30
Health and sanitation
Ultraviolet disinfection
for legionella control
By Environmental Protection Agency, document EPA 810-R-16-001
UV disinfection is a well-established treatment technology for
inactivating pathogens present in the environment.
In the drinking water context, UV disinfection was
initially most widely used in Europe, with hundreds
of installations in place by 1985 (USEPA, 2006c).
In North America, UV disinfection has been more
widely employed in drinking water applications
since 2000 to address health concerns associated
with cryptosporidium.
Characterisation of effectiveness
against legionella
There are several important lessons from installations of
UV disinfection in hospital settings and UV installations
in general. UV disinfection can be effective at controlling
legionella in facility piping (Hall et al.,2003; Franzin et al.,
2002; Liu et al., 1995). In the case of one new facility,
a UV disinfection unit was installed on the incoming
water supply, and none of the 930 cultures of hospital
water were positive. In addition, there were no confirmed
hospital-acquired legionella infections over a 13-year
study period (Hall et al., 2003).
UV is only effective at inactivating legionella in the water
that flows through the UV reactor. For existing facilities
with legionella present in the piping systems downstream
of a UV reactor, supplemental controls such as thermal
treatment or chemical disinfection will be necessary.
UV reactors need to be maintained to remain effective.
The quartz sleeves that house the reactors can be
fouled by iron, manganese, calcium carbonate, or other
deposits that decrease UV output. Lamps and other
reactor components also need to be replaced periodically
in order to maintain treatment effectiveness. Fouling of
the UV lamps was found to decrease effectiveness of the
UV treatment. Liu et al. (1995) added filters to prevent
scaling on UV lamps installed near the point of use in a
hospital’s cold and hot water systems. After treatment
with superheat/flush and shock chlorination, and
installation of filters to remove particles that foul the UV
lamps, the UV intensity of the lamps remained at 100%
throughout the experiment and the showers remained
legionella-free for a period of three months.
Relatively low UV doses appear to inactivate L. pneumophila
(Exhibit 2-1). A dose of 1mJ/cm 2 was found adequate,
in a recirculating model system, to achieve 99% (2-log)
reduction in six different legionella species. (Gilpin et
al., 1985).
Exhibit 2-1: UV doses (mJ/cm2) for inactivation of L. pneumophila
L. pneumophila
strain Lamp
type 1-log 2-log 3-log 4-log Reference
Philadelphia Type 2 LP 0.92 1.84 2.76 No data Antopol and Ellner,
1979
Philadelphia 1
(no light repair) LP 0.5 1.0 1.6 No data Knudson, 1985
Philadelphia 1
(with light repair) LP 2.3 3.5 4.6 No data Knudson, 1985
Philadelphia 1
ATCC33152 LP 1.6 3.2 4.8 6.5 Oguma et al., 2004
Philadelphia 1
ATCC33152 MP 1.9 3.8 5.8 7.7 Oguma et al., 2004
Notes:
LP = Low-pressure lamps, which have a single output of UV peaking around a wavelength of 254 nanometers.
MP = Medium-pressure lamps, which have polychromatic (or broad spectrum) output of UV at m