3.
Planning
and
Design
Aspects
for
UV
Installations
This
chapter
discusses
the
key
planning
and
design
features
for
UV
installations.
The
planning
section
helps
identify
the
parameters
and
constraints
to
be
considered
prior
to
design
of
the
UV
installation,
and
the
design
section
presents
factors
that
should
be
considered
during
detailed
design.

The
focus
of
Chapter
3
is
UV
disinfection
implementation
issues,
not
the
determination
of
whether
UV
disinfection
is
the
most
appropriate
technology.
Throughout
Chapter
3,
it
is
assumed,
unless
otherwise
stated,
that
the
water
to
be
disinfected
is
filtered
water
meeting
applicable
regulatory
requirements.
Appendices
G,
H,
and
I
provide
additional
information
on
unfiltered,
ground
water,
and
small
systems,
respectively.
The
planning
and
detailed
design
for
any
UV
installation
is
site­
specific.
Given
the
wide
range
of
treatment
scenarios
that
are
possible,
a
document
of
this
nature
cannot
address
or
anticipate
all
possible
treatment
conditions.
The
information
presented
here
should
be
used
within
the
context
of
sound
engineering
judgment
as
it
can
be
applied
on
a
case­
by­
case
basis.
In
addition,
this
Guidance
Manual
was
written
with
the
understanding
that
UV
technology
will
continue
to
expand
and
evolve.

The
organization
of
this
chapter
is
presented
below
by
the
question
that
each
section
addresses.

 
What
are
the
goals
of
the
UV
installation?
.................................................
Section
3.1.1
 
What
are
the
potential
installation
locations?
.............................................
Section
3.1.2
 
What
design
parameters
need
to
be
defined?..............................................
Section
3.1.3
 
How
does
the
UV
reactor
selection
affect
design?
.....................................
Section
3.1.4
 
What
are
the
options
for
validation?........................................................
Section
3.1.4.3
 
How
should
potential
installation
locations
be
evaluated?
.........................
Section
3.1.6
 
What
are
the
existing
hydraulic
conditions
and
UV
installation
hydraulic
needs?.......................................................................................
Section
3.1.6.1
 
What
should
be
considered
when
estimating
the
process
footprint
of
the
UV
installation?
..............................................................
Section
3.1.6.2
 
How
can
the
installation
options
be
evaluated?..........................................
Section
3.1.7
 
What
are
the
options
for
UV
reactor
procurement?......................................
Section
3.2
 
What
are
the
options
for
addressing
hydraulic
constraints
and
what
are
the
critical
hydraulic
system
components?...................................
Section
3.3.1
UV
Disinfection
Guidance
Manual
3­
1
June
2003
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
2
June
2003
 
How
does
the
control
strategy
influence
the
design
of
the
process
instrumentation
and
control
for
the
UV
installation?
....................
Section
3.3.2
 
What
are
the
elements
in
the
process
instrumentation
and
control
system?
...........................................................................................
Section
3.3.3
 
What
are
the
necessary
electric
power
arrangements?
...............................
Section
3.3.4
 
What
elements
need
to
be
considered
for
the
UV
installation
layout?.........................................................................................................
Section
3.3.5
 
What
information
should
the
equipment
specification
include?.................
Section
3.3.6
 
What
are
the
necessary
drawings
and
specifications
for
the
UV
installation?
.................................................................................................
Section
3.3.7
 
What
should
be
reported
to
the
State
and
when?..........................................
Section
3.4
The
process
of
planning
and
designing
a
UV
installation
is
presented
as
a
flowchart
in
Figure
3.1.
In
the
United
States
to
date,
the
majority
of
the
utilities
undertaking
the
construction
of
UV
installations
have
pre­
purchased
the
UV
reactors
prior
to
design.
Therefore,
the
design
flowchart
is
based
on
the
pre­
purchase
of
the
UV
reactors
and
the
use
of
a
traditional
design­
bidbuild
approach
for
the
project.
Chapter
3
is
generally
organized
to
follow
the
flowchart.
UV
installations
can
be
successfully
constructed
using
any
of
the
equipment
procurement
and
contractor
selection
approaches
currently
used
within
the
industry.
It
is
the
utility's
and
engineer's
responsibility
to
select
the
most
appropriate
project
approach.
Whatever
approach
is
utilized,
the
planning
and
design
components
discussed
in
Chapter
3
should
be
addressed
even
though
the
actual
order
of
completion
may
vary.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
Figure
3.1
Flowchart
for
Planning,
Design,
and
Construction
of
UV
Installations1
Define
the
goals
of
the
UV
installation
and
identify
the
target
microorganism(
s)

Section
3.1.1
Identify
the
potential
UV
facility
Section
3.1.2
Determine
design
parameters

Water
quality

Flowrate

Power
quality
Section
3.1.1
Evaluate
potential
UV
equipment
Section
3.1.4
Evaluate
equipment
validation
options**

Section
3.1.4.2
Evaluate
operational
and
control
strategies
Section
3.1.5
Identify
alternative
UV
facility
locations
by
evaluating:


Hydraulic
constraints
and
requirements

Footprint

Existing
Infrastructure
Section
3.1.6
Compare
options
and
costs
and
select
UV
facility
location
Section
3.1.7
Evaluate
and
select
procurement
option
Section
3.2
Design
system
hydraulics
Section
3.3.1
Determine
operating
and
control
strategy
Section
3.3.2
Design
instrumentation
and
controls
Section
3.3.3
Design
electric
power
systems
Section
3.3.4
Develop
specifications
and
procure
equipment
Section
3.3.6
Complete
facility
layout
Section
3.3.5
Finalize
hydraulic
design
Section
3.3.7
Finalize
design
layout
Section
3.3.7
Finalize
P&
IC
and
electrical
design
Section
3.3.7
Develop
and
bid
UV
facility
drawings
and
specifications
Section
3.3.7
Report
to
Primacy
Agency
Section
3.4
CONSTRUCTION
PLANNING
DESIGN
3.1.4.3
2
1
Flowchart
is
based
on
pre­
purchase
of
UV
reactors
and
the
traditional
design­
bid­
build
approach
2
The
timing
of
UV
reactor
validation
testing
depends
on
whether
it
has
been
validated
off­
site
or
if
on­
site
validation
is
necessary.

UV
Disinfection
Guidance
Manual
3­
3
June
2003
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
4
June
2003
3.1
UV
Installations
Planning
The
planning
process
for
a
UV
installation
is
similar
to
the
process
that
would
be
employed
for
any
retrofit,
upgrade,
or
new
construction
project
at
a
water
treatment
plant
(
WTP).
In
the
planning
phase,
it
is
important
to
identify
alternatives
and
define
criteria
needed
to
select
the
appropriate
application
and
to
facilitate
detailed
design.
For
a
UV
installation,
this
includes
the
following
steps:

 
Defining
disinfection
goals
 
Identifying
potential
locations
for
UV
disinfection
 
Defining
design
parameters
 
Evaluating
potential
UV
reactors
 
Evaluating
control
strategies
 
Evaluating
hydraulic
factors
and
process
footprint
 
Preparing
preliminary
costs
and
selecting
an
installation
option
This
section
provides
planning
guidance
for
each
of
these
steps
with
a
focus
on
specific
elements
that
should
be
considered
for
UV
disinfection.

3.1.1
Defining
UV
Disinfection
Goals
A
comprehensive
disinfection
strategy
provides
multiple
barriers
to
reduce
microbial
risk
while
minimizing
disinfection
byproduct
(
DBP)
formation.
UV
disinfection
is
a
tool
that
can
contribute
to
a
comprehensive
disinfection
strategy
by
providing
a
cost­
effective
method
of
inactivating
target
pathogens
that
are
more
resistant
to
more
traditional
disinfection
methods.
The
specific
objectives
of
a
given
UV
installation
should
be
clearly
defined
during
the
planning
stages.
This
can
ensure
that
the
design
meets
the
utility's
and
the
State's
expectations
based
on
the
regulatory
requirements,
target
microorganism(
s),
and
the
overall
disinfection
strategy.
Chapter
1
presents
the
regulatory
requirements
that
must
be
met
for
the
overall
water
treatment
process
and
specific
requirements
for
UV
disinfection.

The
UV
doses
necessary
for
Cryptosporidium
and
Giardia
inactivation
are
lower
than
that
those
needed
to
inactivate
viruses.
Accordingly,
the
capital
costs
for
inactivating
Cryptosporidium
and
Giardia
should
be
lower.
One
study
estimated
capital
costs
for
Cryptosporidium
and
Giardia
inactivation
by
UV
disinfection
to
be
approximately
50
percent
lower
than
the
costs
associated
with
the
UV
inactivation
of
viruses
(
Cotton
et
al.
2002).
Therefore,
the
target
microorganism
and
inactivation
level
should
be
determined
early
in
the
planning
process.

Repair
of
UV
light­
induced
damage
is
discussed
in
section
2.3.2.
As
discussed
previously,
repair
has
not
been
observed
in
Cryptosporidium
and
viruses,
and
Giardia
only
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
5
June
2003
exhibited
repair
when
exposed
to
very
low
UV
doses
(
0.5
mJ/
cm2).
Therefore,
repair
of
UVinduced
damage
of
Cryptosporidium,
Giardia,
and
viruses
do
not
need
to
be
considered
in
the
UV
installation
design.
However,
bacteria
have
been
shown
to
repair
of
UV
damage.
The
residual
disinfectant
concentration
(
either
chlorine
or
chloramines)
in
the
distribution
system
will
most
likely
prevent
repair
of
UV
damage
in
bacteria.
Therefore,
microbial
repair
of
bacteria
also
does
not
affect
UV
installation
design.

To
a
degree,
UV
disinfection
can
replace
chemicals
used
to
disinfect
chlorine­
resistant
pathogens
(
e.
g.,
Cryptosporidium
and
Giardia),
thereby
reducing
DBP
formation.
However,
UV
disinfection
is
not
as
efficient
at
inactivating
viruses
as
more
traditional,
chlorine­
based
disinfection
processes.
Because
of
its
effectiveness
at
treating
viruses
and
the
need
to
maintain
a
disinfectant
residual
in
the
distribution
system,
some
chlorine­
based
disinfectant
(
chlorine
or
chloramines)
will
be
needed
even
if
UV
disinfection
is
implemented.
Also,
chemicals
that
serve
as
disinfectants
may
be
added
in
the
treatment
process
to
oxidize
other
constituents
present
in
the
water
(
e.
g.,
iron,
manganese,
or
taste
and
odor
causing
compounds).
Utilities
that
currently
add
a
chemical
disinfectant
prior
to
the
location
of
a
future
UV
installation
and
plan
to
curtail
the
use
of
such
chemicals,
following
implementation
of
UV
disinfection
should
assess
the
effect
that
a
reduction
in
pre­
oxidant
use
may
have
on
water
quality
at
the
point
of
UV
application.
Therefore,
a
utility
considering
a
change
in
disinfection
strategy
should
evaluate
all
water
quality
goals
to
ensure
they
are
met
and
must
prepare
a
disinfection
benchmark
as
discussed
in
Chapter
1.

3.1.2
Identifying
Potential
Locations
for
UV
Installations
It
is
strongly
recommended
that
the
UV
disinfection
process
be
placed
after
filtration.
Although
UV
disinfection
can
potentially
be
applied
anywhere
along
the
treatment
train
from
the
raw
water
intake
to
after
high­
service
pumping,
there
are
significant
drawbacks
to
placing
the
UV
installation
upstream
of
filtration
in
conventional
WTPs.
Prior
to
filtration,
UV
absorbance
at
254
nm
(
A254)
is
higher
(
UV
transmittance
(
UVT)
is
lower)
due
to
higher
concentrations
of
natural
organic
matter,
turbidity,
and
particles.
Coagulation
can
enmesh
microorganisms
in
flocs
and
may
block
the
UV
light
from
reaching
the
microorganisms,
which
affects
the
UV
doseresponse
of
the
microorganism.
In
addition,
Long
Term
2
Enhanced
Surface
Water
Treatment
Rule
(
LT2ESWTR)
UV
dose
requirements
apply
only
to
post­
filter
and
unfiltered
supplies
that
meet
the
criteria
for
filtration
avoidance
(
40
CFR
141.729
(
d)).
Therefore,
this
section
focuses
on
the
post­
filtration
use
of
UV
disinfection.

This
section
presents
the
general
post­
filtration
locations
that
may
be
considered
for
the
UV
installation.
For
a
location
to
be
feasible,
the
UV
installation
hydraulic
needs
should
be
met
(
section
3.1.6.1)
and
the
equipment
must
physically
fit
in
the
proposed
location.
Hydraulic
profiles
and
preliminary
drawings
should
be
developed
for
each
location
under
consideration
to
address
these
controlling
criteria.
Also,
LT2ESWTR
requires
that
all
UV
reactors
be
validated
(
40
CFR
141.729
(
d)),
and
the
validation
protocol
(
Chapter
4)
recommends
specific
piping
configurations
for
both
validation
testing
and
UV
installation.
These
recommendations
for
inlet
and
outlet
conditions
can
affect
the
feasibility
of
the
potential
locations.
Detail
on
the
recommended
inlet
and
outlet
hydraulics
for
both
validation
and
installation
is
given
in
section
3.1.4.3.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
6
June
2003
3.1.2.1
Combined
Filter
Effluent
Installation
(
Upstream
of
Clearwell)

A
combined
filter
effluent
installation
is
defined
here
as
the
application
of
UV
disinfection
to
the
filter
effluent
after
it
has
been
combined
(
as
opposed
to
individual
filters)
and
prior
to
the
clearwell
as
shown
in
Figure
3.2.
This
installation
is
typically
in
a
separate
building.
Of
the
three
options
described,
the
combined
filter
installation
is
generally
preferred
when
conditions
permit.

Figure
3.2
Schematic
for
UV
Installation
Upstream
of
Clearwell
Source
Water
Rapid
Mix
Flocculation
Sedimentation
Basin
Filters
UV
Disinfection
Clearwell
To
Distribution
System
There
are
several
advantages
to
this
type
of
design
and
installation:

 
The
UV
reactor
operation
is
largely
independent
of
the
operation
of
individual
filters,
which
provides
flexibility
for
design
and
operation.

 
If
the
entire
UV
installation
failed,
a
WTP
could
still
provide
disinfection
by
adding
a
chemical
disinfectant
to
the
clearwell.
(
Note
that
backup
chemical
disinfection
may
not
provide
Cryptosporidium
inactivation.)

 
Surge
and
pressure
issues
that
are
concerns
with
UV
reactors
installed
immediately
downstream
or
upstream
of
high
service
pumps
(
HSPs)
are
usually
not
an
issue
for
this
installation
location.

 
Because
the
UV
installation
will
typically
be
constructed
in
a
new
building
for
this
installation
location,
there
may
be
greater
flexibility
in
maintaining
the
recommended
inlet
and
outlet
hydraulic
conditions
for
the
UV
reactors
(
section
3.1.6.1).

The
primary
disadvantages
of
this
type
of
installation
are
that
an
additional
building
may
be
necessary
and
that
piping
and
fittings
may
result
in
higher
headloss
than
alternative
configurations.

3.1.2.2
Individual
Filter
Effluent
Piping
Installation
Individual
filter
effluent
piping
installations
are
defined
here
as
installations
with
UV
reactors
dedicated
to
each
individual
filter
effluent
pipe.
The
installation
is
typically
within
the
existing
filter
gallery.
Figure
3.3
schematically
represents
this
type
of
installation.
The
main
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
7
June
2003
advantage
of
this
installation
is
that
a
new
building
would
not
be
necessary,
which
may
lower
construction
costs.

Figure
3.3
Schematic
of
Individual
Filter
Effluent
Piping
Installation
in
Filter
Gallery
Source
Water
Rapid
Mix
Flocculation
Sedimentation
Basin
Filters
UV
Disinfection
Clearwell
To
Distribution
System
HSPs
However,
there
are
several
disadvantages
to
this
installation
location.
Many
filter
galleries
do
not
have
sufficient
space
within
existing
effluent
piping
to
accommodate
a
UV
reactor.
The
existing
piping
may
also
put
constraints
on
how
the
UV
reactor
is
validated
because
of
the
unique
inlet
and
outlet
conditions
that
may
be
present
(
section
3.1.6.1).
In
addition
to
accommodating
the
UV
reactors,
there
needs
to
be
sufficient
space
in
the
filter
gallery
or
a
nearby
area
for
the
control
panels
and
electrical
equipment.
Access
to
existing
equipment
may
be
impaired
by
the
UV
reactor,
and
access
to
UV
reactor
components
for
maintenance
may
be
more
restricted
than
for
a
combined
filter
effluent
installation.
Also,
the
environmental
conditions
(
e.
g.,
moisture)
in
the
filter
gallery
may
not
be
appropriate
for
the
installation
of
the
UV
reactors,
associated
control
panels,
and
electrical
equipment
without
improvements
to
the
heating,
ventilating,
and
air
conditioning
system
for
the
area.

The
in­
line
installation
may
also
complicate
treatment
plant
operations
and
limit
operational
flexibility
as
described
below.

 
In
general,
this
option
results
in
an
increased
number
of
UV
reactors
compared
to
a
combined
filter
installation
because
the
number
of
filters
dictates
the
number
of
UV
reactors.
This
may
increase
operation
and
maintenance
costs
in
comparison
to
the
combined
filter
effluent
installation
where
the
number
of
UV
reactors
is
determined
by
the
design
flow,
water
quality
constraints,
UV
reactor
capacity,
and
redundancy
needs.

 
The
increased
headloss
of
the
UV
reactors
may
affect
the
operation
of
the
filters
and
the
clearwell.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
8
June
2003
 
With
one
UV
reactor
for
each
filter,
the
operation
of
each
filter
will
be
dependent
on
the
reliable
operation
of
each
UV
reactor
and
vice
versa.

 
The
UV
reactor
operation
during
a
filter
backwash
can
complicate
UV
reactor
operations.
The
lamp
cooling
need
to
addressed
if
it
remains
energized
during
a
backwash
because
the
lamps
should
not
be
energized
in
stagnant
water
or
air.
If
a
UV
reactor
is
off
during
a
backwash,
the
flow
during
the
UV
reactor
warm­
up
(
section
3.1.3.3)
is
off­
specification,
which
may
cause
problems
with
exceeding
offspecification
requirements
and
recommendations
(
section
3.1.3).

3.1.2.3
UV
Disinfection
Downstream
of
the
Clearwell
It
may
be
possible
for
a
WTP
to
build
the
UV
installation
after
the
clearwell
either
upstream
or
downstream
of
the
high
service
pumps
(
Figure
3.4).
In
many
WTPs,
the
HSPs
take
water
directly
from
the
clearwell,
limiting
space
and
the
availability
of
suitable
piping
for
installation
of
the
UV
installation
upstream
of
the
HSPs.
Installation
downstream
of
the
HSPs
may
provide
greater
space
and
flexibility
in
locating
the
UV
reactors.
Either
configuration
may
be
advantageous
if
there
is
insufficient
space
or
head
to
allow
installation
of
the
UV
reactors
between
the
filters
and
the
clearwell;
however,
there
are
significant
disadvantages
to
these
options.

Figure
3.4
UV
Disinfection
Downstream
of
High
Service
Pumps
Source
Water
Rapid
Mix
Flocculation
Sedimentation
Basin
Filters
UV
Disinfection
Clearwell
To
Distribution
System
HSPs
UV
installations
located
downstream
of
the
clearwell
will
experience
greater
fluctuations
in
flowrate
since
actual
flowrates
are
more
closely
matched
to
system
demand
changes.
This
may
increase
the
UV
reactor
size
or
more
UV
reactors
to
accommodate
the
flow
fluctuations.

In
post­
HSP
installations,
the
water
will
be
at
distribution
system
pressure.
The
UV
reactor
housing
may
need
to
be
reinforced
because
of
these
high
pressures,
which
would
increase
the
cost
of
the
UV
reactors.
In
addition,
these
locations
are
more
prone
to
water
hammer
because
of
their
proximity
to
the
HSPs
and
subsequent
high
pressures,
which
could
lead
to
sleeve
damage.
If
a
lamp
sleeve
is
damaged,
the
enclosed
lamp
may
break,
releasing
mercury
into
the
water.
Hydropneumatic
tanks
or
pressure
relief
valves
may
be
necessary
in
this
installation
location
to
avoid
water
hammer.
This
issue
is
discussed
in
more
detail
in
section
3.1.6.1
and
section
N.
2.1.3.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
9
June
2003
A
UV
reactor
located
after
the
HSPs
will
reduce
the
discharge
pressure
to
the
distribution
system
In
summary,
UV
installations
downstream
of
the
clearwell
are
not
recommended
because
of
the
i
3.1.3
efining
Design
Parameters
Water
quality,
lamp
fouling/
aging
factor,
flowrate,
and
power
quality
affect
the
sizing
of
the
UV
V
reactors
are
required
to
be
validated
by
LT2ESWTR
to
demonstrate
the
UV
installa
blishes
ng
is
o
the
extent
practical,
UV
reactors
should
be
designed
with
process
monitoring
and
control
ion
d
limit
The
UV
reactors
are
off­
specification
when
any
of
the
following
conditions
occur:

The
flow,
UV
intensity,
or
lamp
status
is
outside
of
the
validated
range.

The
UVT
or
UV
intensity
is
outside
of
the
validated
range
(
if
the
UV
intensity
and
 
The
calculated
dose
is
outside
of
the
validated
range
at
a
given
flow
(
if
the
calculated
 
All
UV
lamps
in
all
UV
reactors
are
off
because
of
a
power
interruption
or
power
,
and
a
UV
installation
located
between
the
clearwell
and
HSPs
will
reduce
the
suction
head
available
for
the
pumps.
As
a
result,
discharge
pressures
and
storage
utilization
could
be
impacted
at
these
two
locations.

ncreased
potential
for
adverse
pressure
conditions
within
the
UV
reactor
and
the
increased
reliability
and
size
considerations.
In
general,
these
installations
should
only
be
considered
if
the
combined
filter
effluent
and
in­
line
filter
effluent
locations
are
not
feasible.

D
reactors
and
associated
support
facilities.
These
design
parameters
need
to
be
determined
to
ensure
compliance
with
LT2ESWTR
requirements.

U
tion
achieves
the
required
UV
dose
(
40
CFR
141.729(
d)).
Validation
testing
esta
the
conditions
under
which
the
UV
reactors
must
be
operated
to
ensure
the
required
dose
delivery
(
40
CFR
141.729(
d)).
Off­
specification
is
defined
as
a
UV
reactor
that
is
operati
outside
of
its
validated
limits.
(
For
example,
the
UV
reactor
is
operating
with
a
flowrate
that
higher
than
the
UV
reactor
was
validated.)

