· UV disinfection provides no residual disinfection in the
distribution system that protects’ against recontamination (USEPA,
2003). However, chlorine or chloramines can be provided for this and
the EPA requires a chlorine or chloramine residual in the distribution
system regardless of whether UV disinfection is used.
· Poor water quality can reduce the effectiveness of UV disinfection.
Turbidity and dissolved or suspended matter may shield microorganisms
from UV radiation or scatter UV light to prevent it from reaching
the target microorganism (USEPA, 2003 ; Maggi et al., 2001). Some
organic and inorganic compounds may absorb UV energy reducing transmittance
(Maggi et al., 2001). When ferric chloride is used as a coagulant,
the residual iron can absorb UV light, lowering the UV transmittance
and increasing power costs (Maggi et al., 2001). Note, the LT2ESWTR
and the draft Ultraviolet Disinfection Guidance Manual recommend
a post-filter installation so that these should normally be lesser
concerns.
· Inactivation applies a less than lethal UV dose that is
designed to render Cryptosporidium non-reproducible. Damage
to RNA/DNA by sub-lethal doses of UV light may be repaired through
two major mechanisms: photoreactivation and dark repair. Photoreactivation
occurs when cells are exposed to visible light after UV disinfection.
The light activates enzymes that repair the thymine dimers. Dark repair
enzymes excise the dimerized pyrimidine bases and reinsert undimerized
bases (Ajy et al., 2001). However, this possibility is not believed
to be a significant concern for properly designed and operated systems.
· Configuration of a water treatment plant or the available
head may not allow for the installation of UV reactors between filters
and the clear well (Bishop et al., 2003). This may necessitate UV
installation after the high service pumps which is an undesirable
location (USEPA, 2003).
· Lamp fouling is another concern. Lamp fouling is the buildup
of chemical and biological films on the lamp sleeve surface, because
it reduces the amount of UV energy reaching the water. Water hardness
can create problems with keeping lamp sleeves clean (USEPA, 2003).
According to Mackey et al. and Cushing as reported in the draft UV
Disinfection Guidance Manual (2003), standard cleaning protocols and
wiper frequencies (one sweep every 15 min.) are adequate for total
calcium and hardness less than 140 mg/L and iron less than 0.1 mg/L
at sites tested.
· Since lamps contain mercury, their breakage is a potential
concern. If both the lamp and sleeve were broken, mercury would be
released into the drinking water. Potential causes of on-line lamp
breakage include debris breaking the lamp, overheated lamps, temperature
differentials between the water and the lamp, pressures exceeding
sleeve tolerances, electrical surges, improper handling and /or maintenance
of lamp and sleeve, misaligned cleaning mechanism, and improper design
and maintenance. While this is a concern, actual mercury releases
are uncommon. If the potential causes of on-line lamp breakage are
addressed in the design, operation, and maintenance of a UV disinfection
facility, the potential for release of mercury should be minimal.
· UV disinfection does not provide inactivation of viruses
at the dosage levels required for the inactivation of Cryptosporidium.
Of the pathogens tested in drinking water, viruses are the most resistant
to UV disinfection (USEPA, 2003). The UV doses for the inactivation
of viruses are based on the dose-response of adenovirus because it
appears to be the most UV resistant virus of those tested and is a
widespread waterborne pathogen (USEPA, 2003). In general the UV dose
required for 3-log inactivation of viruses ranges from 3 to 18 times
that required for Cryptosporidium. If viruses are of concern
in an individual water treatment system, the engineer and utility
may need to consider another disinfectant.