3–0.5 mg/L
were effective in reducing or eradi­
cating Legionella. However, a sig­
nificant reduction in chlorine diox­
ide residual was observed between
the reservoir and distal hot water
outlets in the buildings. Average
2013 chlorine dioxide concentra­
tion (Monday–Friday during opera­
tion of the chlorine dioxide genera­
tors) was 0.35 mg/L in the reservoir,
0.36 mg/L at the cold water point of
entry, and 0.09 mg/L at the hot
water point of entry. A reduction in
chlorine dioxide residual is still
observed between the reservoir/cold
water and distal hot water outlets.
This is to be expected because of
long residence time in the distribu­
tion system and elevated tempera­
tures in the hot water systems,
which result in the accelerated
decay of chlorine dioxide. The chlo­
rine dioxide residuals being main­
tained by the hospital are providing
effective control of Legionella in the
distribution system, as evidenced by
the environmental sampling results.
Conclusion 4. No cases of LD
were detected at the hospital after
the use of chlorine dioxide began,
despite sporadic isolation of Legionella at distal hot water outlets. Sam­
Frank Sidari (to
should be sent) is
vice-president of
consulting at SPL
Services, 1401 Forbes Ave., Suite
209b, Pittsburgh, PA 15219;
[email protected]
He is a registered professional
engineer, board-certified
environmental engineer, certified
construction document
Chlorine dioxide
Tuesday Wednesday Thursday
Day of Week Sampled
FIGURE 3 Hot water distal outlet monitoring completed during the long-term
evaluation of ClO2 at the hospital
Distal site positivity
Hot water chlorine dioxide
Phase 1
Phase 2
(April 1995– (June 2000– (Feb. 2001–
May 2000)
Jan. 2001)
April 2002)
Phase 3
(Oct. 2003–
Aug. 2012)
Phase 4
(Oct. 2013)
Hot Water Distal Site Legionella Positivity—%
The long-term sampling conducted
by the hospital and the independent
sampling performed by the authors
demonstrates that, after 13 years of
chlorine dioxide application, Legionella continues to be successfully
controlled at the hospital. Using the
chlorine dioxide system to apply a
residual at the point of entry to the
campus distribution system, the hos­
pital is able to provide a chlorine
dioxide residual in its building water
systems. Legionella in the cold water
has not been detected since applica­
tion of chlorine dioxide began. Legionella positivity at the hot water
distal outlets has continued to
decrease since application began and
was only detected in one sample dur­
ing long-term monitoring by the hos­
pital. Most important, the applica­
tion of chlorine dioxide has
prevented LD at the hospital, dem­
onstrating the long-term efficacy of
chlorine dioxide in a secondary dis­
tribution system.
Reservoir Chlorine Dioxide Residual—mg/L
FIGURE 2 2013 average daily chlorine dioxide residual in reservoir, showing
impact of suspended application on weekends
Average Hot Water Chlorine
Dioxide Residual—mg/L
pling conducted during phases 3 and
4 detected Legionella at distal hot
water outlets, but at a reduced level
from phases 1 and 2. The hospital
has continued to perform clinical
screening for suspected cases of LD
and no cases have been detected after
13 years of chlorine dioxide applica­
tion. This is perhaps the best evi­
dence to demonstrate the long-term
efficacy of chlorine dioxide in a sec­
ondary distribution system.
Monitoring by Phase
ClO2—chlorine dioxide
technologist, and a certified
Hazard Analysis and Critical
Control Points auditor. Sidari
holds a master’s of science degree
in civil and environmental
engineering from Carnegie Mellon
University and a bachelor’s of
science in forest engineering from
the State University of New York
Environmental Science and
Forestry. His project work has
focused on distribution systems,
pumping, storage, treatment, and
disinfection on systems ranging in
size from 10 to 70 mil gal; he
specializes in engineering
assessments of water systems
affected by Legionella. Janet E.