T
components
(
e.
g.,
alarms,
shut­
off
valves)
to
prevent
water
from
entering
the
distribut
system
when
a
UV
reactor
is
operating
outside
of
validated
conditions.
Unfiltered
systems
that
use
UV
disinfection
to
meet
the
Cryptosporidium
treatment
requirement
of
the
LT2ESWTR
must
demonstrate
that
at
least
95
percent
of
the
water
delivered
to
the
public
during
each
month
is
treated
by
UV
reactors
operating
within
validated
limits
(
40
CFR
141.721(
c)(
2)).
Or
in
other
words,
the
UV
reactor
cannot
be
off­
specification
for
more
than
5
percent
of
the
water
delivere
to
the
public.
The
LT2ESWTR
does
not
state
an
off­
specification
requirement
for
filtered
systems;
however,
States
may
establish
requirements
for
their
filtered
systems,
including
a
for
off­
specification
operation.

 

 

UVT
setpoint
approach
is
used
(
section
3.1.5)).

dose
approach
is
used
(
section
3.1.5)).

quality
problem,
and
water
is
flowing
through
the
reactors.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
10
June
2003
It
is
important
to
determine
the
appropriate
design
values
for
water
quality,
lamp
nts
and
ctors
3.1.3.1
Assessing
Water
Quality
As
highlighted
in
Chapter
2,
the
following
water
quality
parameters
are
the
primary
parame
Parameters
that
affect
UV
dose
delivery
0
­
300
nm
(
germicidal
range)

 
Parameters
that
typically
determine
sleeve
and
UV
intensity
sensor
fouling
emperature
It
should
be
reiterated
that
this
manual
is
focused
on
post­
filtration
applications;
therefo
s)
and
y
Water
quality
data
should
be
collected
from
locations
that
are
representative
of
the
potenti
and
on
e
The
four
main
considerations
for
assessing
water
quality
are
A254,
fouling
potential,
lamp
fouling
fouling/
aging
factor,
flowrate,
and
power
quality
because
of
these
LT2ESWTR
requireme
recommendations.
If
the
design
parameters
are
not
chosen
conservatively
enough,
the
UV
reactors
may
be
operating
off­
specification
and
be
out
of
compliance.
However,
overly
conservative
design
values
may
result
in
unnecessarily
large
UV
reactors
or
more
UV
rea
than
necessary.

ters
that
affect
UV
installation
planning
and
design:

 

­
UV
absorbance/
transmittance
from
20
­
Upstream
chemical
additives
­
Calcium
­
Alkalinity
­
Hardness
­
Iron
­
pH
­
Lamp
t
re,
it
is
assumed
that
turbidity
is
low
(
1
nephelometric
turbidity
units
(
NTU)
or
les
results
in
insignificant
particle
effects
on
UV
dose
delivery
(
Linden
et
al.
2002b).
It
is
also
assumed
that
the
water
meets
applicable
maximum
contaminant
levels
(
MCLs)
and
secondar
MCLs.

al
UV
installation
location.
The
duration
of
sampling,
number
of
samples
collected,
data
analyses
used
to
evaluate
water
quality
for
UV
disinfection
are
similar
to
the
approaches
taken
for
other
water
treatment
technologies.
The
data
collection
frequency
should
be
a
based
flow
variability,
the
consistency
of
the
source
and
treated
water
qualities,
and
the
potential
for
obtaining
cost
and
energy
savings
by
refining
the
design
criteria.
The
extent
of
water
quality
data
collected
should
be
left
to
the
discretion
of
the
utility
and
the
designer
based
on
experienc
and
professional
judgment.
States
may
desire
to
provide
input
on
data
collection
needs.

/
aging
factor
and
upstream
chemical
impacts.
Each
of
these
is
discussed
in
the
following
sections.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
11
June
2003
UV
Absorbance
As
discussed
in
Chapter
2,
the
A254
1
of
the
water
directly
influences
UV
dose
delivery.
The
A254
254
254
254
data
should
be
evaluated
to
select
a
design
A
value.
The
design
A
along
with
the
specified
UV
dose
and
flowrate
will
be
used
by
the
UV
manufacturer
to
determine
the
appropriate
UV
reactor.
In
addition,
UV
manufacturers
may
use
the
A
range
at
the
WTP
to
determine
the
turndown
(
i.
e.,
power
modulation)
needs
of
the
UV
reactors.

Overly
conservative
design
A254
values
(
i.
e.,
low
UVT)
can
result
in
over­
design
and
increased
capital
costs.
Conversely,
inappropriately
low
design
A254
values
can
result
in
UV
reactor
operation
outside
the
validated
operating
range
and
potential
non­
compliance.
As
with
most
designs,
the
larger
the
data
set,
the
more
refined
the
final
design
can
be.
A
utility
with
very
stable
A254
might
only
need
one
or
two
months
of
data,
while
a
utility
that
experiences
seasonal
changes
would
benefit
from
more
frequent
data
collection
during
seasonal
events
and
over
a
longer
recording
period.

The
A254
sampling
plan
should
include
collection
of
A254
2
measurements
in
grab
samples
or
continuously
with
an
on­
line
A254
monitor.
If
A254
peaks
occur
regularly
during
the
Spring
and
Fall,
increased
sampling
frequency
during
these
periods
will
better
capture
the
magnitude
and
duration
of
the
peaks.
If
different
sources
or
combination
of
sources
(
i.
e.,
blending)
are
used
during
the
year,
the
A254
of
the
potential
source
water
blends
should
be
characterized
to
properly
identify
the
appropriate
water
quality
conditions.
In
addition,
the
maximum
A254
may
not
correspond
to
the
period
of
maximum
water
production.
The
relationship
between
seasonal
production
rates
and
A254
data
should
be
considered
when
developing
design
criteria.

A
cumulative
frequency
(
CF)
diagram
of
the
A254
data
may
assist
the
utility
in
determining
its
design
A254
value.
Figure
3.5
presents
a
CF
diagram
for
three
filtered
waters;
the
CF
percentile
(
x­
axis)
shows
the
percentage
of
the
dataset
that
is
lower
than
a
given
value
of
A254
over
the
data
collection
period.
For
example,
if
the
90th
percentile
A254
is
0.043
cm­
1,
then
90
percent
were
lower
and
only
10
percent
of
the
measurements
were
higher
than
0.043
cm­
1
over
the
period
of
record.

1
A254
in
this
section
implies
A254
measurement
specifically
at
254
nm
unless
otherwise
noted
2
A254
measurements
for
developing
the
design
basis
for
UV
disinfection
systems
should
be
performed
on
unfiltered
samples,
not
with
the
0.45
µ
m
pre­
filtered
samples
typically
used
to
characterize
NOM.
However,
if
only
measurements
that
have
been
filtered
are
available,
they
still
provide
valuable
information.
It
should
be
noted
that
pre­
filtered
measurements
are
typically
biased
low
(
in
terms
of
absorbance),
but
this
bias
is
generally
minimal.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
12
June
2003
Figure
3.5
Example
CF
Diagram
for
Three
Filtered
Waters
0.000
0.020
0.040
0.060
0.080
0.100
0.120
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%

Cumulative
Frequency
Percentile
UV
Absorbance
at
254
nm
(
cm­
1)
Filtered
Water
1
Data
Collected
from
May
1991
to
July
2001
Filtered
Water
2
Data
Collected
from
January
1991
to
March
1998
Filtered
Water
3
Data
Collected
from
February
1993
to
July
2001
In
Figure
3.5,
the
A254
data
for
Filtered
Waters
1,
2,
and
3
display
different
characteristics.
The
A254
values
for
Filtered
Water
1
are
relatively
constant
between
the
5th
and
85th
percentiles,
indicating
consistent
water
quality
approximately
80
percent
of
the
time.
Values
above
the
85th
percentile
increase
to
a
plateau,
and
then
increase
again
above
the
95th
percentile.
A254
data
for
Filtered
Waters
2
and
3
exhibit
greater
variability
by
the
gradually
increasing
slope
between
the
5th
and
95th
percentile.
Selecting
an
appropriate
design
A254
value
for
these
waters
depends
on
an
assessment
of
this
variability
as
compared
to
the
percentage
of
time
that
offspecification
water
could
be
delivered.

For
example,
a
CF
percentile
of
95
percent
would
most
likely
meet
the
off­
specification
criteria
for
unfiltered
systems.
However,
a
95
percent
CF
percentile
may
be
overly
conservative,
depending
on
the
flow
observed
at
the
planned
UV
installation.
Therefore,
plotting
the
A254
with
the
WTP
flow
can
indicate
if
high
A254
and
high
flow
co­
occur,
which
would
be
the
worse
case
water
quality
condition.
Figure
3.6
presents
Filtered
Water
3'
s
flow
and
A254
variation
and
illustrates
a
seasonal
variation
in
A254.
For
this
example
WTP,
the
high
A254
typically
occurs
during
the
high
flow
period;
therefore,
a
more
conservative
design
A254
of
0.077
cm­
1
(
84%
UVT),
which
is
a
CF
percentile
of
95
percent,
may
be
warranted
for
Filtered
Water
3.
However,
a
less
conservative
design
A254
would
be
appropriate
if
the
high
A254
occurred
during
flows
less
than
the
design
flow
(
e.
g.,
average
flow)
because
the
UV
reactors
should
have
enough
turndown
(
e.
g.,
power
modulation)
to
accommodate
high
A254
at
lower
flows
than
the
design
flow.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
13
June
2003
Figure
3.6
Example
Flow
and
UV
Absorbance
(
at
254
nm)
Data
20
60
100
140
180
220
Jan­
97
Apr­
97
Jul­
97
Oct­
97
Jan­
98
Apr­
98
Jul­
98
Oct­
98
Jan­
99
Apr­
99
Jul­
99
Oct­
99
Jan­
00
Apr­
00
Jul­
00
Oct­
00
Jan­
01
Apr­
01
Jul­
01
Plant
Flow
(
mgd)

0.00
0.04
0.08
0.12
0.16
0.20
UV
Absorbance
at
254
nm
(
cm­
1)
Flow
UV
Abs
at
254
nm
Filtered
Water
3
WTP
Capacity:
220
mgd
The
design
A254
(
e.
g.,
a
CF
percentile)
also
should
be
a
function
of
the
utility's
preferred
level
of
conservatism
and
the
site­
specific
A254and
flow
data.
The
UV
reactor
sizing
and
cost
are
not
directly
proportional
to
A254
but
will
increase
for
increased
A254
design
values.
However,
by
evaluating
the
CF
plot
and
collaborating
with
the
UV
manufacturer
to
assess
the
cost
implications
of
using
a
lower
A254
value,
the
utility
and
designer
can
select
the
most
appropriate
design
A254
for
the
water
quality
and
disinfection
objectives
of
the
project.

Typically,
the
UV
manufacturers
work
in
terms
of
UVT;
therefore,
the
design
A254
is
typically
converted
to
a
design
UVT3.
Because
UV
manufacturers
use
UVT
in
their
design
and
control
of
the
UV
reactors,
the
remainder
of
this
chapter
will
use
UVT
as
opposed
to
A254.

The
spectral
absorbance
of
the
water
over
a
range
of
wavelengths
(
200
­
400
nm)
should
also
be
collected,
especially
if
medium
pressure
(
MP)
reactors
are
being
considered.
MP
lamps
emit
light
at
a
range
of
wavelengths
across
the
200
nm
to
300
nm
range.
The
UV
absorbance
of
water
varies
with
wavelength,
typically
decreasing
with
increasing
wavelength.
As
such,
the
attenuation
of
UV
light
in
a
UV
reactor,
the
corresponding
disinfection
performance,
and
the
UV
intensity
sensor
response
depend
on
the
absorbance
at
each
of
the
emitted
wavelengths.
Sitespecific
spectral
absorbance
can
be
used
to
model
MP
reactors
and
may
be
incorporated
into
UV
dose
monitoring
and
control
systems
by
some
UV
manufacturers.
Spectral
absorbance
may
3
254
10
100
(%)
A
UVT
 
 
=

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
14
June
2003
exhibit
seasonal
variation;
therefore,
spectral
absorbance
should
be
collected
at
different
times
during
the
year
to
assess
this
variation.
Also,
the
spectral
absorbance
may
be
used
to
determine
the
appropriate
UV­
absorbing
chemical
for
validation
of
the
UV
reactors
that
will
be
installed,
which
is
discussed
in
section
4.3.3.2.

Fouling
Potential
The
rate
of
fouling
and
the
corresponding
frequency
of
sleeve
cleaning
depend
on
hardness,
alkalinity,
lamp
temperature,
pH,
and
certain
inorganic
constituents
(
e.
g.,
iron
and
calcium).
Fouling
is
typically
caused
by
precipitation
of
compounds
with
low
solubility
or
compounds
where
the
solubility
decreases
as
temperature
increases
(
e.
g.,
CaCO3).
If
significant
seasonal
shifts
in
any
of
the
parameters
are
expected,
these
trends
should
be
captured
in
the
monitoring
period.
Again,
a
CF
diagram
may
assist
in
the
selection
of
the
appropriate
design
criteria.

While
the
specific
rate
of
fouling
and
optimal
cleaning
protocol
for
any
given
application
cannot
currently
be
predicted,
a
proper
cleaning
protocol
and
sleeve­
fouling
factor
can
be
adequately
estimated
for
most
water
sources
without
pilot­
or
demonstration­
scale
testing
and
then
adjusted
during
normal
operation.
Extensive
data
have
been
generated
from
pilot­
scale
testing
on
waters
of
low
to
moderate
hardness
and
iron
content
(
Mackey
et
al.
2001
and
Mackey
and
Cushing
2003).
At
total
and
calcium
hardness
levels
below
140
mg/
L
and
low
iron
(
less
than
0.1
mg/
L),
standard
cleaning
protocols
and
wiper
frequencies
(
one
sweep
every
15
minutes
to
an
hour)
were
more
than
adequate
to
address
the
impact
of
sleeve
fouling
at
the
sites
tested.
At
sites
with
hardness
or
iron
that
exceed
these
levels,
it
may
be
advantageous
to
evaluate
fouling
rates
on
a
site­
specific
or
worst
case
basis
via
pilot
or
demonstration
testing
(
described
in
Appendix
J)
or
during
UV
reactor
start­
up
(
section
5.1)
to
identify
how
best
to
address
fouling.

Although
fouling
is
not
expected
to
be
a
significant
problem
for
most
utilities,
the
listed
water
quality
parameters
(
page
3­
10)
should
be
monitored
prior
to
designing
the
UV
installation
unless
adequate
water
quality
data
are
available.
It
is
important
to
provide
these
data
to
the
UV
manufacturer
to
assist
them
in
a
qualitative
assessment
of
the
fouling
potential
for
their
UV
reactors
and
to
assist
the
designer
in
determining
what
cleaning
system
should
be
specified.
In
addition,
the
lamp
fouling/
aging
factor
will
depend
on
the
initial
assessment
of
potential
fouling,
which
is
discussed
in
the
next
section.

Lamp
Fouling/
Aging
Factor
Sleeve
fouling,
lamp
aging,
and
UV
intensity
sensor
window
fouling
(
if
applicable)
affect
long­
term
UV
reactor
performance.
Accumulation
of
foulants
on
the
lamp
sleeve
surface
can
reduce
transmittance
of
the
lamp
energy
to
the
water.
The
rate
of
fouling
depends
on
the
factors
discussed
in
the
previous
section.
In
addition,
lamp
output
decreases
over
time
due
to
its
physical
aging.
The
rate
at
which
lamp
output
will
decrease
is
a
function
of
the
lamp
physical
characteristics,
lamp
hours
in
operation,
number
of
on/
off
cycles,
and
power
applied
per
lamp
length.
In
MP
reactors,
UV
lamp
aging
can
also
result
in
a
change
in
the
spectral
output
over
time.
Lamp
aging
is
discussed
in
detail
in
section
A.
3.1.6.

A
reduction
in
lamp
output
results
in
a
reduction
in
UV
dose.
The
effects
of
these
parameters
are
typically
incorporated
into
the
UV
reactor
design
by
specifying
a
lamp
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
15
June
2003
fouling/
aging
factor,
which
includes
the
effects
of
both
sleeve
fouling
and
lamp
aging.
The
lamp
fouling/
aging
factor
will
be
site­
specific
and
based
on
the
assessment
of
fouling
described
previously
and
lamp
aging
information.
The
lamp
aging
characteristics
can
be
obtained
from
the
UV
manufacturer
and
should
be
certified
by
an
independent
third
party.
The
lamp
fouling/
aging
factor
is
used
by
the
manufacturer
to
assist
in
the
selection
of
the
appropriate
UV
reactor.
For
example,
if
a
0.5
lamp
fouling/
aging
factor
is
specified,
the
UV
manufacturer
will
choose
a
UV
reactor
the
appropriate
lamps
(
or
number
of
lamps)
where
the
specified
UV
dose
can
be
achieved
at
half
of
the
initial
UV
lamp
output
(
after
burn­
in)
with
all
the
lamps
energized
at
full
power.
The
lamp
fouling/
aging
factor
typically
ranges
from
0.5
(
NWRI
2000)
to
0.9.

The
lamp
fouling/
aging
factor
is
typically
specified
with
a
corresponding
guaranteed
UV
lamp
life
(
e.
g.,
5000
hours).
These
items
are
typically
specified
together
to
ensure
that
the
UV
lamp
replacement
frequency
does
not
occur
more
frequently
than
specified
by
the
guaranteed
lamp
life
given
the
specified
lamp
fouling/
aging
factor.
The
lamp
fouling/
aging
factor
can
be
estimated
based
on
the
designer's
experience
and
UV
manufacturer
input.
In
addition,
pilot
and
demonstration
tests
can
be
completed
to
estimate
the
lamp
fouling/
aging
factor
as
described
in
Appendix
J.

Selection
of
a
lamp
fouling/
aging
factor
and
a
guaranteed
lamp
life
value
is
a
trade­
off
between
maintenance
costs
(
the
frequency
of
lamp
replacement
or
chemical
cleans
necessary)
and
capital
costs
(
the
size
of
the
UV
reactors).
A
lower
lamp
fouling/
aging
factor
means
the
utility
will
have
less
frequent
lamp
replacements
because
the
UV
reactors
are
designed
with
higher
powered
lamps
or
more
lamps
to
achieve
the
necessary
UV
output
at
the
guaranteed
lamp
life.
However,
designing
a
UV
reactor
with
higher
powered
lamps
or
more
lamps
will
increase
the
size
of
the
needed
UV
reactor.
Thus,
the
use
of
a
fouling/
aging
factor
that
is
too
conservative
could
result
in
the
over­
design
of
the
UV
reactors.
Conversely,
the
use
of
a
lamp
fouling/
aging
factor
that
is
not
conservative
enough
may
result
in
the
underestimated
reduction
in
the
output
of
the
lamp
due
to
fouling/
aging
and
potentially
result
in
off­
specification
operation
or
more
frequent
lamp
replacement.

Impacts
of
Upstream
Treatment
Processes
Unit
processes
upstream
of
UV
reactors
can
have
a
significant
impact
on
the
UV
reactor
performance.
The
three
potential
ways
that
upstream
processes
may
affect
UV
performance
are
(
1)
to
increase
UVT
by
increasing
organics
removal
or
oxidizing
organics,
(
2)
to
decrease
UVT
because
certain
chemicals
will
absorb
UV
light,
and
(
3)
to
affect
the
lamp
sleeve
fouling
rate.

It
is
possible
to
increase
filtered
water
UVT
by
increasing
the
coagulant
dose;
however,
the
results
will
be
site­
specific.
In
one
study,
the
UVT
was
increased
from
80%
to
89%
by
increasing
the
alum
dose
from
15
to
45
mg/
L
(
Cushing
et
al
2001).
However,
the
UVT
increase
from
an
increased
alum
dose
should
be
considered
against
the
increased
alum
chemical
costs
and
sludge
production.
UVT
increases
would
also
probably
be
observed
if
other
iron
coagulant
and
poly­
aluminum
chloride
coagulant
doses
were
increased.

Properly
implemented,
ozone
disinfection
prior
to
UV
disinfection
has
the
potential
to
increase
the
UVT
from
oxidation
of
organic
matter.
Conversely,
ozone
disinfection
can
decrease
UVT
if
a
residual
ozone
concentration
is
present
in
the
UV
reactors.
If
the
ozone
residual
is
adequately
quenched,
a
net
increase
in
the
UVT
will
be
observed
(
Malley
2002);
an
example
of
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
16
June
2003
this
increase
for
an
unfiltered
water
is
shown
in
Figure
3.7.
If
a
UVT
increase
is
desired,
then
a
combination
of
coagulant
increase
and
ozone
disinfection
will
likely
give
the
greatest
UVT
increase
(
Cushing
et
al
2001).

Figure
3.7
Example
Effect
of
Pre­
ozonation
on
UV
Absorbance
if
Ozone
is
Quenched
Prior
to
UV
Disinfection
W
a
v
e
le
n
g
th
(
nm
)
2
0
0
2
2
0
2
4
0
2
6
0
2
8
0
3
0
0
3
2
0
UV
Absorbance
(
cm­
1)

0
.0
0
.1
0
.2
0
.3
0
.4
0
.5
P
re
o
z
o
n
a
te
d
U
V
In
flu
e
n
t
(
n
o
d
e
te
c
ta
b
le
o
z
o
n
e
re
s
id
u
a
l)
U
V
In
flu
e
n
t
­
N
o
o
z
o
n
e
Most
common
water
treatment
chemicals
themselves
will
not
significantly
impact
UVT.
The
following
common
water
treatment
chemicals
do
not
significantly
affect
UVT
at
typical
concentrations
present
in
filtered
water:
Alum,
aluminum,
ammonia,
ammonium,
zinc,
phosphate,
calcium,
hydroxide,
and
ferrous
iron
(
Fe+
3)
(
Cushing
et
al
2001).

However,
hypochlorite
(
ClO­),
ferric
iron
(
Fe+
2),
permanganate,
and
ozone
were
the
only
commonly
used
chemicals
examined
that
might
reduce
UVT
(
Cushing
et
al
2001)
as
described
below.

 
Residual
ClO­
has
only
a
slight
effect
on
UVT.
For
example,
a
ClO­
residual
of
3.5
mg/
L
will
cause
the
UVT
to
decrease
from
91%
to
90%
(
Cushing
et
al
2001).
However,
in
most
cases,
a
hypochlorite
residual
that
high
will
not
be
flowing
through
the
UV
reactor.

 
It
is
unlikely
that
ferric
iron
will
be
present
in
filtered
waters
because
ferric
iron
is
only
present
when
there
is
low
dissolved
oxygen.

 
Permanganate
is
a
strong
absorber
of
UV
light;
however,
it
is
typically
added
in
the
raw
water
to
oxidize
taste
and
odor
or
iron
and
manganese.
Therefore,
when
applied
to
raw
water,
there
should
not
be
a
significant
permanganate
concentration
in
the
filtered
water.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
17
June
2003
 
Ozone
residual
can
be
quenched,
and
then
the
UVT
will
not
be
decreased.
Care
should
be
taken
when
choosing
the
quenching
agent
because
one
popular
choice,
thiosulfate
(
often
used
in
the
form
of
calcium
thiosulfate),
is
a
strong
absorber
of
UV
light
(
section
A.
4.1.3,
Table
A.
5)
and
will
decrease
the
UVT.
Sodium
bisulfite,
an
alternative
to
calcium
thiosulfate,
will
not
significantly
impact
UVT.

The
possible
UVT
variation
from
upstream
processes
should
be
assessed
by
collecting
UVT
data
during
various
operating
conditions
(
e.
g.,
a
range
of
alum
doses)
that
are
typically
observed.
Potential
treatment
process
upsets
should
also
be
considered
in
the
water
quality
analysis
to
determine
the
extent
to
which
they
impact
the
design
UVT
and
cleaning
regime.

Some
unit
processes
that
use
metal­
based
coagulants
may
affect
the
rate
of
fouling;
these
effects
will
be
site­
specific.
Mackey
et
al.
(
2001)
found
that
iron
levels
less
that
0.1
mg/
L
could
be
adequately
cleaned
by
standard
protocols
as
described
previously.
In
addition,
lime
softening
has
been
shown
to
reduce
fouling
potential
(
Mackey
et
al.
2001).
Overall,
the
effect
of
upstream
coagulant
addition
and
residual
metals
should
be
considered
in
the
fouling
data
monitoring
described
previously.

3.1.3.2
Determining
Design
Flowrate
Flowrate
is
a
fundamental
design
parameter
that,
in
combination
with
water
quality,
UV
dose,
and
lamp
fouling/
aging
factor
determines
the
necessary
size
and
number
of
UV
reactors.
The
design
criteria
should
identify
the
average,
maximum,
and
minimum
flowrates
that
the
UV
reactors
will
experience.
Potential
methods
for
determining
the
design
flow
for
the
three
described
retrofit
locations
are
shown
in
Table
3.1.
In
addition,
potential
future
changes
in
plant
capacity
should
be
considered
when
determining
the
UV
installation
design
flow.

Table
3.1
Potential
Method
to
Determine
Design
Flow
Retrofit
Location
Design
Flow
Basis
Combined
filter
retrofit
Combined
rated
capacity
of
all
duty
filters1
Individual
filter
retrofit
Rated
design
flow
for
individual
filter
Post­
HSP
Rated
capacity
of
the
HSP
station
1Flow
does
not
include
redundant
filters
3.1.3.3
Assessing
Electrical
Power
The
sensitivity
of
UV
reactors
to
power
fluctuations
make
electrical
power
supply
a
critical
component
of
the
UV
installation
planning
and
design.
In
addition,
the
electrical
system
design
needs
to
ensure
that
the
UV
installation
will
meet
the
requirements
or
recommendation
of
operating
within
validated
conditions
(
i.
e.,
maximum
allowed
off­
specification).
Also,
it
is
impossible
to
meet
inactivation
goals
if
the
power
quality
causes
the
reactor
to
go
down
(
i.
e.,
no
disinfection)
for
longer
than
the
need
to
obtain
the
desired
treatment
level.
For
example,
if
a
2­
log
Cryptosporidium
inactivation
is
desired,
the
UV
reactors
cannot
be
down
while
more
than
1
percent
of
the
flow
passes
through
them.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
18
June
2003
UV
lamps
can
potentially
their
lose
arc
if
a
voltage
fluctuation,
power
quality
anomaly,
or
a
power
interruption
occurs.
For
example,
voltage
sags
that
vary
from
10
to
15
percent
from
normal
operating
conditions
for
as
low
as
2
to
5
cycles
(
0.03
to
0.08
seconds)
may
cause
UV
lamps
to
lose
their
arc.

Low
pressure
(
LP)
lamps
generally
can
return
to
full
operating
status
within
15
seconds
after
power
is
restored.
However,
low­
pressure
high
output
(
LPHO)
and
MP
reactors
that
are
more
typically
used
in
drinking
water
applications
exhibit
significant
restart
times
if
power
is
interrupted.
The
start­
up
and
restart
behavior
for
LPHO
and
MP
lamps
is
summarized
in
Table
3.2.

Table
3.2
Start
and
Restart
Times
for
LPHO
and
MP
Lamps
1
Lamp
Type
Cold
Start2
Warm
Start3
LPHO
2
min
warm­
up
+
4­
5
min
to
full
power
total
time:
6
 
7
minutes
2
min
warm­
up
+
2­
5
min
to
full
power
total
time:
4
 
7
minutes
MP
No
warm­
up
or
cool
down
+
5
min
to
full
power4
total
time:
5
minutes
5
min
cool
down
+
5
min
to
full
power4
total
time:
10
minutes
1
Information
shown
in
table
is
compiled
from
Calgon
Carbon,
Severn
Trent,
Trojan,
and
Wedeco.
2
A
cold
start
occurs
when
UV
lamps
are
started
when
they
have
not
been
operating
for
a
significant
period
of
time.
3
A
warm
start
occurs
when
UV
lamps
are
started
after
they
have
just
lost
their
arc
(
e.
g.,
due
to
voltage
sag).
4
60
percent
intensity
is
obtained
after
3
minutes.

The
effects
of
temperature
can
increase
or
decrease
the
times
listed
in
Table
3.2
and
should
be
discussed
with
the
UV
manufacturer.
Individual
manufacturers
report
that
colder
water
temperatures
(
below
50
degrees
Fahrenheit,
10
degrees
Centigrade)
can
result
in
slower
startups
for
LPHO
lamps
than
listed
in
Table
3.2.
Conversely,
MP
manufacturers
report
shorter
re­
start
times
with
colder
temperatures
because
the
cold
water
accelerates
the
condensation
of
mercury
(
i.
e.,
cool
down),
which
is
necessary
for
re­
striking
the
arc.

To
minimize
the
potential
for
off­
specification
operation,
utilities
should
evaluate
the
reliability
and
quality
of
their
power
supply.
Local
power
suppliers
can
often
provide
power
quality
and
reliability
data
and
should
be
the
first
source
of
information
on
power
quality.
For
those
locations
where
power
quality
is
unknown,
a
power
quality
assessment
is
recommended.
An
assessment
may
be
as
simple
as
reviewing
operating
records
of
power
quality
incidents
(
if
available)
and
power
interruptions
or
Supervisory
Control
and
Data
Acquisition
(
SCADA)
information
for
the
existing
plant.
More
advanced
assessments
may
include
the
installation
of
power
quality
monitors
or
the
retention
of
an
outside
consultant
to
conduct
a
detailed
power
quality
assessment.
Generally,
personnel
with
a
working
knowledge
of
electrical
supply
and
installation
will
be
able
to
review
power
supply
data
and
determine
if
power
quality
problems
exist.
If
a
problem
is
identified,
however,
tracing
it
back
to
its
source
and
determining
an
appropriate
remedy
is
often
best
left
to
an
expert
that
specializes
in
this
area.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
19
June
2003
The
most
common
sources
of
power
quality
problems
are
as
follows:

 
Faulty
wiring
and
grounding
 
Off­
site
accidents
(
e.
g.,
transformer
damaged
by
a
car
accident)

 
Weather­
related
damage
 
Animal­
related
damage
 
Facility
and
equipment
modifications
 
Power
transfer
to
emergency
generator
or
alternate
feeders
In
specific
locations
that
are
subject
to
frequent
power
fluctuations
or
outages,
the
following
options
should
be
considered
to
minimize
off­
specification
operation
and
ensure
regulatory
compliance:

1.
Installation
of
a
backup
generator
2.
Connection
to
a
second,
independent
power
source
3.
Installation
of
power
conditioning
equipment
or
a
battery­
supported
uninterruptible
power
supply
(
UPS)

These
options
will
have
different
response
and
backup
periods
associated
with
them.
For
example,
a
backup
generator
cost­
effectively
provides
backup
power
if
an
extended
power
interruption
occurs;
however,
it
will
not
ensure
a
continuous
power
supply
to
avoid
UV
reactor
shutdown
due
to
voltage
sags.
Connection
to
a
second,
independent
power
source
may
have
the
same
issues
as
the
backup
generator,
depending
on
the
power
quality
associated
with
the
second
power
source.

Power
conditioning
equipment
will
provide
high
quality
power
even
if
voltage
sags
or
other
power
quality
problems
occur.
However,
power­
conditioning
equipment
does
not
provide
backup
power
for
extended
power
outages.
A
battery­
supported
UPS
provides
continuous,
high
quality
power
(
i.
e.,
prevent
voltage
sags)
and
a
specific
amount
of
backup
power
for
a
longer
outage.
UPS
systems
can
provide
as
much
battery
backup
as
specified;
however,
typically
UPS
systems
for
this
purpose
range
between
2
and
15
minutes
of
battery
backup.

The
most
suitable
option
will
depend
on
the
power
quality
of
the
utility,
requirements
limiting
off­
specification
operation,
and
preferences
of
the
utility
and
State.
For
example,
an
unfiltered
system
with
poor
power
quality
that
experiences
multiple
voltage
sags
everyday
and
periodic
interruptions
lasting
over
3
minutes
may
consider
installing
a
UPS
system
with
5
minutes
of
backup
batteries
to
ensure
the
95
percent
requirement
of
operating
within
validated
ranges
is
met
(
40
CFR
141.721(
c)(
2)).
However,
a
filtered
system
that
experiences
two
or
three
voltage
sags
a
month
and
no
long­
term
power
interruptions
may
not
need
to
provide
any
additional
power
or
power
conditioning
equipment.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
20
June
2003
Any
equipment
needed
to
address
power
quality
problems
affects
both
the
cost
and
the
feasibility
of
implementing
UV
disinfection.
For
example,
the
UV
reactor
cost
and
installation
footprint
has
been
estimated
to
increase
by
approximately
25
percent
if
a
UPS
system
with
5
minutes
of
backup
capacity
is
installed
(
Cotton
et
al.
2002).
Other
power
conditioning
options
without
backup
power
are
less
expensive
and
have
lower
footprint
needs.

It
is
important
that
a
utility
have
a
complete
WTP­
wide
assessment
of
its
power
quality
when
considering
UV
disinfection.
Any
actions
involving
the
electrical
system
may
also
affect
the
WTP
power
quality
and
equipment
performance.
For
example,
the
impact
of
the
WTP's
maintenance
program
for
backup
generators
(
e.
g.,
routine
startup
and
exercise)
should
be
considered
during
the
planning
and
design
of
the
UV
reactors
to
ensure
that
the
program
supports
compliance
goals
and
does
not
cause
excessive
UV
reactor
shutdown
times.
Other
items
that
may
affect
power
quality
include
future
integration
or
upgrade
of
equipment
(
particularly
equipment
with
a
large
power
demand
or
variable
frequency
operation),
testing
of
backup
power
supplies,
deterioration
of
existing
facility
wiring
(
resulting
in
poorly
grounded
circuits),
overload
of
electrical
circuits,
and
any
other
activity
that
may
affect
the
electrical
supply
or
distribution
within
the
facility.

3.1.4
Evaluating
Potential
UV
Reactors
It
is
important
to
evaluate
the
available
UV
reactors
in
the
planning
process
because
each
manufacturer's
UV
reactors
are
unique
and
proprietary.
Process
footprints
and
related
installation
needs
(
e.
g.,
UV
reactor
to
control
panel
distances)
are
different,
depending
on
the
UV
manufacturer.
This
section
provides
a
brief
overview
of
different
UV
reactors,
their
impact
on
space
requirements,
and
UV
reactor
validation
issues.
More
detailed
UV
reactor
information
is
presented
in
section
2.4.
In
addition,
UV
manufacturers
should
be
contacted
directly
to
gain
a
better
understanding
of
the
UV
reactors
available
and
what
UV
reactors
are
applicable
to
the
utility's
installation
locations
given
the
design
criteria
developed
in
section
3.1.3.

3.1.4.1
UV
Reactors
There
are
different
types
of
UV
reactors
for
disinfecting
drinking
water
with
unique
characteristics,
such
as
lamps,
lamp
configuration
in
the
reactor,
cleaning
systems,
ballasts,
and
control
systems
(
section
2.4.).
This
section
briefly
highlights
the
differences
in
UV
reactors
that
affect
design
of
the
UV
installation.

UV
reactors
can
generally
be
characterized
based
on
lamp
type
with
LPHO
and
MP
lamps
being
the
most
applicable
to
WTPs.
One
of
the
fundamental
differences
between
LPHO
and
MP
reactors
is
the
lamp
intensity
output,
which
influences
the
UV
reactor
configuration
and
size,
lamp
life,
number
of
lamps,
electrical
needs,
and
ballasts.
Each
has
its
inherent
advantages
and
disadvantages.
While
a
competitive
procurement
can
be
made
among
these
two
reactor
types
when
the
construction
contract
is
bid,
the
overall
layout
and
supporting
facilities
will
be
different
for
each.

The
UV
reactor
footprint
depends
on
the
UV
reactor
configuration
and
UV
lamp
type.
There
are
several
different
UV
reactor
configurations.
Typically,
LPHO
reactors
are
in­
line
(
i.
e.,

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
21
June
2003
configured
like
a
pipe).
However,
MP
reactors
can
be
in­
line,
S­
shaped,
or
U­
shaped,
depending
on
the
UV
manufacturer
and
the
site
constraints
of
the
specific
installation
location.
Typically,
LPHO
reactors
have
a
larger
footprint
than
MP
reactors
because
more
UV
lamps
are
needed
to
deliver
the
same
UV
dose.
MP
reactor
footprints
will
also
vary,
depending
on
lamp
orientation
(
e.
g.,
parallel
versus
perpendicular
to
flow).
When
evaluating
locations
for
installation,
the
largest
UV
reactor
footprint
of
those
being
considered
should
be
used
to
estimate
the
UV
installation
footprint.

Lamp
life
also
varies
between
LPHO
and
MP
reactors.
Most
manufacturers
provide
warrantees
of
8,000
to
12,000
hours
for
LPHO
lamps.
Guaranteed
life
for
MP
lamps
range
from
4,000
to
8,000
hours.
Although
the
lamp
life
for
LPHO
is
greater
than
that
for
MP
reactors,
due
to
the
need
for
a
greater
number
of
lamps,
the
actual
number
of
lamps
that
are
replaced
during
a
given
period
may
be
less
for
a
MP
reactor.
It
is
important
to
consider
the
labor
associated
with
lamp
replacement,
as
well
as
the
actual
unit
cost
of
the
replacement
lamps,
when
estimating
the
operating
and
maintenance
costs
of
the
two
technologies.
In
addition,
while
LPHO
reactors
typically
have
more
lamps,
the
actual
power
input
is
less
than
that
for
similarly
sized
MP
reactors
because
MP
lamps
are
less
efficient
in
converting
the
power
input
to
germicidal
wavelengths
for
disinfection.
This
may
result
in
a
higher
input
power
and
an
increase
in
the
overall
power
consumption
for
MP
reactors
compared
to
LPHO
reactors.

The
lamp
sleeve
cleaning
systems
can
also
be
different
between
LPHO
and
MP
reactors.
LPHO
reactors
may
have
off­
line
chemical
cleaning
(
OCC)
systems
instead
of
on­
line
mechanical
cleaning
(
OMC)
because
of
the
larger
number
of
lamps.
With
OCC
systems,
the
UV
reactors
must
be
taken
off
line
to
be
cleaned.
OMC
and
OCC
systems
are
described
in
section
2.4.5.
This
may
result
in
higher
maintenance
costs
for
LPHO
reactors,
depending
on
the
extent
to
which
cleaning
is
necessary.

Finally,
the
type
of
ballast
used
will
affect
the
UV
installation
layout.
Ballasts
regulate
the
power
supply
at
the
appropriate
level
needed
for
energizing
and
driving
the
UV
lamps.
UV
reactors
may
use
electronic
ballasts,
electromagnetic
ballasts,
or
transformers.
Transformers
are
typically
more
stable
than
electronic
or
electromagnetic
ballasts
and
allow
a
greater
separation
distance
between
the
UV
reactor
and
control
panel.
However,
most
transformers
allow
only
step
adjustment
of
lamp
intensity.
Compared
to
transformers,
ballasts
have
the
capability
to
provide
almost
continuous
intensity
adjustment
but
may
increase
lamp
aging
and
spectral
shift
and
have
lower
allowable
separation
distances
between
the
UV
reactor
and
control
panel.
It
is
important
to
discuss
the
implications
of
these
various
components
with
the
UV
manufacturers
to
determine
their
effect
on
the
UV
installation
layout
and
design.
Specific
items
that
should
be
discussed
include
ballast
cooling
needs,
allowable
separation
distances,
and
intensity
adjustment
capabilities.

The
differences
described
above
imply
that
UV
reactor
evaluation
should
not
be
based
solely
on
capital
costs.
Operation
and
maintenance
costs,
including
energy
usage
and
labor,
will
be
important
in
an
overall
life
cycle
cost
comparison.
This
is
discussed
in
greater
detail
in
section
3.1.7.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
22
June
2003
3.1.4.2
UV
Reactor
Control
Strategies
There
are
currently
three
different
control
strategies
for
UV
reactors,
which
affect
how
UV
reactors
are
validated
and
operated.
The
three
general
control
strategies
relate
to
three
methods
for
monitoring
dose­
delivery
and
are
summarized
in
Table
3.3.
The
first
strategy
utilizes
one
or
more
UV
intensity
sensors
located
at
a
distance
from
the
lamps
that
yields
an
intensity
signal
that
is
proportional
to
UV
dose
(
UV
intensity
setpoint
approach),
and
the
intensity
sensor
measurement
and
flowrate
are
used
to
monitor
dose
delivery.
The
second
and
third
methods
utilize
UV
intensity
sensors
that
are
positioned
close
to
the
lamps
(
so
that
there
is
minimal
absorbance
by
the
water)
and
separate
monitors
for
UVT.
The
second
approach
incorporates
a
validated
setpoint
value
for
UVT,
in
addition
to
setpoints
for
UV
intensity
and
flowrate,
to
ensure
a
given
dose
(
UV
intensity
and
UVT
setpoint
approach).
In
the
third
approach,
the
UV
dose
is
calculated
based
on
these
measurements
of
flowrate,
UV
intensity,
and
UVT
via
a
validated
computational
algorithm
developed
by
the
manufacturer
(
calculated
dose
approach).

Table
3.3
UV
Reactor
Control
Strategies
Control
Strategy
Dose
Delivery
Monitoring
and
Control
Basis
UV
Intensity
Setpoint
UV
intensity
sensor
measurement
UVT
and
UV
Intensity
Setpoint
UV
intensity
sensor
and
UVT
measurement
Calculated
Dose
The
calculated
UV
dose1
1
The
UV
reactor
calculates
a
UV
dose,
using
the
UV
intensity
sensor
measurement,
the
UVT
of
the
water,
and
the
flowrate.

In
the
planning
phase,
these
control
strategies
need
to
be
evaluated
by
the
designer
and
utility
to
determine
if
a
particular
control
strategy
is
preferable
based
on
the
ease
of
integration
into
their
existing
operation
and
control
system.
The
impacts
of
the
control
strategy
on
the
instrumentation
and
controls
are
discussed
in
section
3.3.2,
and
the
specific
validation
recommendations
for
each
control
strategy
are
presented
in
section
C.
4.9.

3.1.4.3
Equipment
Validation
Issues
The
LT2ESWTR
requires
that
UV
reactors
be
validated
(
40
CFR
141.729(
d)).
A
utility's
approach
to
UV
reactor
validation
will
affect
the
UV
installation
design.
The
issues
to
consider
are
the
hydraulic
parameters
for
validation
and
whether
equipment
will
be
validated
on­
site
or
off­
site.

This
section
describes
how
these
issues
affect
the
design
and
installation
footprint
estimation.
Chapter
4
provides
an
overview
of
validation,
and
Appendix
C
details
UV
reactor
validation
guidelines
in
detail.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
23
June
2003
Validation
Hydraulics
The
inlet
and
outlet
hydraulics
of
the
UV
reactor
can
significantly
affect
dose
delivery;
therefore,
the
following
validation
and
corresponding
installation
strategies
are
recommended
in
the
validation
protocol
(
section
C.
3.1.5)
and
are
presented
in
Table
3.4.

Table
3.4
Summary
of
Recommended
Hydraulic
Configurations
for
Validation
and
Installation
Option
Validation
UV
Installation
1
The
inlet
and
outlet
configuration
is
the
same
as
the
installation
for
10
diameters
upstream
and
5
diameters
downstream
of
the
UV
reactor.
Inlet
and
outlet
configuration
is
the
same
as
when
the
UV
reactor
was
validated
for
10
diameters
upstream
and
5
diameters
downstream
of
the
UV
reactor.

2
The
UV
reactor
is
validated
with
a
90­
degree
bend
directly
upstream
of
the
UV
reactor.
The
UV
reactor
is
defined
to
include
a
specific
amount
of
straight
pipe
upstream
or
downstream
of
the
UV
reactor
as
specified
by
the
UV
manufacturer.
The
UV
reactor
should
be
installed
with
a
minimum
of
5
pipe
diameters
of
straight
piping
between
the
UV
reactor
and
any
upstream
hydraulic
configuration.
1
3
The
velocity
at
the
validation
facility
is
measured
at
evenly
spaced
points
through
a
given
cross
section
of
the
flow
upstream
and
downstream
of
the
UV
reactor.
The
velocity
at
the
installation
is
measured
at
evenly
spaced
points
through
a
given
cross
section
of
the
flow
upstream
and
downstream
and
is
within
20
percent
of
the
theoretical
velocity
determined
during
validation.
1
This
approach
is
not
acceptable
if
the
upstream
fitting
is
an
expansion
or
if
the
upstream
valve
will
be
used
for
flow
control.
A
valve
that
will
be
exclusively
used
for
open/
close
service
(
e.
g.,
isolation)
is
acceptable.

Option
1
is
most
applicable
when
unique
piping
configurations
are
needed
or
if
the
inlet
and
outlet
conditions
validated
in
Option
1
cannot
be
achieved
because
of
site
constraints.
For
example,
Option
1
may
be
the
only
validation
option
for
an
individual
filter
effluent
location,
which
probably
does
not
have
5
diameters
of
straight
pipe
before
the
UV
reactors
(
Option
2)
because
of
existing
site
constraints.

The
validation
and
installation
of
a
particular
UV
reactor
should
meet
one
of
these
options.
Option
2
provides
more
general
applicability
for
validation
and
installation
of
UV
reactors.
For
example,
the
inlet
and
outlet
piping
configuration
for
installations
in
a
new
building
could
be
designed
based
on
how
the
procured
UV
reactor
was
validated.
Option
3
also
provides
flexibility
but
may
have
the
practical
limitation
of
measuring
the
velocity
through
a
cross
section
at
the
installation.

Off­
site
Versus
On­
site
Validation
Manufacturers
will
likely
validate
UV
reactors
over
a
wide
range
of
flowrates
and
water
quality
(
e.
g.,
UVT)
conditions
at
off­
site
testing
facilities.
The
inlet
and
outlet
hydraulic
conditions
during
validation
will
probably
be
selected
so
the
UV
reactors
can
be
installed
in
most
WTPs.
Off­
site
validation
has
several
advantages,
including
simplicity,
cost,
and
the
ability
to
design
around
a
UV
reactor
with
known
performance
characteristics
and
inlet
and
outlet
hydraulics.
However,
the
LT2ESWTR
requires
that
the
site­
specific
installation
and
operating
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
24
June
2003
conditions
must
fall
within
the
range
of
conditions
used
when
the
installed
UV
reactor
was
validated
off­
site
(
40
CFR
141.729(
d)).
If
the
validation
conditions
do
not
encompass
the
utility's
design
criteria
or
inlet
and
outlet
piping
configurations,
the
utility
may
request
that
the
UV
manufacturer
re­
validate
the
unit
off­
site
under
specific
testing
conditions
that
closely
match
those
of
the
proposed
installation.
Alternatively,
on­
site
validation
can
be
performed.