Stout is president of Special
Pathogens Laboratory, Pittsburgh,
2014 © American Water Works Association
Pa. Scott Duda is a project
engineer for SPL Consulting
Services, Pittsburgh, Pa. Doug
Grubb is with facilities operations,
and Alan Neuner is the vicepresident of facilities operations,
both at the Geisinger Medical
Center, Danville, Pa.
1DIOX Water Hygiene 1000 series electrolytic
generators, Klenzoid Inc., Conshohocken, Pa.
2Severn Trent Capital Control Titrator, Wash­
ington, Pa.
Colville, A.; Crowley, J.; Dearden, D.; Slack,
R.C.B.; & Lee, J.V., 1993. Outbreak of
Legionnaires’ Disease at University
Hospital, Nottingham. Epidemiology,
Microbiology and Control. Epidemiology &
Infection, 110:1:105.
Cooper, I.R. & Hanlon, G.W., 2010. Resistance
of Legionella Pneumophila Serotype 1
Biofilms to Chlorine-Based Disinfection.
Journal of Hospital Infection, 74:2:152.
Garcia, M.T.; Baladron, B.; Gil, V.; Tarancon,
M.L.; Vilasau, A.; Ibanez, A.; Elola, C.; &
Pelaz, C., 2008. Persistence of ChlorineSensitive Legionella Pneumophila in
Hyperchlorinated Installations. Journal
of Applied Microbiology, 105:3:837.
Giao, M.S.; Wilks, S.; Azevedo, N.F.; Vieira, M.J.;
& Keevil, C.W., 2009. Incorporation of
Natural Uncultivable Legionella
Pneumophila Into Potable Water Biofilms
Provides a Protective Niche Against
Chlorination Stress. Biofouling, 25:4:345.
Heffelfinger, J.D.; Kool, J.L.; Fridkin, S.; Fraser,
V.J.; Hageman, F.; Carpenter, J.; &
Whitney, C.G. 2003. Risk of HospitalAcquired Legionnaires’ Disease in Cities
Using Monochloramine versus Other
Water Disinfectants. Infection Control
and Hospital Epidemiology, 24:8:569.
Helms, C.M.; Massanari, R.M.; Wenzel, R.P.;
Pfaller, M.A.; Moyer, N.P.; & Hall, N.,
1988. Legionnaires’ Disease Associated
With a Hospital Water System. A FiveYear Progress Report on Continuous
Hyperchlorination. Journal of the
American Medical Association,
Kilvington, S. & Price, J., 1990. Survival of
Legionella Pneumophila Within Cysts of
Acanthamoeba Polyphaga Following
Chlorine Exposure. Journal of Applied
Bacteriology, 68:5:519. http://dx.doi.
King, C.H.; Shotts, E.B., Jr.; Wooley, R.E.; &
Porter, K.G., 1988. Survival of Coliforms
and Bacterial Pathogens within
Protozoa during Chlorination. Applied
and Environmental Microbiology,
Kool, J.L.; Carpenter, J.C.; & Fields, B.S., 1999.
Effect of Monochloramine Disinfection of
Municipal Drinking Water on Risk of
Nosocomial Legionnaires’ Disease.”
The Lancet, 353:9149:272. http://dx.doi.
Kuchta, J.M.; States, S.J.; McNamara, A.M.;
Wadowsky, R.M.; & Yee, R.B. 1983.
Susceptibility of Legionella Pneumophila
to Chlorine in Tap Water. Applied and
Environmental Microbiology, 46:5:1134.
Lin, Y.E.; Stout, J.E.; & Yu, V.L. 2011. Controlling
Legionella in Hospital Drinking Water: An
Evidence-Based Review of Disinfection
Methods. Infection Control and Hospital
Epidemiology, 32:2:166. http://dx.doi.
Lin, Y.E.; Stout, J.E.; Yu, V.L.; & Vidic, R.D. 1998.
Disinfection of Water Distribution
Systems for Legionella. Seminars in
Respiratory Infections, 13:2:147.