The
advantage
of
on­
site
validation
are
that
the
UV
reactors
can
be
validated
under
the
exact
piping
hydraulic
conditions
at
which
it
will
operate,
and
the
UVT
will
more
accurately
represent
the
UV
installation
even
if
a
UV­
absorbing
chemical
is
added.
In
addition,
the
equipment
necessary
for
on­
site
validation
will
also
provide
the
flexibility
for
future
testing
to
optimize
the
UV
reactor
performance
under
specific
hydraulic
and
water
quality
conditions
even
if
they
are
not
completed
for
the
initial
validation.
However,
a
disadvantage
of
on­
site
validation
is
that
the
UV
installation
is
designed
and
constructed
without
prior
validation
of
the
performance
of
the
UV
reactors.
This
may
lead
to
the
UV
installation
failing
to
meet
performance
requirements,
and
it
may
be
difficult
to
increase
UV
disinfection
efficiency
after
the
UV
reactors
are
already
installed.
In
addition,
on­
site
validation
is
limited
to
the
highest
UVT
available
at
the
time
of
testing.
Consequently,
UV
reactor
performance
characteristics
cannot
be
determined
at
higher
UVT,
and
the
UV
reactors
may
need
to
be
operated
at
conditions
other
than
optimal,
resulting
in
higher
power
use
and
faster
lamp
and
ballast
replacement
frequencies.
Other
disadvantages
include
the
logistics
and
cost
of
the
testing.
For
example,
one
unit
must
be
isolated
from
the
system
to
allow
validation
testing
to
occur,
and
a
permit
may
be
needed
to
discharge
the
non­
pathogenic
challenge
microorganism.

If
on­
site
validation
is
desired,
then
the
UV
installation
design
should
be
adapted
to
enable
testing.
The
UV
reactor
design
would
need
to
incorporate
feed
and
sample
ports,
static
mixers,
space
for
tanks
near
the
UV
installation
(
for
the
addition
of
the
challenge
microorganism
and
UV
absorbing
chemical),
and
adequate
facilities
for
laboratory
testing,
and
discharge
of
the
treated
water.

3.1.5
Evaluating
Operational
Strategies
The
operational
strategy
is
defined
in
this
manual
as
the
method
in
which
the
utility
chooses
to
operate
the
UV
reactors
given
the
UV
reactor's
control
strategy
and
validation
data.
It
is
important
for
the
utility
to
understand
the
control
strategies
unique
to
various
UV
reactors
(
section
3.1.4.2)
and
select
equipment
consistent
with
their
operating
philosophy
and
energy
efficiency
objectives.
The
control
strategy
is
defined
as
the
method
that
the
UV
reactor
uses
to
monitor
and
control
the
UV
lamp
power
based
on
flow
and
UVT
to
deliver
the
specified
UV
dose.
For
each
UV
reactor,
the
operating
conditions
must
be
defined
based
on
validation
testing
results
(
40
CFR
141,
Subpart
W,
Appendix
D),
and
the
validation
data
will
vary
with
different
control
strategies.
The
validation
data
can
be
utilized
in
different
ways
that
facilitate
a
simple
or
complex
operating
strategy;
three
potential
approaches
are
described
in
Table
3.5.
Detailed
examples
of
how
to
determine
the
operational
parameters
for
these
operational
strategies
are
described
in
section
5.5.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
25
June
2003
Table
3.5
Potential
Operational
Strategies
Operational
Strategy
Description
Advantages
Disadvantages
Single
Operation
Setpoint
One
setpoint
is
used
for
all
flows
and
UVT
values
that
were
validated
Simplest
operational
strategy
Not
as
energy
efficient
because
the
UV
reactor
is
over­
dosing
at
low
flows
Variable
Setpoint
Operation
A
setpoint
would
be
used
for
a
given
flowrate
and
UVT
range
using
a
lookup
table
Increased
energy
efficiency
over
the
single
setpoint
approach
More
complex
operation
compared
to
single
setpoint
approach
and
may
necessitate
more
advanced
controls
for
the
UV
reactor
Setpoint
Interpolation
The
setpoints
are
calculated
as
a
function
of
flowrate,
typically
automatically
using
the
UV
reactor
controls1
The
most
energy
efficient
operation
and
may
reduce
operational
hours
needed
if
operated
automatically
Potentially
more
validation
data
is
needed
(
which
may
increase
validation
costs)
and
necessitates
advanced
reactor
controls
1
Only
an
option
for
UV
intensity
setpoint
and
calculated
dose
setpoint
approach
because
the
UV
intensity
and
UVT
setpoint
approach
is
controlled
as
function
of
flowrate
and
UVT
(
as
opposed
to
only
flowrate)

3.1.6
Evaluating
Hydraulics
and
Process
Footprint
The
potential
locations
for
UV
disinfection
identified
in
section
3.1.2
can
be
evaluated
based
on
an
understanding
of
the
candidate
UV
reactors,
the
hydraulics,
and
the
estimated
process
footprint.
This
section
discusses
the
principle
criteria
that
affect
the
feasibility
of
a
UV
installation
location
 
(
1)
hydraulic
needs
and
limitations
and
(
2)
space
availability
and
site
constraints.

3.1.6.1
Hydraulic
Considerations
When
selecting
the
appropriate
location
for
UV
reactors,
the
hydraulic
needs
should
be
addressed.
Headloss
through
a
UV
installation
is
dependent
on
the
specific
UV
reactor
and
flowrate
and
generally
varies
from
0.5
to
3
feet.
Characteristic
headloss
data
should
be
obtained
from
the
UV
manufacturer(
s)
for
all
candidate
UV
reactors.
In
addition
to
the
headloss
associated
with
the
UV
reactor
itself,
the
headloss
associated
with
piping,
valves,
flow
meters,
and
flow
distribution
devices
should
be
considered
when
assessing
the
feasibility
and
location
of
the
installation.
The
overall
headloss
of
a
UV
installation
is
typically
between
1
and
8
feet.

If
the
headloss
through
the
UV
installation
is
greater
than
the
available
head,
modifications
to
the
plant
design
and/
or
operation
may
be
necessary.
Some
potential
modifications,
alone
or
in
combination,
that
may
be
considered
to
address
hydraulic
limitations
are
listed
below
followed
by
details
about
each:

 
Eliminating
existing
hydraulic
inefficiencies
within
the
facility
to
improve
head
conditions
(
e.
g.,
replace
undersized
or
deteriorated
piping
and
valves)

 
Modifying
the
operation
of
the
clearwell
to
accommodate
the
UV
installation
 
Modifying
the
operation
of
the
filters
to
accommodate
the
UV
installation
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
26
June
2003
 
Installing
booster
pumps
 
Modifying
the
UV
reactor
design
(
through
the
UV
manufacturer)
to
reduce
headloss.
If
the
UV
reactor
design
is
modified,
it
must
be
validated
in
its
modified
condition
to
ensure
it
meets
performance
requirements
Eliminating
Existing
Hydraulic
Inefficiencies
Replacing
undersized
piping
and
valves
with
larger
diameter
piping
and
valves
may
increase
the
available
head
for
the
proposed
UV
installation.
Older
piping
can
also
produce
excessive
headloss
if
the
inner
pipe
surface
is
pitted
or
scaled
or
if
the
original
pipe
material
has
a
high
coefficient
of
friction.
Slip­
lining
the
interior
of
existing
pipe
with
a
lower
coefficient
of
friction
pipe
material
(
e.
g.,
high
density
polyethylene)
is
one
method
of
reducing
friction
losses.
Re­
lining
the
existing
pipe
interior
with
a
smooth
coating
will
also
reduce
headloss.

Modifying
Clearwell
Operation
A
utility
may
increase
head
available
to
a
UV
installation
by
lowering
the
surface
water
level
of
the
clearwell.
However,
this
strategy
decreases
the
storage
volume
available
to
meet
peak
demands.
In
addition,
a
lower
clearwell
level
will
reduce
the
contact
time
available
in
the
clearwell
for
chemical
disinfectants
and
may
affect
pump
discharge
head.
It
is
important
to
evaluate
any
potential
reduction
in
disinfection
credit
if
contact
time
in
the
clearwell
is
used
for
calculating
CT.
The
UV
installation,
though,
may
reduce
the
Giardia
CT
requirements
sufficiently
to
offset
the
reduction
in
contact
time
.

Modifying
Filter
Operation
A
treatment
facility
may
alter
the
operation
of
its
filters
to
increase
the
head
available
for
the
UV
installation.
However,
this
may
reduce
filter
run
times,
unit
filter
run
volumes,
and
result
in
more
frequent
backwashing.
If
conditions
upstream
of
the
filters
are
such
that
additional
freeboard
and
hydraulic
head
are
available,
a
second
option
is
to
increase
the
water
surface
elevation
over
the
filters
to
help
minimize
the
reduction
in
head
available
for
filtration.

Installing
Booster
Pumps
When
modifications
to
the
existing
facility
or
operations
will
not
provide
adequate
head
for
the
UV
reactors,
booster
pumps
can
be
installed.
Booster
pumping
provides
additional
flexibility
in
the
location
of
the
UV
reactors.
The
installation
of
booster
pumps
will
increase
facility
operation
and
maintenance
cost
and
space
requirements.
The
reliability
of
the
pumps
should
also
be
considered
in
the
evaluation
because
the
pumps
become
a
critical
operating
component.
Additional
detail
on
booster
pump
design
is
provided
in
section
3.3.1.6.

Modifying
UV
reactors
Modifying
a
UV
reactor
to
reduce
headloss
(
e.
g.,
removing
baffles)
can
affect
disinfection
performance
and
should
only
be
considered
in
careful
collaboration
with
the
UV
manufacturer.
Any
resulting
gains
in
system
head
must
be
weighed
against
diminished
disinfection
efficiency,
which
could
result
in
more
UV
reactors
being
needed
to
accommodate
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
27
June
2003
the
flow
and
provide
the
necessary
UV
dose.
Any
modified
UV
reactors
will
also
need
to
be
validated
in
its
modified
condition.

Other
Options
to
Address
Hydraulic
Constraints
If
none
of
the
above
options
are
feasible,
the
utility
could
consider
installing
the
UV
reactors
upstream
or
downstream
of
the
HSPs.
If
a
location
adjacent
to
the
HSPs
is
selected,
the
potential
for
damage
from
pressure
surges
is
increased
and
a
surge
analysis
should
be
completed.
Most
lamp
sleeves
are
designed
to
withstand
continuous
positive
pressures
of
at
least
120
pounds
per
square
inch
gauge
(
psig)
(
Roberts
2000;
Aquafine
2001;
Dinkloh
2001).
However,
lamp
sleeves
are
vulnerable
to
negative
gauge
pressure
transients
associated
with
water
hammer.
The
tolerance
level
of
the
sleeve
depends
on
the
quality
of
the
quartz
and
the
thickness
and
length
of
the
sleeve.
However,
pressures
of
negative
1.5
psig
have
been
shown
to
negatively
affect
sleeve
integrity
(
Dinkloh
2001).
Hydropneumatic
tanks,
surge
relief
valves,
air
release
valves,
or
air
vacuum
valves
on
pumps
or
at
different
locations
along
the
pipeline
can
be
used
to
help
control
surge
conditions.

3.1.6.2
Process
Footprint
The
process
footprint
should
be
estimated
in
the
planning
phase
to
determine
potential
UV
installation
locations.
One
critical
component
needed
to
estimate
the
UV
installation
footprint
is
the
number,
capacity,
and
configuration
of
the
UV
reactors.
The
number
of
UV
reactors
depends
on
the
redundancy
chosen.
UV
reactor
redundancy
should
be
determined
early
in
the
planning
process
and
should
use
sound
engineering
approaches
similar
to
those
used
for
other
major
equipment
(
e.
g.,
capacity
to
provide
full
treatment
with
the
largest
unit
out­
ofservice
The
specific
level
of
redundancy
should
be
determined
by
the
utility
based
on
operating
history
and
process
requirements
and
should
take
into
account
the
site
constraints.
For
example,
one
UV
reactor
dedicated
to
each
filter
may
have
different
redundancy
needs
than
UV
installations
treating
combined
filter
effluent.
Any
excess
capacity
that
may
be
available
within
the
UV
reactors
(
e.
g.,
incorporation
of
additional
lamps
or
change
in
lamp
power)
should
also
be
considered.

The
number
of
UV
reactors
necessary
is
also
affected
by
the
acceptable
power
turndown
of
the
UV
reactors
and
the
LT2ESWTR
requirement
that
the
UV
reactors
must
operate
within
their
validated
flow
range
(
40
CFR
141.729(
d)).
Some
UV
reactors
will
operate
at
low
power
efficiency
at
reduced
flowrates,
and
more
UV
reactors
with
a
lower
capacity
may
increase
energy
efficiency,
depending
on
water
quality
and
flowrates.
For
the
potential
combinations
of
number
and
capacity
of
UV
reactors,
the
available
turndown
should
be
determined
for
each
configuration
with
respect
to
the
anticipated
flow
range
and
power
modulation
capabilities
of
the
UV
reactors.

The
overall
UV
reactor
and
piping
configuration
will
be
affected
by
site
constraints.
For
example,
a
vertical
orientation
of
the
UV
reactors
may
be
necessary
to
reduce
building
footprint
because
of
little
land
availability.
Ultimately,
the
selected
configuration
should
balance
the
capital
cost
of
the
equipment,
which
may
be
lower
for
designs
incorporating
high
capacity
reactors,
with
the
improved
operating
efficiency
and
flexibility
that
may
be
achieved
using
a
larger
number
of
lower
capacity
reactors.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
28
June
2003
The
following
items
should
be
considered
when
estimating
the
UV
installation
footprint
in
the
planning
phase:

 
The
number,
capacity,
and
configuration
of
UV
reactors
(
including
redundancy
and
connection
piping)

 
The
configuration
of
the
connection
piping
and
the
inlet/
outlet
piping
necessary
before
and
after
each
UV
reactor,
based
on
validated
hydraulic
conditions
and
UV
manufacturer
recommendations
 
Booster
pumps
(
if
necessary)

 
The
space
needed
for
electrical
equipment
including
control
panels,
transformers,
ballasts,
backup
generator(
s),
and
possible
UPS
systems
 
The
maximum
allowable
separation
distance
between
the
UV
reactors
and
electrical
controls
since
distance
limitations
may
apply
 
Access
for
maintenance
and
replacement,
room
for
storage
of
spare
parts,
and
chemicals
(
if
needed),
and
lifting
capability
for
heavy
equipment
 
Provisions
for
on­
site
validation
(
if
applicable)

Once
the
UV
installation
footprint
is
estimated,
the
feasible
site
locations
may
be
determined
based
on
the
available
land
and
buildings
to
accommodate
the
installation
footprint.
UV
installation
layout
is
discussed
in
more
detail
in
section
3.3.5.

3.1.7
Preparing
Preliminary
Costs
and
Selecting
the
UV
Installation
Option
The
amount
of
analysis
necessary
to
determine
the
appropriate
application
point
for
a
UV
installation
is
site­
specific.
Some
options
will
clearly
be
infeasible
while
others
may
necessitate
a
more
detailed
comparison
of
the
installation
options.
Once
feasible
alternatives
are
identified,
the
development
of
life
cycle
costs
can
be
useful
in
selecting
among
alternatives.

Preliminary
life­
cycle
cost
estimates
should
include
both
capital
cost
and
operation
and
maintenance
cost
elements.
Capital
cost
elements
includes
the
cost
of
the
UV
reactors,
pumping
(
if
necessary),
electrical
and
instrumentation
provisions,
and
site
work;
contractor
overhead
and
profit;
piloting
and
validation
costs;
engineering,
legal,
and
administrative
costs.
Depending
on
the
detail
of
the
cost
estimates
being
developed,
the
existing
infrastructure
may
need
to
be
evaluated
to
develop
the
cost
estimate.
These
issues
are
discussed
in
detail
in
section
3.3.5.

The
average
conditions
for
flowrate
and
UVT
are
typically
the
most
representative
for
determining
annual
operating
costs,
as
opposed
to
the
maximum
design
flowrate
and
minimum
UVT.
Nevertheless,
the
specific
operating
limitations
of
the
equipment
and
the
electrical
cost
rate
structure
for
the
installation
should
be
considered.
If
a
utility's
electricity
charge
includes
both
a
usage
and
a
demand
component,
the
demand
charge
may
need
to
be
estimated
based
on
the
worst­
case
operating
conditions
to
accurately
represent
the
cost
to
the
utility.
Similarly,
it
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
29
June
2003
may
be
important
to
correlate
the
anticipated
energy
demand
for
the
UV
reactors
to
the
actual
rate
structure
for
the
facility
if
power
costs
vary
based
on
the
time
of
day
and
the
flowrate
and
UVT
fluctuate
significantly.

Selection
of
the
best
option
may
not
be
based
solely
on
capital
and
operation
and
maintenance
costs.
The
final
selection
criteria
should
also
consider
the
following
factors:

 
Cost­
effectiveness
and
ability
to
meet
the
utility
disinfection
and
design
objectives.

 
Ease
of
installation
(
where
applicable).

 
Operational
flexibility
and
reliability.

 
Specific
maintenance
needs.

 
Flexibility
for
future
treatment
expansion
(
if
applicable)

3.2
Equipment
Procurement
Options
The
same
equipment
procurement
options
that
are
used
to
acquire
traditional
equipment
(
e.
g.,
pumps)
within
the
water
industry
can
also
be
used
for
UV
reactors.
Owner
pre­
purchase;
base
bid,
under
which
the
design
is
based
on
a
single
UV
manufacturer
but
is
open
to
alternatives
at
the
discretion
of
the
owner;
and
contractor
selection,
in
which
operating
or
performance
criteria
are
established
and
final
equipment
selection
is
left
to
the
discretion
of
the
contractor
are
the
most
common
methods
of
procurement
for
traditional
design­
bid­
build
projects.
Because
the
use
of
UV
reactors
in
drinking
water
treatment
plant
applications
has
been
limited
in
the
United
States,
many
of
the
projects
to
date
have
pre­
purchased
the
UV
reactors.
Pre­
purchase
allows
the
designer
to
work
more
closely
with
the
UV
manufacturer
during
design,
reducing
the
potential
for
errors
that
could
occur
with
an
evolving
technology.
However,
pre­
purchasing
may
necessitate
that
a
more
detailed
assessment
be
completed
during
the
planning
stage
of
the
project
to
ensure
that
the
appropriate
equipment
is
selected
and
that
a
second
set
of
contract
documents
be
prepared.
Further,
this
may
result
in
the
owner
assuming
increased
responsibility
for
equipment
delivery
and
performance
when
compared
to
base
bid
or
contractor
selection.
If
owner
pre­
purchase
is
selected,
these
factors
need
to
be
carefully
considered
and
addressed
by
the
designer
during
development
of
the
equipment
procurement
document.

The
advantages
and
disadvantages
of
the
procurement
methods
with
respect
to
designing
and
constructing
a
UV
installation
are
described
in
Table
3.6.
It
should
be
noted
that
funding
sources
or
municipalities
might
have
specific
bidding
and
procurement
requirements.
These
requirements
are
site­
specific
and
should
be
reviewed
prior
to
establishing
a
project
approach
to
ensure
all
requirements
are
met.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
30
June
2003
Table
3.6
Potential
UV
Reactor
Procurement
Options
Procurement
Method
Advantages
Disadvantages
Owner
Pre­
Purchase
 
Single
design
around
selected
equipment.
 
Actual
UV
reactor
pricing
is
better
defined
earlier
in
project.
 
Owner
receives
equipment
warranty
directly
from
UV
manufacturer.
 
May
result
in
shorter
project
schedule
if
equipment
fabrication
time
occurs
during
design
and
bidding
phases
of
the
UV
installation.
 
Option
may
necessitate
the
preparation
of
two
sets
of
documents;
an
equipment
procurement
document
and
contract
documents
for
the
construction
of
the
overall
UV
installation.
 
Option
may
not
be
possible
under
some
procurement
codes.
 
Except
where
procurement
is
assigned,
installation
contractor
is
not
single
point
of
responsibility
for
equipment.
 
Equipment
disputes
need
to
be
dealt
with
by
owner.
Base
Bid
 
Single
design
around
selected
UV
reactors.
 
Contractor
handles
all
pricing
and
coordination
with
UV
manufacturer.
 
Low
incentive
for
contractor
to
bid
alternates
to
selected
UV
manufacturer.
 
It
is
difficult
to
prevent
supplier
"
packaging"
of
UV
reactors.
 
UV
reactor
disputes
are
problematic
because
contractor
was
directed
to
use
equipment.
Contractor
Selection
 
Fits
most
procurement
codes.
 
UV
reactor
disputes
are
the
responsibility
of
the
contractor.
 
Contractor
is
likely
to
select
UV
reactors
with
lowest
capital
cost
rather
than
lowest
life­
cycle
cost.
 
Multiple
UV
installation
designs
may
be
necessary,
increasing
engineering
effort
and
cost.

As
discussed
previously,
Chapter
3
is
organized
in
the
same
manner
as
the
flow
chart
shown
in
Figure
3.1,
utilizing
equipment
pre­
purchase
and
a
design­
bid­
build
approach
for
project
implementation.
It
should
be
noted
that
successful
implementation
of
UV
installations
can
be
accomplished
using
any
of
the
equipment
procurement
and
contractor
selection
approaches
currently
available.

3.3
UV
Installation
Design
Elements
Additional
design
concepts
are
expanded
and
refined
in
this
section,
particularly
with
regard
to
hydraulic
issues,
control
strategy,
instrumentation,
and
electrical
power.
The
section
concludes
with
considerations
for
the
layout
of
UV
installations.

3.3.1
UV
Installation
Hydraulics
Following
the
selection
of
an
installation
option
during
the
planning
phase,
a
more
detailed
evaluation
of
system
hydraulics
should
be
conducted,
including
flow
control,
distribution,
and
measurement.
It
is
important
that
design
of
the
inlet
and
outlet
conditions
be
coordinated
with
the
validation
process
to
ensure
that
the
proposed
configuration
can
be
cost­

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
31
June
2003
effectively
validated
and
will
provide
hydraulic
conditions
that
result
in
dose
delivery
equal
to
or
better
than
that
provided
during
validation
testing.

3.3.1.1
Inlet
and
Outlet
Piping
Configuration
Optimal
hydraulic
conditions
vary
based
on
the
UV
reactor
design
and
lamp
configuration,
but
turbulent
flow
with
a
reasonably
uniform
velocity
profile
is
generally
preferred.
Turbulent
flow
conditions
can
be
achieved
at
very
low
flowrates
when
compared
to
the
actual
capacity
of
a
given
pipe
cross
section.

The
recommended
inlet
and
outlet
conditions
for
validation
and
the
installation
are
summarized
in
section
3.1.4.3
and
described
in
detail
in
section
C.
3.1.5.
These
recommendations
should
be
considered
when
designing
the
inlet
and
outlet
conditions
for
the
UV
reactors.
The
designer
should
contact
the
UV
manufacturer
to
determine
how
the
procured
UV
reactors
were
validated
and
what
the
inlet
and
outlet
piping
constraints
are
for
the
UV
installation.
If
on­
site
validation
is
planned,
the
inlet
and
outlet
hydraulics
should
be
designed
as
recommended
by
the
UV
manufacturer
and
as
the
site­
specific
constraints
permit.

3.3.1.2
Flow
Distribution,
Control,
and
Measurement
Regulations
specify
flowrate,
UV
intensity,
and
lamp
status
as
the
minimum
operating
conditions
a
utility
must
routinely
monitor
(
40
CFR
141,
Subpart
W,
Appendix
D).
Accordingly,
proper
flow
distribution
and
measurement
are
essential
for
compliance
monitoring
of
the
UV
reactors.
Confirmation
of
compliance
will
be
dependent
on
understanding
the
flow
through
each
UV
reactor,
regardless
of
the
dose
monitoring
or
control
strategy
used
by
the
utility.
Moreover,
UV
reactors
are
validated
within
specific
flow
ranges
and
have
associated
operating
characteristics
that
demonstrate
dose
delivery
as
a
function
of
flow.
Therefore,
the
flowrate
through
the
UV
reactor
must
be
known
to
ensure
that
proper
dose
delivery
is
achieved.

This
section
discusses
different
methods
for
ensuring
proper
flow
distribution
and
measurement
through
UV
reactors.
In
some
instances,
flow
can
be
determined
through
flow
splitting
and
proper
hydraulic
design
without
an
individual
flow
measurement
device
for
each
UV
reactor.
Nevertheless,
the
need
for
individual
flow
measurement
for
each
UV
reactor
is
at
the
discretion
of
the
State.
Utilities
implementing
UV
disinfection
are
encouraged
to
discuss
flow
measurement
requirements
with
their
State
during
the
planning
and
preliminary
design
phases.

Flow
Distribution
and
Control
Two
approaches
for
flow
measurement
and
control
have
generally
been
used.
The
first
involves
the
installation
of
a
dedicated
flow
meter
and
flow
control
valve
for
each
UV
reactor.
The
second
involves
the
use
of
passive
flow
distribution,
with
confirmation
of
equal
flow
split
by
monitoring
pressure
differential
across
identical
pipe
segments
(
or
the
UV
reactor)
or
with
flow
meters.
For
identical
reactors,
the
differential
pressure
across
each
parallel
UV
reactor
train
should
be
the
same
if
equal
flow
distribution
is
occurring
and
valves
are
in
the
same
operating
position.
The
use
of
dedicated
flow
meters
and
modulating
downstream
valves
to
control
flow
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
32
June
2003
through
the
UV
reactors
provides
the
greatest
hydraulic
control
in
applications
with
widely
varying
flowrates.

Assuming
multiple,
parallel
UV
reactors
of
the
same
capacity,
the
UV
reactors
should
be
sized
and
configured
to
provide
approximately
equal
headloss
through
each
treatment
train
(
i.
e.,
portion
of
distribution
and
recombination
channel
or
manifold,
lateral
piping,
and
UV
reactor
with
associated
valves
and
flow
measurement).
This
is
particularly
important
if
passive
flow
distribution
is
used.
Because
flowrates
may
deviate
from
equal
distribution,
the
maximum
design
flowrate
for
each
reactor
should
account
for
any
potential
distribution
imbalance.
Equation
3.1
can
be
used
to
determine
the
appropriate
upper
design
flowrate
for
each
UV
reactor:

N
E
Q
Q
total
reactor
)
1
(
+
 
=
Equation
3.1
where
Qreactor
=
UV
reactor
design
flow
Qtotal
=
Plant
maximum
design
flow
E
=
Calculated
maximum
flow
distribution
error
(
percent
as
a
decimal)
N
=
Number
of
on­
line
UV
reactors
The
maximum
flow
distribution
error
(
E)
should
be
determined
through
hydraulic
calculations
or
hydraulic
modeling
of
the
UV
installation.
For
example,
ideally
two
identical,
parallel
reactors
would
have
a
50/
50
flow
split.
If
the
actual
flow
split
between
the
reactors
is
calculated
or
modeled
to
be
60/
40
percent,
then
a
20
percent
(
E=
0.20)
maximum
flow
distribution
error
(
E=(
60­
50)/
50=
0.2)
would
be
used
in
the
above
equation
to
estimate
the
proper
design
flow
for
the
reactor.

The
reactor
flow
should
be
estimated
over
the
range
of
anticipated
operating
reactors
(
i.
e.,
number
of
operating
reactors).
In
general,
with
passive
distribution,
as
the
number
of
UV
reactors
increases
and
flowrate
decreases,
the
potential
for
flow
distribution
imbalance
is
magnified.
Effective
passive
distribution
relies
on
the
headloss
through
each
treatment
lateral
being
significantly
greater
than
the
headloss
through
the
common
influent
manifold
or
chamber.
Under
the
conditions
of
reduced
flow
and
an
increased
number
of
operating
reactors,
the
relative
amount
of
headloss
through
each
lateral
becomes
less
significant
when
compared
to
the
headloss
through
the
manifold,
resulting
in
less
controlled
distribution.

For
utilities
that
use
distribution
and
recombination
channels
(
as
opposed
to
influent
and
effluent
manifolds),
designers
typically
have
two
basic
choices
to
achieve
passive
flow
distribution
(
Figure
3.8):
a
series
of
individual
weirs
set
at
the
same
elevation
or
a
series
of
orifices
submerged
into
the
individual
UV
reactor
laterals.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
33
June
2003
Figure
3.8
Open­
Channel
Flow
Distribution
Options
A.
Flow
Splitting
Weirs
Plan
Weir
Plan
Section
Section
B.
Submerged
Orifices
Flow
Flow
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
34
June
2003
Flow
Measurement
Depending
on
the
design
and
control
strategy
of
the
UV
reactor,
a
number
of
options
are
available
for
flow
measurement.
As
discussed
previously,
flow
measurement
devices
installed
specifically
for
the
UV
reactors
may
not
be
needed
in
all
applications.
It
should
be
noted
that
some
level
of
inaccuracy
or
drift
is
likely
to
occur
with
all
flow
meters.
This
potential
error
should
be
accounted
for
during
design
and
validation.

If
a
single
UV
reactor
is
installed,
the
plant's
raw
water
metering
station
can
be
used
to
determine
a
reasonably
accurate
flow
through
the
reactor.
The
use
of
raw
water
flow
metering
data,
however,
may
not
account
for
backwash
and
residuals
flow
losses,
which
would
create
flow
measurement
inaccuracies
for
UV
reactors
installed
downstream
of
the
filters
or
clearwell.
For
applications
where
the
UV
reactor
is
dedicated
to
a
rate­
of­
flow
control
filter,
flow
information
from
the
filters
may
be
used
to
determine
the
flowrate
through
the
UV
reactors.

If
equal
flow
distribution
between
multiple
UV
reactors
can
be
achieved
passively
under
all
hydraulic
conditions,
a
single,
common
flow
meter
(
new
or
existing)
may
be
used
to
measure
flow.
The
total
flow
can
then
be
divided
by
the
number
of
operating
UV
reactors
to
determine
the
flow
through
each
UV
reactor.
If
this
approach
is
selected,
some
method
of
confirming
the
equal
flow
split
should
also
be
incorporated
(
e.
g.,
differential
pressure
measurement).

A
single
flow
meter
for
the
entire
UV
installation
or
individual
meters
(
with
or
without
rate­
of­
flow
control)
should
be
considered
to
provide
increased
flow
measurement
accuracy.
Magnetic
flow
meters
or
other
meter
types,
such
as
doppler,
that
do
not
protrude
into
the
flow
path
have
the
least
effect
on
the
velocity
profile,
which
minimizes
the
potential
effect
on
reactor
inlet
or
outlet
hydraulics.
The
desired
means
of
flow
measurement
for
the
UV
reactors
should
be
selected
based
on
the
level
of
flow
measurement
accuracy
needed
to
accomplish
the
operating
and
control
strategy
for
the
installation
and
satisfy
validation
criteria,
as
well
as
an
understanding
of
the
variability
in
the
plant
flowrate.
Several
options
are
listed
in
Table
3.7
and
are
illustrated
in
Figure
3.9.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
Table
3.7
Comparison
of
Techniques
for
UV
Installation
Flow
Measurement
Flow
Measurement
Method
Description
Flow
Control
Method
Advantages
Disadvantages
A.
Raw
or
Filtered
Water
Flow
Measurement
(
Figure
3.8A)
Use
plant
flow
information
upstream
of
UV
reactors.
Passive
flow
control
such
as
a
weir
or
orifice.
 
New
flow
meters
not
needed
 
Minimizes
UV
installation
capital
cost
 
Simplifies
control
strategy
 
Introduces
potential
errors
in
measured
flow
versus
actual
UV
reactor
flow
 
Relies
on
adequate
flow
distribution
between
UV
reactors
 
Relies
on
a
single
meter
 
May
need
oversized
UV
reactors
to
provide
adequate
dose
delivery
at
all
times
B.
Single
Flow
Meter
for
Flow
Measurement
to
Entire
UV
Installation
Measure
total
UV
reactor
flow.

(
Figure
3.8B)
Passive
flow
control
such
as
a
weir
or
orifice.
 
Measures
flows
accurately
 
Only
one
new
flow
meter
needed
 
Relies
on
adequate
flow
distribution
between
UV
reactors
 
Relies
on
a
single
meter
 
May
need
oversized
UV
reactors
to
provide
adequate
dose
delivery
at
all
times
C.
Individual
UV
Reactor
Flow
Measurement
(
Figure
3.8C
without
flow
control
valve)
Measure
flow
for
each
UV
reactor.
Passive
flow
control
such
as
a
weir
or
orifice.
 
Measures
UV
reactor
flows
accurately
 
Does
not
have
one
meter
as
a
single
point
of
failure
 
Equal
flow
distribution
is
not
necessary
for
dose
control
 
Relies
on
adequate
flow
distribution
 
Increases
capital
cost
 
Increases
UV
installation
complexity
 
Increases
installation
footprint
to
achieve
necessary
meter
hydraulics
 
Increases
reactor
headloss
D.
Individual
UV
Reactor
Flow
Measurement
and
Control
(
Figure
3.8C)
Measure
and
control
flow
for
each
UV
reactor.
Individual
flow
control
(
valve)
for
each
UV
reactor.
 
Does
not
introduce
potential
errors
in
measured
flow
 
Does
not
rely
on
adequate
flow
distribution
 
Does
not
rely
on
a
single
meter
 
Increases
capital
cost
 
Increases
UV
installation
complexity
 
Increases
installation
footprint
due
to
hydraulics
of
UV
reactor,
meter,
and
valves
 
Increases
reactor
headloss
UV
Disinfection
Guidance
Manual
3­
35
June
2003
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
36
June
2003
Figure
3.9
Flow
Measurement
and
Control
Options
A.
Raw
Water
Meter
for
Flow
Measurement
M
C.
Individual
UV
Reactor
Flow
Measurement
and
Control
M
M
B.
Single
UV
Disinfection
Facility
Meter
for
Flow
Measurement
M
M
M
M
Raw
Water
Flowmeter
Raw
Water
Flowmeter
UV
Facility
Flowmeter
Raw
Water
Flowmeter
Individual
Flowmeters
3.3.1.3
Level
Control
The
UV
reactors
must
be
flowing
full
at
all
times
during
operation.
Therefore,
the
reactors
should
be
placed
below
the
hydraulic
grade
line
elevation.
There
are
two
basic
options
commonly
used
to
maintain
the
level
in
the
reactors.
One
option
is
with
a
fixed
downstream
weir;
in
many
WTPs,
a
fixed
weir
is
already
located
in
a
clearwell
and
can
be
used
for
this
purpose.
If
not,
another
option
is
to
install
a
weir
immediately
downstream
of
the
UV
reactor
or
at
another
location
that
ensures
full
pipe
flow
through
the
UV
reactors.
A
final
option
is
to
use
flow
control
valves
to
monitor
and
maintain
the
downstream
hydraulic
grade
line.

3.3.1.4
Air
Relief
and
Pressure
Control
Valves
UV
reactors
should
be
kept
free
of
air
to
prevent
lamp
overheating.
The
formation
of
negative
pressures
or
surge
effects
within
the
UV
reactors
should
also
be
prevented
to
avoid
damage
to
the
lamp
sleeve
and
UV
lamps.
The
use
of
air
release
valves,
air/
vacuum
valves,
or
combination
air
valves
may
be
appropriate
to
prevent
air
pockets
and
negative
pressure
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
37
June
2003
conditions.
The
locations
of
the
valves
will
be
dictated
by
the
specific
configuration
of
the
installation
and
should
be
determined
during
design.

3.3.1.5
Flow
Control
and
Isolation
Valves
Each
UV
reactor
should
have
the
capability
of
being
isolated
and
taken
out
of
service.
This
will
necessitate
a
valve,
gate,
or
other
isolation
device
upstream
and
downstream
of
the
UV
reactor.
Valves
are
generally
preferred,
since
they
provide
a
tighter
seal.
Utilities
that
use
passive
flow
distribution
will
rely
on
the
valves
primarily
for
isolation
and
sequencing
of
UV
reactor
operation
(
as
opposed
to
flow
control).
Valves
downstream
of
the
UV
reactor
should
be
equipped
with
an
actuator
to
automatically
open
or
close
on
a
critical
alarm
occurrence
and
to
enable
start­
up
sequencing.

If
the
isolation
valves
are
used
for
flow
control,
either
the
upstream
or
the
downstream
valve
can
be
used.
However,
it
is
generally
recommended
that
the
valve
downstream
of
the
UV
reactor
be
used
to
minimize
disturbance
of
the
flow
entering
the
UV
reactor,
particularly
if
the
separation
distance
between
the
upstream
valve
and
the
UV
reactor
is
relatively
small.
The
flow
characteristic
curve
of
the
valve
and
the
operating
speed
of
the
actuator
should
be
matched
to
the
flow
control
needs
of
the
UV
reactors.
During
design,
the
valve
configuration
should
be
discussed
with
the
UV
manufacturer
to
ensure
that
UV
reactor
performance
will
not
be
adversely
affected
by
the
location
or
operation
of
the
valves.
It
is
important
to
coordinate
the
location
of
the
valves
with
the
validation
conditions
for
the
reactor,
as
discussed
in
section
C.
3.1.1.

Valve
seats
and
other
in­
pipe
seals
and
fittings
within
the
straight
pipe
lengths
adjacent
to
the
UV
reactors
should
be
constructed
of
materials
that
are
resistant
to
UV
light
to
avoid
degradation.
If
in­
place
rehabilitation
of
existing
piping
is
used
to
improve
system
hydraulics,
the
materials
used
to
slip­
line
or
reline
the
piping
adjacent
to
the
proposed
UV
reactors
should
also
be
resistant
to
degradation
from
exposure
to
UV
light.
Organic
materials
and
plastics
that
have
not
incorporated
UV­
resistant
additives
are
typically
most
susceptible
to
UV
degradation.

3.3.1.6
Intermediate
Booster
Pumps
A
detailed
evaluation
and
design
of
a
booster
pumping
system
is
recommended
if
it
is
determined
during
the
planning
phase
that
head
constraints
necessitate
the
installation
of
a
pumping
system.
Pumps
common
in
water
treatment
plants
(
i.
e.,
vertical
turbine,
end­
suction
centrifugal,
and
split­
case
centrifugal
pumps)
tend
to
have
higher
discharge
pressures
than
intermediate
pumping
applications
and
are
generally
not
appropriate
for
this
application.
Mixed
flow
or
axial
flow
pumps
with
high­
flow
and
low­
head
operating
characteristics
are
typically
more
appropriate.
However,
additional
headloss
may
need
to
be
added
to
the
system,
based
on
the
capabilities
of
the
pump.
Smaller
diameter
piping,
backpressure
valves,
or
control
valves
can
be
used
to
increase
the
system
head
to
more
closely
match
the
pump
discharge
curve.

Pumps
may
be
installed
before
or
after
the
UV
reactors,
allowing
more
flexibility
in
the
UV
installation's
design
elevations
and
the
location
of
the
UV
reactors.
Regardless
of
pump
location,
some
form
of
wetwell
should
be
provided
upstream
of
the
pump
station.
Existing
clearwells,
recombination
channels,
or
dedicated
pump
wetwells
may
be
used.
Direct
connection
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
38
June
2003
Proposal
Draft
clearwells,
recombination
channels,
or
dedicated
pump
wetwells
may
be
used.
Direct
connection
to
filter
effluent
piping
may
adversely
affect
upstream
process
performance
and
should
be
avoided.
Booster
pump
operation
may
be
controlled
by
the
water
level
within
the
upstream
wetwell.
The
use
of
variable
frequency
drives
(
VFDs)
to
moderate
flow
peaks
is
recommended.
This
is
especially
important
if
the
pump
station
is
upstream
of
the
UV
reactors.
By
minimizing
hydraulic
peaks,
the
UV
reactors
can
be
sized
to
more
closely
match
the
flow
through
the
WTP.

If
pumps
are
located
adjacent
to
the
UV
reactors,
the
impact
of
surge
conditions
should
be
evaluated.
Of
particular
concern
is
the
potential
for
surge
if
the
pumps
are
operating
and
power
is
lost.
Pump
start­
up
procedures
should
be
carefully
selected
with
possible
inclusion
of
pump
control
valves.
Control
of
individual
UV
reactor
isolation
valves
should
be
coordinated
with
pump
starts
and
stops
and
with
pump
control
valves
where
appropriate.
Likewise,
the
warm­
up
time
associated
with
the
start­
up
of
the
UV
reactors
must
be
taken
into
account
with
the
sequencing
of
the
pump
operation.

3.3.2
Operational
Strategy
Determination
Once
the
UV
reactors
are
procured,
a
utility
should
determine
the
preferred
operational
strategy
given
the
UV
reactor's
control
strategy
and
available
validation
data.
The
different
operational
strategies
are
described
in
section
3.1.5,
and
an
example
of
how
to
interpret
the
validation
data
to
develop
an
operational
strategy
is
described
in
section
5.5.

The
power
needs
for
UV
reactors
can
be
moderately
high,
and
an
inefficient
UV
installation
can
result
in
unnecessarily
high
operating
costs.
When
considering
what
operational
strategy
to
use
for
a
particular
installation,
the
operational
complexity
should
be
compared
to
the
potential
for
energy
savings.
It
should
be
noted,
however,
that
intensity
adjustment
does
not
correlate
directly
to
the
amount
of
energy
that
is
saved.
Lamp
output
efficiency
may
decrease
as
the
lamp
intensity
is
reduced,
resulting
in
a
reduced
energy
savings.
Lamp
output
efficiency
as
a
factor
of
intensity
should
be
discussed
with
the
UV
manufacturer
and
considered
when
determining
the
potential
cost
savings
associated
with
dose
pacing.
An
operational
strategy
consistent
with
the
procured
UV
reactor
should
be
selected
to
facilitate
the
instrumentation
and
control
design.

3.3.3
Instrumentation
and
Control
After
the
hydraulic
needs
of
the
UV
reactors
have
been
addressed
and
a
dose
control
strategy
has
been
selected,
the
instrumentation
and
controls
necessary
to
satisfy
both
can
be
identified.
The
level
of
instrumentation
and
control
that
is
needed
will
depend
on
the
flow
control,
flow
distribution,
and
flow
measurement
approach
that
is
selected,
as
well
as
the
dose
control
strategy
that
is
employed.
Passive
flow
distribution
with
an
intensity
setpoint
dose
control
strategy
is
a
relatively
simple
operation
and
demands
limited
instrumentation
and
control.
Operating
flexibility
and
the
ability
to
optimize
UV
reactor
energy
efficiency,
however,
are
reduced.
The
use
of
dedicated
flow
meters
and
flow
control
valves,
in
combination
with
on­
line
transmittance
monitors
and
dose
pacing,
demands
a
higher
level
of
instrumentation
and
control.
However,
this
approach
provides
significant
operating
flexibility
and
the
ability
to
optimize
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
39
June
2003
The
specific
instrumentation
and
control
elements
included
with
the
UV
reactors
may
not
be
known
until
a
final
UV
reactor
selection
is
made.
Most
of
the
equipment
manufacturers,
however,
share
common
instrumentation
and
control
attributes
and
alarm
conditions
in
the
designs
of
their
UV
reactors.
To
enable
a
procurement
document
to
be
prepared,
a
control
strategy
should
be
established.
To
the
extent
practicable,
the
designer
should
identify
the
elements
of
the
control
system
that
are
preprogrammed
into
the
UV
reactor
control
panel
and
those
that
will
be
addressed
through
the
installation
of
supplemental
controls
and
equipment.
At
a
minimum,
the
LT2ESWTR
requires
that
UV
lamp
intensity,
flowrate,
and
lamp
status
be
monitored
(
40
CFR
141.729(
d)).
The
final
instrumentation
and
control
design
can
be
modified
as
needed
after
equipment
is
selected.

3.3.3.1
UV
Reactor
Start­
up
Regardless
of
the
UV
reactors
that
are
selected,
the
start­
up
cycle
will
likely
be
the
same.
For
a
UV
reactor
that
is
starting
cold
(
i.
e.,
previously
off
as
opposed
to
shutdown
for
a
very
short
period
due
to
power
interruption),
a
typical
control
sequence
will
open
the
isolation
valves,
ignite
the
lamps,
and
bring
the
lamps
to
full
power.
During
the
typical
control
sequence,
the
water
being
treated
will
be
off­
specification
until
the
lamps
reach
full
operating
power,
which
can
take
up
to
10
minutes.
However,
the
amount
of
off­
specification
water
can
be
reduced
by
providing
a
low
flow
of
cooling
water
that
can
be
discharged
to
waste.
Alternatively,
if
a
LP
or
LPHO
reactor
is
procured,
the
downstream
valve
may
remain
closed
as
the
UV
lamps
are
warming
up.
However,
the
designer
should
consult
the
LP
or
LPHO
manufacturer
to
ensure
this
strategy
is
feasible.
It
is
recommended
that
the
utility
discuss
these
practices
with
the
State
to
confirm
their
acceptance.

3.3.3.2
UV
Reactor
Automation
Depending
on
the
size
and
complexity
of
the
UV
reactor,
its
operation
can
range
from
manual
to
fully
automatic.
Manual
operation
includes
manual
initiation
of
lamp
start­
up
and
shut
down,
and
appropriate
valve
actuation.
Different
levels
and
types
of
automation
can
be
added
to
the
manual
sequence.
A
first
level
of
automation
includes
the
sequencing
of
lamp
startup
and
valve
actuation
to
bring
individual
UV
reactors
on­
line
after
manual
initiation.
Further
levels
of
automation
could
include
starting
up
UV
reactors,
activating
rows
of
lamps,
or
making
lamp
intensity
adjustments
based
on
lamp
condition,
water
quality,
and/
or
flowrate.