Marchesi, I.; Stefano, C.; Marchegiano, P.;
Frezza, G.; Borella, P.; & Bargellini, A,
2012. Control of Legionella
Contamination in a Hospital Water
Distribution System by Monochloramine.
American Journal of Infection Control,
Mayack, L.A.; Soracco, R.J.; Wilde, E.W.; &
Pope, D.H., 1984. Comparative
Effectivness of Chlorine and Chlorine
Dioxide Biocide Regimes for Biofouling
Control. Water Research, 18:5:593. http://
Morris, R.D.; Audet, A.M.; Angelillo, I.F.;
Chalmers, T.C.; & Mosteller, F., 1992.
Chlorination, Chlorination By-products,
and Cancer: A Meta-analysis.” American
Journal of Public Health, 82:7:955.
Muraca, P.W.; Yu, V.L.; & Goetz, A., 1990.
Disinfection of Water Distribtuion
Systems for Legionella: A Review of
Application Procedures and
Methodologies. Infection Control and
Hospital Epidemiology, 11:2:79.
Sidari III, F.P.; Stout, J.E.; VanBriesen, J.M.;
Bowman, A.M.; Grubb, D.; Neuner, A.;
Wagener, M.; & Yu, V.L., 2004. Keeping
2014 © American Water Works Association
Legionella Out of Water Systems.
Journal AWWA, 96:1:111.
Simpson, G.D.; Laxton, G.D.; Miller, R.F.; &
Clements, W.R., 1993. A Focus on
Chlorine Dioxide: The Ideal Biocide.
(accessed July 29, 2014).
States, S.J.; Conley, L.F.; Kuchta, J.M.; Oleck,
B.M.; Lipovich, M.J.; Wolford, R.S.;
Wadowsky, R.M.; McNamara, A.M.;
Sykor, J.L.; & Keleti, G., 1987. Survival
and Multiplication of Legionella
Pneumophila in Municipal Drinking
Water Systems. Applied and
Environmental Microbiology, 53:5:979.
Stout, J.E.; Duda, S.; Kandiah, S.; Hannigan,
J.; Yassin, M.; Fabizio, M.; Ferrelli, J.;
Hariri, R.; Goepfert, J.; Bond, J.; &
Rogers, D., 2012. Evaluation of a New
Monochloramine Generation System for
Controlling Legionella in Building Hot
Water Systems. Association of Water
Technologies Annual Convention and
Exposition, Palm Springs, Calif.
Stout, J.E. & Yu, V.L., 2003. Experiences of the
First 16 Hospitals Using Copper–Silver
Ionization for Legionella Control:
Implications for the Evaluation of Other
Disinfection Modalities. Infection Control
and Hosptial Epidemiology, 24:8:563.
USEPA (US Environmental Protection
Agency), 2014. Water: Basic Information
about Regulated Drinking Water
Contaminants. Basic Information about
Disinfection Byproducts in Drinking
Water: Total Trihalomethanes,
Haloacetic Acids, Bromate, and
disinfectionbyproducts.cfm (accessed
May 22, 2014).
USEPA, 1999. Chlorine Dioxide. EPA Guidance
Manual Alternative Disinfectants and
Oxidants. Office of Water, Washington.
USEPA, 1998. National Primary Drinking
Water Regulations: Disinfectants and
Disinfection Byproducts. Federal
Register, 63:241:69389.
Yee, R B. & Wadowsky, R.M., 1982.
Multiplication of Legionella Pneumophila
in Unsterilized Tap Water. Applied
and Environmental Microbiology,
Zhang, A.; McCann, C.; Hanrahan, J.;
Jencson, A.; Joyce, D.; Fyffe, S.; S
Piesczynski.; Hawks, R.; Stout, J.E.; Yu,
V.L.; & Vidic, R.D., 2009. Legionella
Control by Chlorine Dioxide in Hospital
Water Systems. Journal AWWA,