Automatic
UV
reactor
shutdown
under
critical
alarm
conditions
(
e.
g.,
high
temperature,
lamp
or
sleeve
failure,
loss
of
flow)
is
important
for
all
operating
approaches,
including
manual
operation.
The
shutdown
cycle
will
be
site­
specific.
However,
to
the
extent
practicable,
the
downstream
flow
control
or
isolation
valve
should
be
closed
whenever
the
UV
reactor
is
shut
down
to
minimize
the
distribution
of
water
that
has
not
been
disinfected
by
the
UV
installation.

3.3.3.3
UV
Intensity
and
Calculated
Dose
(
If
Applicable)

Signals
from
UV
intensity
sensors
should
be
displayed
locally
or
on
the
UV
reactor
control
panel.
Because
the
output
from
the
UV
intensity
sensor
is
integral
to
the
determination
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
40
June
2003
of
adequate
dose
delivery,
the
UV
intensity
sensor
output
should
be
monitored
continuously.
If
the
calculated
dose
control
strategy
is
used,
the
calculated
dose
should
be
displayed
locally
and
be
monitored
continuously.

3.3.3.4
UV
Transmittance
An
on­
line
UVT
monitor
or
bench­
top
spectrophotometer
may
be
used
to
monitor
UVT,
depending
on
the
control
strategy
(
section
3.1.4.2).
An
on­
line
UVT
monitor
is
typically
used
for
the
UV
intensity
and
UVT
setpoint
approach
and
the
calculated
UV
dose
setpoint
approach.
However,
for
utilities
that
have
water
with
a
stable
UVT,
periodic
grab
samples
may
be
adequate.
Results
from
a
bench­
top
spectrophotometer
can
be
manually
input
into
a
SCADA
system
or
other
control
system.
Output
from
an
on­
line
UVT
monitor
can
be
input
directly
into
a
control
loop
for
most
UV
reactors,
a
SCADA
system,
or
both.

If
the
UV
intensity
setpoint
approach
is
used,
UVT
does
not
need
to
be
monitored
because
the
UVT
is
accounted
for
in
the
UV
intensity
measurement.
However,
it
may
be
advantageous
to
monitor
UVT
with
an
on­
line
UVT
monitor
or
bench­
top
unit
to
assist
with
troubleshooting
UV
reactor
performance
issues.

The
specific
size
and
operating
characteristics
of
the
UVT
monitor
will
vary
dependent
on
the
UV
manufacturer.
If
an
on­
line
UVT
monitor
is
included
in
the
design,
it
is
important
to
provide
adequate
space
and
access
to
an
electrical
supply
for
installation
of
the
monitor
and
to
include
appropriate
sample
taps
and
drains
in
the
design
for
the
withdrawal
and
discharge
of
sample
water.
The
sample
line
should
be
equipped
with
a
valve
to
isolate
the
unit.
If
insufficient
pressure
is
available
in
the
system,
then
a
sample
pump
should
be
installed.

3.3.3.5
Flow
Measurement
Flowrate
is
one
of
the
operating
conditions
a
utility
must
routinely
monitor
(
40
CFR
141,
Subpart
W,
Appendix
D).
To
maintain
regulatory
compliance,
the
flowrate
through
a
UV
reactor
must
be
known
to
ensure
that
flow
is
within
the
validated
range.
Section
3.3.1.2
discusses
flow
measurement
and
control
options.
If
flow
meters
are
installed,
the
flow
signal
should
be
displayed
locally
or
be
input
directly
into
a
control
loop
for
the
UV
reactor,
a
SCADA
system,
or
both.

3.3.3.6
Lamp
Age
Each
lamp
or
an
integral
bank
of
lamps
should
be
monitored
for
operating
time.
Lamp
replacement
should
be
based
on
the
dose
delivery
and
the
age
of
the
lamp.
Initially,
the
number
of
lamp
hours
used
to
trigger
lamp
replacement
can
be
estimated
based
on
UV
manufacturer
recommendation
and
validation
data.
Later,
the
actual
frequency
of
replacement
should
be
correlated
to
the
operating
performance
of
the
UV
installation.
Frequent
restarting
of
the
lamps
may
reduce
their
useful
life.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
41
June
2003
3.3.3.7
Lamp
and
Reactor
Status
Lamp
status
is
one
of
the
operating
conditions
a
utility
must
routinely
monitor
(
40
CFR
141,
Subpart
W,
Appendix
D).
In
addition
to
the
status
of
individual
lamps,
whether
the
reactor
is
on­
line
or
off­
line
should
also
be
monitored
and
indicated
locally
and
remotely.
Power
and
valve
status
are
two
methods
that
utilities
can
consider
to
perform
such
monitoring.

3.3.3.8
Alarms
and
Control
Systems
Interlocks
Many
UV
reactor
signals
and
alarms
are
specific
to
the
UV
installation
and
the
level
of
automation
employed.
Alarms
may
be
designated
as
minor,
major,
or
critical,
depending
on
the
severity
of
the
condition
being
indicated.
The
same
alarm
condition
may
represent
a
different
level
of
severity
dependent
on
the
conditions
under
which
the
UV
reactor
was
validated,
the
type
of
UV
reactor,
the
control
strategy,
and
the
disinfection
objectives
of
the
utility.
For
example,
if
a
UV
reactor
was
validated
with
one
lamp
out
of
service,
a
single
lamp
failure
alarm
may
be
a
minor
alarm.
Had
the
reactor
been
validated
with
all
lamps
in
operation,
then
a
single
lamp
failure
may
be
a
major
alarm.
At
a
minimum,
alarm
conditions
should
be
displayed
locally.
The
use
of
an
audible
alarm
may
be
beneficial.
If
UV
reactors
will
frequently
be
unstaffed,
provisions
should
also
be
included
in
the
design
to
allow
remote
monitoring.

A
minor
alarm
generally
indicates
that
a
UV
reactor
needs
maintenance
but
that
the
UV
reactor
is
not
operating
out
of
compliance.
For
example,
a
minor
alarm
would
occur
when
the
end­
of­
lamp­
life
is
reached,
indicating
the
possible
need
for
lamp
replacement.
A
major
alarm
indicates
that
the
UV
reactor
needs
immediate
maintenance
(
e.
g.,
the
UV
intensity
sensor
value
has
dropped
below
the
validated
setpoint)
and
that
the
unit
may
be
operating
off­
specification.
Based
on
the
utility's
disinfection
objectives,
this
condition
may
also
be
handled
as
a
critical
alarm.
A
critical
alarm
typically
shuts
the
unit
down
until
the
cause
of
the
alarm
condition
is
remedied.
An
example
of
a
more
typical
critical
alarm
is
the
UV
reactor
temperature
exceeding
a
pre­
determined
maximum
value,
resulting
in
automatic
shutdown
to
prevent
overheating
and
equipment
damage.

The
designer
should
work
with
the
UV
manufacturer
to
determine
what
elements
of
the
control
system
are
integral
to
the
UV
reactor
and
what
will
be
addressed
through
the
installation
of
supplemental
controls
and
equipment.
For
installations
with
multiple
UV
reactors,
a
common,
master
control
panel
may
be
necessary
to
enable
sequencing
of
the
UV
reactors
and
to
allow
the
UV
reactor
operations
to
be
optimized.
Table
3.8
summarizes
typical
UV
reactor
monitoring
and
alarms;
additional
detail
is
provided
in
section
5.4.
Many
of
the
alarms
shown
will
be
integral
to
the
UV
reactor
control
panel.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
42
June
2003
Table
3.8
Typical
Alarm
Conditions
for
UV
Reactors
Alarm/
Sensor
Purpose/
Descriptions
Lamp
Age
 
Minor
alarm
occurs
when
run­
time
for
lamp
indicates
end
of
defined
operational
lamp
life.
Calibrate
UV
Intensity
Sensor
 
Minor
alarm
occurs
when
UV
intensity
sensor
needs
calibration
based
on
operating
time.
Differential
Pressure
Out
of
Range
(
When
Differential
Pressure
is
Used
for
Flow
Split
Confirmation)
 
Necessary
only
if
a
single
master
flow
meter
is
used.
 
Minor
alarm
occurs
if
pressure
drop
across
parallel,
identical
UV
reactors
indicates
unequal
flow
split.
 
Major
alarm
occurs
if
differential
pressure
across
a
given
UV
reactor
indicates
flow
outside
of
the
validated
range.
Low
UV
Dose
 
Major
alarm
occurs
when
dose
condition
falls
below
required
dose.
 
Triggered
by
signals
gathered
by
control
system
and
compared
to
validated
UV
reactor
dose
requirements.
Low
UV
Intensity
 
Major
alarm
occurs
when
intensity
falls
below
design
conditions.
Low
UV
Transmittance
 
Major
alarm
occurs
when
UVT
falls
below
design
conditions.

High/
Low
Flow
 
Major
alarm
occurs
when
flowrate
falls
outside
of
validated
range.
 
Based
on
measurement
from
dedicated
flow
meters
or
calculated
based
on
total
flowrate
divided
by
number
of
units
operating.
Lamp/
Ballast
Failure
 
Major
alarm
occurs
when
a
single
lamp/
ballast
failure
is
identified.
 
Critical
alarm
occurs
when
multiple
lamp/
ballast
failures
are
identified.
Low
Liquid
Level
 
Critical
alarm
occurs
when
liquid
level
within
the
UV
reactor
drops
and
potential
for
overheating
increases.
High
Temperature
 
Critical
alarm
occurs
when
the
temperature
within
the
UV
reactor
or
ballast
exceeds
a
setpoint.
Mechanical
Wiper
Function
Failure
 
Needed
only
if
a
mechanical
wiper
system
is
used.
 
Critical
alarm
occurs
if
wiper
function
fails.
Note:
Alarm
conditions
and
relative
severity
shown
above
may
vary
dependent
on
specific
conditions
under
which
the
UV
reactor
is
validated,
the
type
of
UV
reactor,
the
control
strategy,
and
the
disinfection
objectives
of
the
utility.

3.3.4
Electrical
Power
Configuration
The
electrical
power
configuration
that
is
used
should
take
into
account
the
findings
of
the
power
quality
assessment
conducted
during
the
planning
phase
described
in
section
3.1.3.3,
the
power
requirements
of
the
selected
equipment,
and
the
disinfection
objectives
and
control
strategy
of
the
utility.
Issues
that
should
be
addressed
include
harmonic
distortion
and
offspecification
operation
due
to
power
quality
problems
(
fluctuation
in
line
voltage).

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
43
June
2003
3.3.4.1
Power
Requirements
The
proper
supply
voltage
and
total
load
requirements
must
be
coordinated
with
the
UV
manufacturer,
considering
the
available
power
supply.
In
addition,
the
power
needs
for
each
of
the
UV
reactor
components
may
differ.
For
example,
the
UV
reactors
may
require
a
3­
phase,
480­
volt
service
while
the
on­
line
UVT
monitor
may
need
a
single
phase,
110­
volt
service.
The
method
of
handling
the
power
feed
must
be
carefully
coordinated
to
ensure
all
electrical
equipment
and
services
are
included
and
to
clearly
establish
the
responsible
party
for
each
element
of
the
electrical
supply
(
e.
g.,
primary
service,
transformer,
secondary
service).
Excluding
high
service
pumping,
the
electrical
load
from
UV
reactors
will
typically
be
one
of
the
larger
loads
at
the
WTP.

Due
to
the
varying
nature
of
UV
reactor
loads,
current
and
voltage
harmonic
distortion
can
be
induced.
Such
disturbances
can
result
in
electrical
system
problems,
including
overheating
of
some
power
supply
components
and
effects
on
other
critical
systems
such
as
VFDs,
program
logic
controllers
(
PLCs),
and
computers.
Proper
selection
of
the
UV
reactors,
including
a
thorough
analysis
of
the
potential
for
the
equipment
to
induce
harmonic
distortion,
should
minimize
the
potential
for
harmonic
distortion.
Another
method
for
controlling
harmonics
is
to
use
a
transformer
with
Delta
Wye
connections
to
isolate
the
UV
reactors
from
the
remainder
of
the
WTP
power
system.
The
delta­
connected
primary
feed
could
be
designed
and
sized
to
trap
and
moderate
any
induced
harmonics.
The
Wye­
connected
secondary
should
be
solidly
grounded
so
that
the
ballasts
are
powered
from
a
grounded
source
in
accordance
with
electrical
code
requirements.
If
a
separate
transformer
for
the
UV
reactors
is
impractical,
harmonic
filters
could
be
added
to
the
UV
reactor
power
supply
to
control
distortion.
Regardless
of
the
method
used
to
address
harmonic
issues,
electrical
acceptance
testing
during
start­
up
should
include
a
harmonic
analysis
to
verify
that
the
UV
reactor
harmonics
are
not
affecting
other
electrical
components
at
the
WTP.

3.3.4.2
Backup
Power
Supply
The
continuous
operation
of
the
UV
reactors
is
highly
dependent
on
its
power
supply.
This
dependence,
when
combined
with
the
sensitivity
of
the
UV
reactors
to
power
fluctuations,
increases
the
importance
of
a
high
quality,
dependable
power
supply.
The
utility
should
work
with
the
State
to
establish
specific
power
reliability
objectives
for
the
UV
installation,
as
power
reliability
may
directly
affect
the
utility's
ability
to
meet
the
State's
allowable
off­
specification
requirements.
As
discussed
in
section
3.1.3.3,
minor
power
transients
can
lead
to
lamp
outages.
If
the
power
reliability
objectives,
and,
consequently,
the
disinfection
objectives,
cannot
be
met
solely
by
relying
on
the
commercial
power
supply,
then
the
use
of
a
backup
power
supply
(
i.
e.,
backup
generator,
separate
commercial
service,
and/
or
battery­
supported
UPS)
may
be
necessary.
If
an
existing
backup
power
supply
is
in
place,
the
load
capacity
of
this
supply
should
be
assessed
to
determine
if
it
is
able
to
accept
the
additional
load
associated
with
the
UV
reactors.
Additionally,
the
time
needed
to
transfer
from
the
primary
power
supply
to
a
backup
power
supply
and
the
potential
effect
of
the
transfer
time
on
compliance
with
the
State's
allowable
off­
specification
operation
should
be
determined.

An
alternate
backup
power
supply
may
be
needed
if
a
backup
power
supply
is
not
inplace
or
the
available
load
capacity
is
insufficient
to
handle
the
new
load
associated
with
the
UV
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
44
June
2003
reactor
equipment.
The
type
of
backup
power
supply
that
is
needed
will
depend
on
the
frequency
and
duration
of
the
power
interruptions
and
the
potential
for
those
interruptions
to
result
in
off­
specification
operation
of
the
UV
reactors.
If
power
quality
issues
are
infrequent
and
of
short
duration
(
on
the
order
of
seconds
or
minutes),
it
is
possible
that
a
backup
power
supply
may
be
unnecessary,
or
a
simple
backup
power
supply
may
be
sufficient.
If
the
frequency
of
power
outages
increases
or
the
duration
of
the
outages
increases,
the
need
for
a
more
extensive
backup
power
supply
becomes
more
significant.

If
a
backup
power
supply
is
necessary,
but
continuous
power
is
not
needed,
the
use
of
a
traditional
diesel
or
natural
gas­
fired
backup
generator
set,
a
standby
UPS,
or
a
rotary
UPS
may
be
adequate.
Should
a
continuous
power
supply
be
needed
to
meet
reliability
objectives,
the
use
of
a
line­
interactive
UPS
will
be
necessary.
The
line­
interactive
UPS
provides
a
continuous
power
supply,
but
is
generally
less
efficient,
has
a
lower
starting
current,
and
costs
more
than
a
similarly
sized
standby
UPS.
Typically,
a
line­
interactive
UPS
would
be
installed
in
conjunction
with
a
backup
generator
to
provide
a
cost­
effective
backup
power
supply
for
longer
duration
power
interruptions
or
for
frequent,
shorter
duration
power
interruptions.
Although
unlikely
to
be
a
requirement
for
compliance
monitoring,
it
may
be
beneficial
to
include
a
data
logger
that
records
instances
of
UPS
operation
as
part
of
the
UPS
system
design.

The
elements
that
should
be
considered
when
assessing
the
need
for
a
backup
power
supply
for
a
UV
reactor
are
somewhat
unique
when
compared
to
those
associated
with
more
typical
WTP
equipment.
However,
once
it
is
determined
that
a
backup
power
supply
is
necessary,
the
design
for
a
UV
reactor
is
very
similar
to
that
for
any
other
equipment
or
treatment
process.
Factors
that
should
be
considered
during
design
include
isolation,
in­
rush
current,
purchase
and
installation
cost,
maintenance
requirements,
voltage
regulation,
electrical
surge
protection,
attenuation
of
harmonic
current,
run­
time,
transformer
continuity,
and
the
ability
to
operate
with
other
power
supply
equipment.
In
most
circumstances,
a
UPS
should
not
be
used
without
a
backup
generator
because
of
the
battery
reserve
necessary
to
power
a
UV
installation
for
longer
durations.
To
minimize
capital
cost,
the
battery
reserve
time
should
be
sufficient
to
allow
the
power
supply
to
switch
to
the
backup
generator.

3.3.4.3
Ground
Fault
Interrupt
and
Electrical
Lockout
Ground
fault
interrupt
(
GFI)
is
an
important
safety
feature
for
any
electrical
system
in
contact
with
water,
including
UV
reactors.
All
UV
reactor
suppliers
should
provide
GFI
circuits
for
their
lamps,
which
should
be
included
in
the
specifications
that
are
developed
for
equipment
procurement.
For
a
GFI
to
function
properly,
the
transformer
in
the
UV
reactor
ballast
must
not
be
isolated
from
the
ground.
If
the
UV
reactor
ballast
isolates
the
output
from
the
ground,
ground
faults
will
not
be
properly
detected,
and
safety
can
be
compromised.

Provisions
enabling
the
UV
reactors
to
be
isolated
and
locked
out
for
maintenance,
both
hydraulically
and
electrically,
should
be
included
in
the
design.
Control
of
all
lockout
systems
should
remain
local;
however,
when
appropriate,
the
status
of
local
lockouts
could
be
monitored
remotely.
In
all
cases,
the
design
must
comply
with
electrical
code
and
policy
requirements
for
equipment
lockout.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
45
June
2003
3.3.5
UV
Installation
Layouts
Once
the
previous
design
elements
(
i.
e.,
section
3.3.1
through
section
3.3.4)
have
been
evaluated,
an
installation
layout
can
be
prepared
as
part
of
the
equipment
procurement
document.
The
layout
should
take
into
account
the
findings
of
all
previous
work.
Because
the
design
process
is
iterative
and
many
elements
of
the
layout
are
dependent
on
the
specific
UV
reactors
that
are
used
and
the
validation
scenario
that
is
proposed,
the
layout
may
change
after
the
UV
reactors
are
selected
and
any
additional
space
constraints
are
identified.

3.3.5.1
Site
Layout
Site
layout
for
a
UV
installation
is
generally
similar
to
the
layout
of
any
treatment
process.
When
locating
the
UV
installation,
access
for
construction,
operation,
and
maintenance
should
be
addressed.
The
availability
of
adequate
existing
infrastructure
(
e.
g.,
power,
drains,
lifting
devices)
is
also
important.
In
general,
when
compared
to
other
treatment
processes
at
a
WTP,
the
UV
installation
has
a
relatively
small
footprint.

3.3.5.2
UV
Installation
Layout
In
large
part,
the
piping
layout
will
be
dictated
by
the
validated
hydraulic
conditions
because
the
inlet
and
outlet
conditions
for
the
installed
UV
reactors
should
be
equal
to
or
better
than
the
hydraulic
conditions
used
during
validation.
Additional
details
on
the
relationship
between
the
validated
inlet
and
outlet
configuration
and
the
actual
installed
configuration
are
given
in
section
C.
3.1.5.
Nevertheless,
the
designer
can
prepare
a
reasonable
UV
installation
layout
based
on
the
type
of
technology
(
i.
e.,
LPHO
versus
MP),
the
number
of
UV
reactors
that
is
needed,
and
the
manner
in
which
flow
is
controlled
and
measured.
This
layout
can
then
be
used
in
the
selection
and
procurement
of
the
UV
reactors.

Most
UV
reactors
available
for
drinking
water
applications
are
of
a
closed­
chamber
type.
Filtered
water
is
conveyed
via
pipes
or
covered
channels
to
a
series
of
UV
reactors
for
primary
disinfection.
As
such,
laying
out
UV
installation
typically
involves
designing
the
method
by
which
water
is
divided
between
UV
reactors
(
channel
or
piping),
and
routing
the
sections
of
pipe
between
inlet
and
discharge
headers
in
which
the
UV
reactors
are
inserted
via
flanged
connections
(
although
other
types
of
connections
may
be
used).
The
number
and
configuration
of
the
UV
reactors
will
vary
depending
on
lamp
type/
reactor
design,
reactor
size,
flow
range
to
be
treated,
control
strategy,
and
degree
of
redundancy.

Although
most
components
of
UV
reactors
are
fairly
compact,
it
is
important
not
to
underestimate
the
necessary
space
for
the
building
that
will
house
the
UV
installation.
In
addition
to
those
items
identified
in
section
3.1.6.2,
the
following
factors
should
be
considered
in
the
layout
for
the
UV
installation:

 
The
length
of
straight­
run
piping
before
and
after
each
flow
meter
to
achieve
the
proper
hydraulic
conditions
for
accurate
and
repeatable
flow
measurement
(
if
applicable)

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
46
June
2003
 
Field
instrumentation
 
Isolation
valves
and
flow
control
devices
 
Control
and
power
panels,
and
code­
required
clear
space
 
Potential
space
for
power
monitors
and
UPS
systems
 
Drain
provisions
for
the
process
area
and
to
permit
UV
reactor
draining
 
Provisions
for
future
expansion
of
UV
disinfection
capacity
Components
of
the
UV
reactors
are
typically
located
inside
a
building
for
protection
from
the
weather
and
to
provide
a
clean,
convenient
area
for
maintenance.
The
UV
reactors
themselves,
associated
electrical
components
and
controls,
and
electrical
support
equipment
such
as
a
UPS
should
be
enclosed.
There
are
installations,
however,
where
UV
reactors
and
control
panels
are
uncovered.
Prior
to
implementing
an
uncovered
installation,
it
is
recommended
that
the
State
and
UV
manufacturer
be
consulted.
Any
exposed
equipment
and
control
panels
should
be
rated
for
the
anticipated
environment
and
appropriate
site
security
should
be
in­
place
to
restrict
public
access.

The
power
and
control
panels
associated
with
UV
reactors
should
be
located
so
that
there
is
adequate
space
for
panel
doors
to
be
opened
without
interference,
and
to
allow
unhindered
access
to
the
UV
reactors
with
panel
doors
open.
In
selecting
the
location
of
the
power
and
control
panels,
UV
manufacturer
cable
length
limitations
should
not
be
exceeded.
The
maximum
allowable
cable
length
is
UV
manufacturer­
specific
and
may
be
less
than
30
feet.
If
harmonic
feedback
is
a
concern,
extra
room
should
be
provided
for
power
conditioning
equipment.

When
allotting
space
for
maintenance
activities,
adequate
space
to
remove
the
lamps
and
the
lamp
wiper
assembly
should
be
provided.
In
some
cases,
access
may
be
needed
on
both
sides
of
the
UV
reactor.
In
addition,
provisions
should
be
included
to
collect
and
convey
water
that
is
discharged
during
maintenance
activities.

Certain
UV
reactors
need
maintenance
involving
an
OCC
procedure
in
which
a
UV
reactor
is
taken
off­
line,
isolated,
drained,
filled
with
a
cleaning
solution,
cleaned,
flushed,
and
returned
to
service.
The
OCC
equipment
is
typically
self­
contained
and
the
cleaning
chemical
is
recirculated.
Where
applicable,
sufficient
space
around
the
UV
reactors
should
be
maintained
to
provide
access
for
the
OCC
procedure.
In
addition,
the
OCC
solution
often
has
specific
handling
requirements.
Appropriate
drains,
storage,
and
health
and
safety
equipment
(
e.
g.,
emergency
eyewash
station)
should
be
provided
as
recommended
by
the
chemical
manufacturer.

Sample
taps
are
recommended
upstream
and
downstream
of
each
UV
reactor
within
the
lateral
pipe.
The
sample
taps
may
be
used
for
the
collection
of
water
quality
samples
or
may
be
used
during
validation
testing
if
on­
site
validation
is
necessary.
If
on­
site
validation
will
be
conducted,
the
number
and
location
of
sample
and
feed
ports
should
be
coordinated
with
the
UV
manufacturer
or
third
party
validation
service
to
comply
with
the
recommendations
of
the
selected
validation
protocol.
Additional
detail
on
the
locations
of
sample
taps
and
other
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
47
June
2003
validation­
related
appurtenances,
as
well
as
the
methods
used
to
validate
a
reactor
are
provided
in
section
C.
3.1.4.

Drain
valves
or
plugs
should
be
located
on
each
lateral
between
the
two
isolation
valves.
In
many
cases,
the
UV
manufacturer
may
have
already
incorporated
a
drain
into
the
UV
reactor
design.
Drain
valves
should
also
be
provided
at
one
or
more
low
points
in
the
UV
installation
to
enable
the
UV
reactor
to
be
fully
drained
for
future
maintenance
activities.

3.3.6
Elements
of
UV
Reactor
Specifications
Table
3.9
summarizes
the
elements
that
should
be
considered
in
developing
equipment
specifications
for
the
UV
reactors.
The
information
included
in
Table
3.9
is
not
exhaustive
and
should
be
modified
to
meet
the
specific
needs
of
the
utility.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
48
June
2003
Proposal
Draft
Table
3.9
Recommended
Content
for
UV
Reactor
Specifications
Specification
Item
Purpose/
Description
Flowrate
Maximum,
minimum,
and
average
flowrates
should
be
clearly
identified.
The
minimum
and
maximum
flowrates
must
be
within
the
range
of
validation
flowrates.
The
minimum
flowrate
is
important
to
avoid
overheating
with
MP
reactors.
UV
Dose
The
required
reduction
equivalent
dose
as
well
as
the
validation
technique
that
will
be
used
to
measure
the
dose
should
be
established.
Additional
detail
is
provided
in
Chapter
4.
Water
Quality
and
Environment
The
following
water
quality
criteria
should
be
included:
­
Influent
temperature
­
Total
Hardness
­
Turbidity
­
pH
­
UV
Transmittance
at
254
nm
­
Iron
­
Spectral
absorbance
200­
300
nm
(
MP
reactors
only)
For
some
parameters,
a
design
range
may
be
most
appropriate.
UV
Intensity
Sensors
It
is
recommended
that
at
least
one
UV
intensity
sensor
be
specified
per
UV
reactor.
The
number
of
reference
sensors
that
should
be
determined
based
on
the
time
and
labor
associated
with
checking
and
maintaining
the
duty
sensors.
Redundancy
If
a
combined
filter
effluent
UV
reactors
are
used,
it
is
recommended
that
at
least
one
completely
redundant
UV
reactor
be
specified
as
a
standby.
For
other
configurations,
the
designer
should
determine
the
appropriate
redundancy
based
on
the
State's
requirements
and
the
utility's
disinfection
objectives.
Hydraulics
The
following
hydraulic
information
should
be
specified:
­
Maximum
system
pressure
at
the
UV
reactor
­
Maximum
allowable
headloss
through
the
UV
reactor
­
Special
surge
conditions
that
may
be
experienced
­
Hydraulic
constraints
based
on
site­
specific
conditions
and
validated
conditions
(
e.
g.,
upstream
and
downstream
straight
pipe
lengths)
Size/
Location
Constraints
Any
size
constraints
or
restrictions
on
the
location
of
the
UV
reactor
or
control
panels
(
e.
g.,
space
constraints
with
in­
line
installation).
Validation
The
specifications
should
establish
the
validation
protocol
that
will
be
followed,
provide
the
conditions
under
which
the
validation
will
be
conducted
(
e.
g.,
water
quality,
flow
range,
hydraulic
conditions,
UVT),
and
require
the
submittal
of
a
validation
report
(
40
CFR
141.730).
Control
Strategy
and
Operating
Sequence
The
specification
should
provide
a
narrative
description
of
the
operating
sequence
and
control
strategy
for
the
UV
reactors.
Lamp
Sleeves
At
a
minimum,
the
following
items
should
be
specified:
­
Lamp
sleeves
should
be
annealed
to
remove
internal
stress.
­
UV
manufacturer
should
perform
QA
/
QC
checks
of
a
fraction
of
each
lot
using
a
polarized
light
or
other
approved
method.
­
UV
manufacturer
should
submit
documentation
on
the
integrity
of
their
sleeve,
monitoring
practices,
and
rationale
for
using
a
given
internal
QA
/
QC
frequency.
­
UV
manufacturer
should
submit
calculations
showing
the
maximum
allowable
pressure
for
the
lamp
sleeves
and
the
maximum
bending
stress
experienced
by
the
lamp
sleeves
under
the
maximum
specified
flow
conditions.
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
49
June
2003
Table
3.9
Recommended
Content
for
UV
Reactor
Specifications
(
continued)

Specification
Item
Purpose/
Description
Safeguards
At
a
minimum,
the
following
UV
reactor
alarms
should
be
specified:
­
Lamp
or
ballast
failure
­
Low
UV
intensity
or
low
UV
dose
(
dependent
on
control
strategy
used)
­
High
temperature
­
Low
or
high
flow
­
Wiper
failure
(
as
applicable)
­
Other
alarms
discussed
in
section
3.3.3.8,
as
appropriate
Control
Systems
At
a
minimum
the
following
signals
and
indications
should
be
specified:
­
UV
reactor
status
­
UV
intensity
­
Individual
lamp
status
­
Lamp
cleaning
cycle
and
history
­
Accumulated
runtime
for
individual
lamps
­
Influent
flowrate
At
a
minimum
the
following
UV
reactor
controls
(
as
applicable)
should
be
specified:
­
UV
dose
setpoints,
lamp
intensity
setpoints,
or
UVT
setpoints
(
dependent
on
control
strategy
used)
­
UV
reactor
on/
off
control
­
UV
reactor
manual/
auto
control
­
UV
reactor
local/
remote
control
­
Manual
lamp
power
level
control
­
Manual
lamp
cleaning
cycle
control
­
Automatic
lamp
cleaning
cycle
setpoint
control
Performance
Guarantee
The
performance
guarantee
should
specify
that
the
equipment
provided
under
the
UV
reactor
specification
should
meet
the
performance
requirements
stated
in
the
specification
for
an
identified
period.
The
following
specific
performance
criteria
may
be
included:
­
Allowable
headloss
at
each
of
the
design
flowrates.
­
Estimated
power
consumption
under
the
design
operating
conditions.
­
Disinfection
capacity
of
each
reactor
under
the
design
water
quality
conditions.

Warranties
A
physical
equipment
guarantee
and
UV
lamp
guarantee
should
be
specified.
The
specific
requirements
of
these
guarantees
will
be
at
the
discretion
of
the
utility
and
engineer.

3.3.6.1
Information
Provided
by
Manufacturer
in
UV
Reactor
Bid
It
is
important
that
UV
manufacturers
provide
adequate
information
when
bidding
to
enable
the
designer
to
conduct
a
proper,
timely
review
of
the
proposed
equipment.
Suggested
information
to
be
obtained
from
the
UV
manufacturer
is
presented
in
Table
3.10.

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
50
June
2003
Table
3.10
Recommended
Information
to
be
Provided
by
UV
Manufacturer/
Vendor
Item
Purpose
Design
Parameters
Demonstration
of
an
understanding
of
the
design
parameters
for
the
UV
reactors.
All
UV
reactor
design
parameters
from
the
contract
documents
should
be
repeated
in
the
proposed
UV
reactor
submittal
information.
Summary
of
Design
A
summary
of
the
equipment
proposed
(
number
of
UV
reactors,
lamp
type)
and
specify
equipment
redundancies.
Reactor
Technical
Specifications
Ability
of
proposed
UV
reactors
to
meet
technical
specifications
and
an
explanation
of
any
exceptions
taken.

UV
Manufacturer's
Experience
Information
on
project
experience,
including
previous
installations
and
references.

UV
Intensity
Sensor
Information
on
the
UV
intensity
sensor(
s)
including
acceptance
angle,
external
dimensions,
working
range
in
mW/
cm2,
spectral
response,
measurement
uncertainty,
environmental
requirements,
linearity
and
temperature
stability.
Data
and
calculations
should
be
provided
showing
how
the
total
measurement
uncertainty
of
the
sensor
is
derived
from
the
individual
sensor
properties.
(
See
sections
4.3.2.3
and
C.
4.7
)
Validation
Data
UV
reactor
validation
data
as
described
in
Appendix
C
of
these
Guidelines.
If
onsite
validation
is
proposed,
validation
data
for
the
UV
reactors
from
other,
similar
installations
should
be
included
to
provide
a
baseline
comparison
to
the
proposed
operating
conditions.
Upstream
and
Downstream
Hydraulic
Requirements
A
statement
of
the
length
of
straight
pipe
and
hydraulic
conditions
necessary
upstream
and
downstream
from
the
UV
reactor
to
ensure
the
desired
flow
profile
is
maintained
and
the
design
conditions
are
met.

Power
Requirements
The
power
needs
of
each
UV
reactor
and
which
elements,
including
electrical
cable
and
wiring,
are
included
as
part
of
their
equipment.
Cleaning
Strategy
The
strategy
that
will
be
employed
for
cleaning
the
UV
lamps
in
the
UV
reactor.

Control
Strategy
The
proposed
UV
reactor
control
strategy,
including
manual
and
automatic
control
schemes
and
a
listing
of
inputs,
outputs,
and
the
types
of
signals
that
are
available
for
remote
monitoring
and
control.
Reactor
Data
The
materials
of
construction,
dimensions
of
the
UV
reactors
and
ancillary
equipment,
a
listing
of
spare
parts,
and
a
sample
operations
and
maintenance
manual.
Safeguards
The
safeguards
built
into
the
UV
reactor
and
accompanying
equipment,
such
as
high
temperature
protection,
wiper
failure
alarms,
and
lamp
failure
alarms.
Warranties
A
statement
of
the
proposed
UV
reactor
guarantees,
including
the
physical
equipment,
the
UV
lamp,
and
the
system
performance
guarantee.
Any
exceptions
should
be
indicated
and
explained.

Warranties
The
UV
reactor
specification
should
include
suitable
written
guarantees
regarding
physical
equipment,
UV
lamps,
and
performance.

It
is
recommended
that
the
UV
lamp
guarantee
specify
that
each
lamp
is
warranted
to
provide
the
lamp
output
necessary
to
meet
the
required
reduction
equivalent
dose
(
RED)
under
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
51
June
2003
the
design
conditions
for
a
minimum
number
of
operating
hours,
which
will
vary
depending
on
lamp
type.
To
limit
the
UV
manufacturer's
liability
and
to
potentially
reduce
the
contingency
costs
included
in
their
equipment
bid
prices,
the
guarantee
could
be
prorated
after
a
specified
number
of
operating
hours.
It
is
important
that
the
appropriate
lamp
fouling/
aging
factor
be
included
in
the
design
conditions
as
discussed
in
section
3.1.3.1.
If
these
specifications
are
not
met,
significant
operation
and
maintenance
costs
may
occur
because
lamps
may
need
to
be
replaced
frequently
for
the
UV
reactors
to
operate
within
the
validated
range.
The
combination
of
lamp
fouling/
aging
factor
and
the
guaranteed
lamp
life
will
make
the
UV
manufacturer
responsible
if
the
UV
lamps
do
not
meet
these
specifications.
The
guaranteed
lamp
life
will
depend
on
the
available
technology
at
the
time
of
the
UV
installation
design
and
will
likely
change
as
lamp
technology
improves.

3.3.7
Final
UV
Installation
Design
After
the
equipment
procurement
document
is
developed
and
competitively
bid,
and
all
bids
have
been
carefully
reviewed,
the
UV
reactors
can
be
selected.
Once
the
UV
reactors
are
selected,
the
designer
can
work
with
the
selected
UV
manufacturer
to
develop
the
final
disinfection
installation
design
based
on
the
specific
needs
and
design
of
the
selected
equipment.
The
hydraulic
design,
instrumentation
and
control
design,
electrical
design,
and
installation
layout
should
be
modified
to
address
the
specific
needs
of
the
selected
equipment
and
to
ensure
that
the
control
strategy
can
be
implemented
within
the
constraints
established
during
the
validation
testing.

Particular
emphasis
should
be
given
to
the
integration
of
the
overall
control
strategy
with
the
alarms,
signals,
and
interlocks
that
are
integral
to
the
UV
reactor
design.
For
designs
with
multiple
UV
reactors,
a
master
control
panel
may
be
necessary
to
enable
the
sequenced
operation
of
the
individual
UV
reactors
and
to
optimize
the
efficiency
of
the
UV
installation.
It
is
critical
that
the
final
design
be
coordinated
with
the
validation
testing
to
ensure
that
validation
criteria
are
sufficient
to
implement
the
proposed
control
strategy
and
to
ensure
that
the
UV
reactors
will
meet
the
utility's
disinfection
objectives
under
the
anticipated
operating
conditions.

3.3.7.1
Design
Drawings
The
drawings
may
include
the
following
content:

 
Existing
conditions
 
Site
work
 
Structural
work
 
Architectural
work
 
Mechanical
work
(
heating,
ventilation,
and
air
conditioning)

Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
52
June
2003
 
Electrical
work
 
Instrumentation
work
3.3.7.2
Specifications
The
content
of
the
specifications
will
vary
dependent
on
the
complexity
and
size
of
the
UV
installation
and
the
selected
method
of
project
delivery.
However,
it
is
likely
that
portions
of
nearly
all
of
the
16­
Division
Construction
Specifications
Institute
(
CSI)
MasterFormat
may
be
necessary.
For
those
UV
installations
that
pre­
purchase
the
UV
reactors,
the
equipment
procurement
document
should
be
included
as
an
appendix
to
the
specifications
to
facilitate
contractor
review
and
installation
of
the
equipment.

3.4
Reporting
To
The
State
Interaction
with
the
State
throughout
the
planning
and
design
phases,
as
well
as
during
development
of
the
reactor
validation
protocol,
is
recommended
to
ensure
that
the
objectives
of
both
the
utility
and
the
State
are
met.

Given
the
relatively
limited
past
use
of
UV
disinfection
in
drinking
water
treatment
and
the
unique
technical
characteristics
of
this
technology,
State
agencies
may
not
have
developed
approval
requirements
specifically
for
UV
disinfection.
This
section
provides
guidance
on
the
information
that
may
be
included
in
submittals
to
the
State.
Utilities
are
urged
to
consult
with
their
State
early
in
their
UV
disinfection
planning
process
to
understand
what
approvals
and
documentation
will
be
required
for
the
use
of
UV
disinfection.

3.4.1
Planning
The
State
may
require
that
a
pre­
design
report
be
submitted
that
summarizes
the
decision
logic
used
to
identify,
evaluate,
and
select
UV
disinfection.
Appendix
K
is
an
example
predesign
report,
including
installation
alternatives
and
analysis.
The
following
items
may
be
addressed
in
the
pre­
design
report:

 
Disinfection
objectives
(
target
organism
and
inactivation)

 
Overall
disinfection
strategy
 
Summary
of
reasons
for
incorporating
UV
disinfection
 
Description
of
the
overall
process
train
 
Description
of
the
proposed
UV
reactors
 
Water
quality
data
Proposal
Draft
3.
Planning
and
Design
Aspects
for
UV
Installations
UV
Disinfection
Guidance
Manual
3­
53
June
2003
 
UV
reactor
reliability
targets
(
i.
e.,
off­
specification
limits)

 
UV
reactor
validation
3.4.2
Equipment
Procurement
If
the
utility
pre­
purchases
the
equipment,
a
separate
procurement
document
would
be
prepared.
The
equipment
procurement
document
should
be
consistent
with
the
pre­
design
report
and
should
include
technical
specifications
and
a
preliminary
layout
of
the
UV
installation.
Details
on
the
recommended
content
of
the
specifications
are
given
in
section
3.3.7.
While
the
State
may
not
require
submittal
of
the
equipment
procurement
document
prior
to
equipment
purchase,
it
is
recommended
that
acceptance
of
a
pre­
design
report
be
received
from
the
State
prior
to
proceeding
with
equipment
purchase.
The
State
should
also
be
notified
of
any
deviations
from
the
pre­
design
report.

3.4.3
Drawings
and
Specifications
The
UV
installation
drawings
and
specifications
should
be
submitted
to
the
State
for
approval.
Under
the
equipment
pre­
purchase
option,
the
drawings
and
specifications
should
address
the
installation
of
the
UV
reactors
and
related
equipment
as
well
as
other
necessary
facility
modifications.
The
specific
items
that
would
be
included
in
this
submittal
are
discussed
in
section
3.3.7.
If
an
alternative
approach
is
used
(
e.
g.,
design­
build
or
design­
build­
operate)
the
level
of
detail
included
in
the
design
documents
will
differ.

3.4.4
Validation
Report/
Start­
up
Confirmation
States
may
request
that
a
validation
report
or
other
preliminary
testing
results
be
submitted.
As
discussed
in
section
3.1.4.3,
validation
may
occur
off­
site
or
on­
site.
If
the
UV
reactors
are
validated
at
an
off­
site
location,
the
validation
report
should
be
available
from
the
UV
manufacturer
and
should
be
a
required
submittal
from
the
UV
manufacturer
as
part
of
either
the
equipment
procurement
documents
or
the
UV
installation
specifications.
If
on­
site
validation
is
used,
a
validation
protocol
should
be
developed
and
accepted
by
the
State
prior
to
implementation.
Following
completion
of
the
on­
site
validation,
a
validation
report
should
be
prepared
and
submitted
to
the
State.
Recommended
validation
protocols
are
provided
in
Appendix
C.

In
addition,
some
States
may
request
that
the
utility
provide
as­
built
documentation
(
i.
e.,
start­
up
confirmation)
certifying
construction
was
completed
in
accordance
with
the
approved
drawings
and
specifications.
Start­
up
confirmation
may
be
most
important
where
alternative
project
delivery
approaches
are
used
and
the
State
does
not
have
the
benefit
of
reviewing
100
percent
design
drawings
and
specifications
prior
to
construction.

Proposal
Draft
4.
Overview
of
Validation
Testing
The
Long
Term
2
Enhanced
Surface
Water
Treatment
Rule
(
LT2ESWTR)
requires
the
use
of
validated
UV
reactors
for
receiving
Cryptosporidium,
Giardia,
or
virus
inactivation
credit
(
40
CFR
141.729(
d)).
The
purpose
of
validating
a
UV
reactor
is
to
provide
confidence
that
the
UV
reactor
can
provide
the
level
of
inactivation
required
for
a
given
application.
The
rule
specifies
only
basic
components
of
a
validation
process
(
presented
in
section
4.1).
Using
those
requirements
as
a
framework,
this
guidance
manual
describes
recommended
procedures
and
data
analysis
for
one
possible
approach
to
validating
a
UV
reactor.
Other
approaches
or
modifications
to
this
approach
may
be
used
at
the
discretion
of
the
State.

The
validation
protocol
provided
in
this
manual
has
two
tiers,
specifying
two
different
methods
for
addressing
uncertainty
with
a
safety
factor
to
determine
the
log
inactivation
credit.
These
tiers
differ
in
level
of
complexity.
Tier
1
is
simplified
while
Tier
2
is
more
complex,
potentially
allowing
for
a
less
conservative
safety
factor
based
on
detailed
knowledge
and
testing
of
equipment
performance.
Appendix
C
provides
all
the
necessary
procedures
and
descriptions
to
complete
a
validation
for
both
Tier
1
and
Tier
2
methods.
This
chapter
provides
a
brief
overview
of
the
validation
process,
describing
all
the
basic
steps
of
the
testing
procedures
and
interpretation
of
results,
with
references
to
Appendix
C
for
more
detailed
descriptions.
For
those
conducting
a
validation
test
of
a
given
reactor,
it
is
important
to
understand
the
background
and
detailed
procedures
described
in
Appendix
C.

4.1
LT2ESWTR
UV
Disinfection
Requirements
This
section
reviews
the
LT2ESWTR
requirements
related
to
UV
reactor
validation
specified
under
40
CFR
141.729(
d)
and
40
CFR141,
Subpart
W,
Appendix
D.

Validation
testing
must
determine
a
set
of
operating
conditions
that
can
be
monitored
by
a
utility
to
ensure
that
the
UV
dose
required
for
a
given
pathogen
inactivation
credit
is
delivered;
and
the
utility
must
then
monitor
to
demonstrate
it
is
operating
within
the
range
of
conditions
under
which
the
reactor
was
validated.

Validation
operating
conditions
must
include,
at
a
minimum,
the
following:

 
UV
intensity
(
as
measured
by
a
UV
intensity
sensor)

 
Flowrate
 
Lamp
status
Many
design
and
equipment
factors
affect
the
UV
dose
delivered
by
the
reactor.
The
validated
operating
conditions
must
account
for
the
following
factors:

 
Lamp
aging
 
Lamp
sleeve
fouling
UV
Disinfection
Guidance
Manual
4­
1
June
2003
Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
2
June
2003
 
UV
transmittance
(
UVT)
of
the
water
 
Inlet
and
outlet
piping
or
channel
configurations
of
the
UV
reactor
 
Dose
distributions
arising
from
the
velocity
profiles
through
the
reactor
 
Failure
of
UV
lamps
or
other
critical
system
components
 
Measurement
uncertainty
of
on­
line
sensors
Unless
the
State
approves
an
alternative
approach,
validation
testing
must
involve
the
following:

 
Full­
scale
testing
of
a
UV
reactor
that
conforms
uniformly
to
the
reactors
used
by
the
utility.

 
Inactivation
of
a
test
microorganism
whose
dose­
response
characteristics
have
been
quantified
with
a
low­
pressure
(
LP)
mercury
vapor
lamp.

4.2
Overview
of
Validation
Process
The
validation
process
determines
the
log
inactivation
achieved
for
a
specific
pathogen
and
relates
it
to
the
operating
conditions
at
the
time
of
the
testing
(
e.
g.,
UV
transmittance
at
254
nm,
or
UVT,
flowrate).
Figure
4.1
shows
the
key
steps
of
a
validation
process,
with
the
differentiation
of
Tier
1
and
Tier
2
approaches.

The
experimental
portion
of
the
validation
process
is
referred
to
as
"
biodosimetry."
It
consists
of
a
UV
reactor
test
that
measures
log
inactivation
of
a
surrogate
(
challenge)
microorganism
under
various
flowrate,
UVT,
and
lamp
power
combinations.
Log
inactivation
is
then
benchmarked
to
the
corresponding
operational
conditions
and
UV
intensity
sensor
values.
Since
the
true
UV
doses
delivered
to
the
challenge
microorganisms
cannot
be
measured
directly
by
the
UV
reactor,
a
separate
test
must
be
conducted
to
relate
the
inactivation
measured
in
the
field
to
a
UV
dose
value.
Current
practice
in
the
UV
industry
uses
a
collimated
beam
to
generate
a
UV
dose­
response
curve
for
a
given
challenge
microorganism
(
log
inactivation
versus
UV
dose).
The
log
inactivation
from
the
biodosimetry
test
is
then
related
to
a
UV
dose
from
the
UV
dose­
response
curve.
This
dose
is
termed
the
reduction
equivalent
dose
(
RED).

Hydraulic
effects,
UV
reactor
equipment,
and
error
in
on­
line
UV
intensity
sensors
all
create
uncertainty
in
translating
an
RED
measured
during
a
validation
test
to
a
given
level
of
pathogen
inactivation
during
routine
operation.
To
account
for
this
uncertainty,
a
safety
factor
should
be
applied
to
the
required
UV
dose
values
for
pathogen
inactivation
credit.
The
required
UV
dose
value
multiplied
by
the
appropriate
safety
factor
is
the
RED
that
should
be
demonstrated
during
a
validation
test
for
a
given
level
of
pathogen
inactivation
credit.
Tier
1
and
Tier
2
approaches
provide
methods
for
incorporating
the
safety
factor
to
determine
the
log
inactivation
credit.

Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
3
June
2003
Figure
4.1
Steps
of
a
Validation
Process
Collimated
Beam
Test
Log
Inactivation
and
corresponding
UV
intensity
sensor
values
Biodosimetry
Test
Challenge
Microorganism
Influent
Sample
Effluent
Sample
1
1
Log
Inactivation
UV
Dose
Generate
UV­
dose
Response
for
Challenge
Microorganism
Determine
UV
Dose
for
Log
Inactivation
Measured
Reduction
Equivalent
Dose
(
RED)

Log
Inactivation
(
from
biodosimetry)

UV
Dose
(
from
collimated
beam)
2
Determine
Log
Inactivation
Credit
Tier
I
(
Preset
Safety
Factors)

­
Experimental
plan
and
results
should
meet
specified
criteria
­
Uses
Tier
I
RED
Target
Tables
(
Table
4.1
and
4.2)
Tier
2
(
Derive
Safety
Factors)

­
Calculate
uncertainties
associated
with
lamps,
sensors,
microbial
measurements,
and
interpolation
of
data.

­
Calculate
bias
associated
with
RED
measurements
of
challenge
microorganism
vs.
pathogen.

­
Calculate
bias
of
MP
lamp
measurements
(
if
applicable).

Calculate
safety
factor
from
the
above
three
results.
3
Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
4
June
2003
4.2.1
Relating
the
Experimental
RED
to
Log
Inactivation
Credit
Chapter
1
presents
the
UV
dose
needed
to
achieve
various
inactivation
credits
for
Cryptosporidium,
Giardia,
and
viruses.
These
dose
requirements
were
derived
from
batch
(
collimated
beam)
dose­
response
data
and
account
for
the
uncertainty
and
statistical
variability
in
the
dose­
response
of
the
pathogen.

There
is
significant,
additional
uncertainty
associated
with
applying
these
batch
data
to
full­
scale,
continuous
flow
testing
results.
This
additional
uncertainty
associated
with
UV
reactor
validation
and
on­
line
dose
monitoring
should
also
be
considered
when
determining
the
log
inactivation
credit
from
UV
reactor
validation.
To
account
for
this
uncertainty,
the
RED
measured
during
validation
should
be
greater
than
the
dose
requirement
multiplied
by
a
safety
factor.
The
safety
factor
incorporates
random
uncertainty
and
corrections
for
expected
variation,
and
is
defined
according
to
Equation
4.1:

(

e
B
B
SF
Poly
RED
+
×
×
=
1
)
Equation
4.1
where
BRED
=
RED
bias
BPoly
=
Polychromatic
bias
e
=
Expanded
uncertainty
expressed
as
a
fraction
The
RED
bias
is
a
correction
that
accounts
for
the
difference
between
the
expected
dose
delivered
to
the
target
pathogen
and
the
actual
dose
measured
using
a
challenge
microorganism
during
biodosimetry.
That
is,
the
RED
measured
for
two
microorganisms
is
not
identical
if
the
dose­
response
behavior
of
the
two
microorganisms
is
different.
The
magnitude
of
the
difference
will
depend
on
the
dose
distribution
of
the
UV
reactor
and
the
unique
inactivation
kinetics
of
the
challenge
microorganism
and
target
pathogen.
If
the
challenge
microorganism
is
more
resistant
to
UV
light
than
the
target
pathogen,
the
RED
measured
during
validation
will
be
greater
than
the
expected
dose
delivered
to
the
pathogen.
If
the
challenge
microorganism
is
as
sensitive
or
more
sensitive
to
UV
light
than
the
target
pathogen,
the
RED
bias
has
a
value
of
one.
Appendix
F
describes
this
concept
in
more
detail.

The
polychromatic
bias
is
a
correction
for
the
spectral
differences
in
the
lamp
output,
lamp
sleeve
UV
transmittance,
water
UVT,
and
action
spectrum
between
validation
and
operation
of
a
UV
reactor.
This
bias
only
applies
to
polychromatic
lamps.

The
expanded
uncertainty,
e,
accounts
for
the
uncertainty
in
the
measurements
taken
during
validation
and
associated
with
the
equipment
(
e.
g.,
UV
intensity
sensors)
used
to
monitor
dose
delivery.

Appendix
F
discusses
in
greater
detail
the
basis
for
the
uncertainty
and
bias
terms
of
the
safety
factor.
Later
sections
of
this
chapter
and
Appendix
C
describe
the
application
of
the
safety
factor.

Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
5
June
2003
4.2.1.1
Tier
1
and
Tier
2
Approaches
for
Establishing
Inactivation
Credit
As
stated
previously,
the
Tier
1
and
Tier
2
approaches
differ
in
the
complexity
of
the
method
used
to
determine
the
log
inactivation
credit
based
on
the
RED
measured
during
biodosimetry.
The
Tier
1
approach
provides
RED
target
values
to
be
met
during
validation
that
correspond
to
the
log
inactivation
credit
(
presented
in
Tables
4.1
and
4.2).
These
RED
values
incorporate
pre­
determined
safety
factors
based
on
characteristics
of
the
UV
reactor
and
validation
testing
(
section
4.6
provides
further
details).
In
the
Tier
2
approach,
the
user
calculates
the
safety
factor
using
detailed
knowledge
of
the
equipment
and
testing
conditions
and
then
applies
it
to
the
required
dose.
This
allows
the
user
to
optimize
their
experimental
methods
which
may
reduce
the
safety
factor.

4.2.2
Location
and
Application
of
Validation
Testing
Validation
testing
may
be
conducted
either
on­
site,
being
the
location
where
the
UV
reactor
will
be
installed
and
operating,
or
off­
site.
Off­
site
validation
may
be
conducted
at
either
a
manufacturer's
facility
or
at
a
centralized
facility
dedicated
to
validating
a
variety
of
UV
equipment.

Reactors
may
be
validated
for
a
specific
WTP
or
may
validate
under
a
wide
array
of
conditions
for
a
variety
of
treatment
applications.
In
addition
to
a
range
of
operating
conditions
(
e.
g.,
flowrate,
UVT),
the
reactors
may
also
be
validated
for
a
wide
range
of
target
doses,
thereby
allowing
reactor
operation
to
be
tailored
to
achieve
different
levels
of
pathogen
inactivation
credit
at
different
WTPs.
The
test
conditions
and
target
doses
can
allow
interpolation
of
the
validation
data
to
conditions
of
flowrate,
UVT,
and
lamp
output
specific
for
application
to
various
WTP
applications.
Section
C.
4.9.3
describes
interpolation
of
validation
results
as
a
function
of
those
variables.

Utilities
installing
a
pre­
validated
UV
reactor
should
ensure
that
validation
conditions
are
appropriate
for
their
plant
operations
and
the
quality
of
testing
is
acceptable
to
their
State.
At
a
minimum,
the
following
hydraulic
and
operating
test
conditions
impact
the
application
of
prevalidated
UV
reactors:

 
UV
reactor
inlet
and
outlet
configurations
 
Flowrate
 
UVT
Validating
on­
site
at
the
WTP
is
not
trivial
and
should
be
regarded
as
a
relatively
complex
experimental
procedure.
Utilities
conducting
on­
site
validation
should
consider
the
following
issues
(
section
C.
3.1
provides
further
details):

 
Obtaining
water
with
a
sufficiently
high
UVT
to
allow
validation
over
the
entire
UVT
range
expected
at
the
WTP
Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
6
June
2003
 
Adequate
facilities
to
culture
the
challenge
microorganism
to
the
necessary
levels
to
demonstrate
the
desired
inactivation
 
Adequate
facilities
and
chemicals
to
adjust
UVT
to
the
range
expected
during
fullscale
operation
 
Providing
sufficient
mixing
of
additives
prior
to
entering
the
UV
reactor
and
mixing
of
the
challenge
microorganisms
after
the
reactor
 
Obtaining
permits
for
the
disposal
of
water
used
for
validation
 
Verification
of
the
behavior
of
UV
intensity
sensors
used
during
validation
(
sections
C.
3.2
and
C.
4.7)

 
Testing
with
inlet
and
outlet
conditions
representative
of
those
conditions
used
at
the
WTP
(
issue
for
off­
site
validation)

UV
reactors
previously
validated
under
existing
protocols
may
receive
inactivation
credit
if
the
validation
used
the
appropriate
challenge
microorganism(
s)
and
test
conditions
met
the
needs
of
the
operating
conditions
at
the
WTP.
Both
the
Austrian
Standard
ONORM
M
5873­
1
and
German
Guideline
DVGW
W294
require
an
RED
of
40
mJ/
cm2,
using
a
microorganism
more
representative
of
Cryptosporidium
(
B.
subtilis)
than
that
used
to
develop
Tier
1
criteria
(
MS2
phage).
Based
on
criteria
in
this
document,
UV
reactors
validated
with
those
protocols
should
be
granted
3
log
Cryptosporidium
and
Giardia
inactivation
credit.
Validation
by
NWRI/
AwwaRF
Guidelines
and
NSF
Standard
55
should
be
evaluated
on
a
case­
by­
case
basis
as
indicated
in
Appendix
C.

4.2.3
Third­
Party
Oversight
Third­
party
oversight
is
recommended
to
ensure
that
validation
testing
and
data
analyses
are
conducted
in
a
technically­
sound
manner
and
without
bias.
The
validation
testing
should
be
overseen
by
a
registered
professional
engineer,
independent
of
the
UV
manufacturer,
with
experience
in
testing
and
evaluating
UV
reactors.
Furthermore,
expert
opinion
should
be
sought
from
additional
parties
in
areas
of
UV
validation
where
the
engineer
has
limited
experience.
These
areas
can
include,
but
are
not
limited
to,
lamp
physics,
optics,
hydraulics,
microbiology,
and
electronics.

4.3
Considerations
for
Validation
Testing
This
section
highlights
the
key
factors
that
should
be
considered
in
the
early
planning
stages
of
UV
reactor
validation.

Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
7
June
2003
4.3.1
Inlet
and
Outlet
Hydraulics
The
inlet
and
outlet
configurations
of
the
validation
location
should
produce
conditions
that
result
in
equal
or
worse
dose
delivery
than
those
that
will
be
obtained
at
the
WTP.
Sections
3.1.4.3
and
C.
3.1.5
provide
recommended
approaches
to
ensure
such
hydraulic
conditions.

Computation
fluid
dynamics
(
CFD)­
based
dose
modeling
can
also
be
used
in
conjunction
with
any
approach
to
conservatively
address
reactor
hydraulics
during
testing.
However,
due
to
uncertainty
in
the
CFD
predictions,
the
predicted
dose
delivery
during
validation
should
be
at
least
20
percent
greater
than
the
dose
delivery
predictions
at
the
WTP.

4.3.2
UV
Equipment
This
section
discusses
the
following
UV
equipment
related
issues:
documentation,
monitoring
control
strategies,
UV
intensity
sensors,
and
lamp
aging
effects.

4.3.2.1
UV
Reactor
Documentation
In
the
weeks
prior
to
testing,
the
UV
manufacturer
should
provide
documentation
identifying
and
describing
the
UV
reactor
to
the
testing
organization
(
or
to
third­
party
oversight
if
the
manufacturer
is
conducting
the
testing
with
their
facilities).
This
documentation
should
include
all
reactor
and
component
information
relating
to
dose
delivery
and
monitoring,
such
as
technical
descriptions
of
all
internal
components,
lamp
and
sleeve
specifications,
UV
intensity
sensor
and
sensor
port
information.
See
section
C.
2.2
for
a
complete
list
and
discussion
of
the
documentation
requirements.

4.3.2.2
Control
Strategies
The
UV
reactor's
control
strategy
for
monitoring
dose
delivery
affects
the
selection
of
test
conditions
(
i.
e.,
flowrate,
UV
intensity,
and
UVT).
At
present,
three
strategies
are
commonly
used
for
monitoring
UV
dose
delivery.
Sections
C.
4.9.4.1
to
C.
4.9.4.3
describe
these
strategies
in
detail
and
recommend
validation
conditions
for
each.
(
Sections
3.1.5
and
5.5
also
describe
these
strategies
with
relation
to
design
and
operation,
respectively).

 
UV
intensity
setpoint
 
relies
on
UV
intensity
measurements
(
i.
e.,
UV
intensity
sensors)
and
flowrate
to
confirm
dose
delivery.
The
system
is
in
compliance
when
the
measured
intensity
value
is
greater
than
the
setpoint
at
that
flowrate.

 
UV
intensity/
UVT
setpoint
 
relies
on
the
UVT
as
well
as
the
UV
intensity
and
flowrate
to
determine
dose
delivery.
The
system
is
in
compliance
when
both
the
UV
intensity
and
UVT
are
greater
than
the
preset
setpoint
values.

 
Calculated
dose
 
relies
on
calculated
dose
delivery
from
UV
intensity,
UVT
(
in
some
cases),
lamp
power
and
flowrate
using
an
algorithm
provided
by
the
UV
reactor
manufacturer.
Typically,
this
method
is
tested
over
a
range
of
Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
8
June
2003
combinations
of
flow,
UVT,
and
lamp
power
to
determine
the
UV
dose
and
validate
the
algorithm.

4.3.2.3
UV
Intensity
Sensor
Monitoring
of
the
UV
dose
is
achieved
through
the
use
of
on­
line
UV
intensity
sensors.
The
properties
of
both
on­
line
and
reference
sensors
should
be
measured
by
an
independent
laboratory
that
is
equipped
to
confirm
sensor
calibration
and
measure
the
sensor's
angular
and
spectral
response,
linearity
over
the
working
range,
and
temperature
response.
The
Tier
1
approach
specifies
criteria
for
sensor
placement
in
the
UV
reactor,
sensor
spectral
response,
and
measurement
uncertainty.

4.3.2.4
Lamp
Aging
Prior
to
the
initiation
of
validation
testing,
all
lamps
should
undergo
100
hours
of
burn­
in.
This
practice
improves
the
stability
of
lamp
output.
Additional
testing
may
also
be
performed,
if
requested,
in
order
to
assess
the
effects
of
lamp
age
on
dose
delivery.
With
time,
mediumpressure
(
MP)
UV
lamps
can
undergo
non­
uniform
aging
that
causes
spectral
shifts
in
output.
These
changes
can
have
an
impact
on
the
dose
delivery
registered
by
the
monitoring
systems.
Manufacturers
should
test
dose
delivery
of
new
and
aged
lamps
to
determine
if
the
aged
lamps
reduce
disinfection
performance.
If
so,
validation
should
be
conducted
using
both
new
and
aged
lamps.
(
Section
C.
4.8
describes
a
procedure
for
testing
new
versus
aged
lamps.)

4.3.3
Additives
Used
in
Validation
Testing
4.3.3.1
Challenge
Microorganism
UV
reactor
validations
should
be
performed
with
a
microorganism
with
the
following
characteristics:
inactivation
kinetics
closely
resembling
those
of
the
target
pathogen
and
the
ability
to
be
cultured
in
a
reproducible
manner
to
high
concentrations.
Currently,
research
has
not
identified
such
a
microorganism
that
is
ideal
for
Cryptosporidium.
Challenge
microorganisms
typically
used
include
MS2
phage
and
Bacillus
subtilis,
both
of
which
are
significantly
more
resistant
to
UV
than
Cryptosporidium.

The
RED
bias,
an
important
component
of
the
safety
factor,
is
due
to
the
differences
in
inactivation
kinetics
between
the
challenge
microorganism
and
the
target
pathogen.
Under
the
Tier
1,
the
RED
bias
is
based
on
MS2
phage
as
the
challenge
organism.
If
a
challenge
microorganism
is
identified
in
the
future
that
exhibits
a
dose­
response
similar
to
the
target
pathogen
(
e.
g.,
Cryptosporidium),
the
RED
bias
could
be
decreased.

Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
9
June
2003
4.3.3.2
UV­
Absorbing
Compound
During
validation,
the
UVT
can
be
lowered
through
the
addition
of
a
UV­
absorbing
compound
to
simulate
the
range
of
UVT
that
may
be
encountered
for
a
given
UV
application.
For
the
validation
of
MP
UV
systems,
the
absorbing
compound
should
have
a
UV
absorbance
spectrum
similar
to
the
water
being
treated
in
the
full­
scale
application.
However,
obtaining
an
exact
replica
is
usually
not
possible.
Coffee
and
lignin
sulfonate
are
commonly
used
UV
absorbing
compounds;
however,
sodium
thiosulfate
and
fluorescein
have
also
been
used
with
some
success.

The
polychromatic
bias,
a
component
of
the
safety
factor
for
only
MP
reactors,
is
determined
as
a
function
of
the
UV­
absorbing
compound.
The
Tier
1
approach
specifies
criteria
for
minimum
UVT
for
MP
reactors
using
UV­
absorbing
compounds
and
applies
a
correction
factor
based
on
validation
testing
performed
to­
date
with
various
UV
absorbing
compounds.

4.4
Validation
Testing
Validation
provides
an
assessment
of
UV
reactor
dose
delivery
and
monitoring
under
specific
conditions
of
flowrate,
UVT,
and
lamp
output.
This
section
briefly
discusses
the
steps
involved
in
conducting
a
validation
test
and
provides
references
to
more
detailed
procedures
in
Appendices
C
and
D.

4.4.1
Microorganism
Preparation
Challenge
microorganisms
should
be
prepared
in
accordance
with
peer­
reviewed
methods.
All
information
regarding
the
source
of
the
host,
media
descriptions,
and
preparation
steps
should
be
documented.
It
is
expected
that
the
microorganism
stock
will
be
prepared
by
laboratory
personnel
familiar
with
methodologies
designed
to
prevent
microbial
stock
contamination.
The
use
of
these
same
techniques
in
the
field
during
validation
is
critical
and
any
personnel
participating
in
the
validation
should
be
familiar
with
them
to
avoid
sample
contamination.

Preparation
methods
for
the
two
most
common
challenge
microorganisms,
MS2
phage
and
B.
subtilis
spores,
are
provided
in
Appendix
D.
Note,
the
same
batch
of
challenge
organisms
should
be
used
for
both
collimated
beam
and
biodosimetry
testing,
as
described
below.

4.4.2
Collimated
Beam
Testing
The
collimated
beam
data
are
used
to
develop
the
dose­
response
curve
for
the
challenge
microorganism.
A
collimated
beam
apparatus
typically
consists
of
an
enclosed
low­
pressure
UV
lamp
and
a
tube
with
a
non­
reflective
inner
surface
(
see
Figure
4.2).
A
sample
of
the
challenge
microorganism
(
preferably
taken
from
the
influent
to
the
biodosimetry
test
stand)
is
placed
in
a
petri
dish
and
exposed
to
the
UV
light
for
a
predetermined
amount
of
time.
The
UV
dose
is
Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
10
June
2003
calculated
using
the
intensity
of
the
incident
UV
light,
UV
absorbance
of
the
water,
and
exposure
time.
Appendix
E
provides
a
complete
description
of
collimated
beam
testing.

At
least
two
water
quality
conditions
should
be
tested 
one
with
the
highest
UVT
(
no
absorbing
chemical
added)
and
a
second
with
the
lowest
UVT
used
in
the
biodosimetry
test.
UV
doses
should
be
selected
to
target
microorganism
inactivations
of
approximately
0.5,
1.0,
2.0,
3.0,
4.0,
and
5.0
log.

Figure
4.2
Collimated
Beam
Test
Apparatus
Low­
Pressure
Mercury
Arc
Lamp
Lamp
Enclosure
Petri
Dish
Containing
Microbial
Suspension
Magnetic
Stirrer
UV
Light
@
254
nm
Collimating
Tube
4.4.3
Biodosimetry
of
Full­
Scale
Reactors
The
biodosimetry
test
is
used
to
determine
the
inactivation
of
the
challenge
microorganism
by
the
UV
reactor
under
continuous­
flow
test
conditions.
Figure
4.3
provides
a
schematic
of
the
components
used
in
a
typical
biodosimetry
test.
Section
C.
3.1
describes
the
key
features.

Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
11
June
2003
Figure
4.3
Biodosimetry
Test
Components
he
following
facilities
are
typically
required
in
biodosimetry
testing:

Injection
of
the
challenge
microorganism
and
UV
absorbing
compound
 
ixing
of
the
added
compounds
upstream
and
downstream
of
the
reactor
before
 
low
measurement
 
ressure
measurement
upstream
and
downstream
of
the
reactor
 
ample
collection
before
and
after
the
UV
reactor
Proper
facilities
should
be
provided,
along
with
appropriate
permits,
to
discharge
the
treated
the
ent
and
p
detailed
description
of
sampling
requirements
is
provided
in
Appendix
C.

4.5
Data
Analysis
esults
from
the
collimated
beam
testing,
biodosimetry
testing,
and
uncertainties
associa
ved
by
.
Developing
a
UV
dose­
response
curve
for
the
challenge
microorganism
from
the
Challenge
Microbe
Backflow
Prevention
Static
Mixer
Static
Mixer
Flow
meter
UV
Reactor
Valve
To
Waste
Effluent
Sample
Port
Influent
Sample
Port
Valve
Water
Supply
Inlet
Piping
UV
Absorber
Outlet
Piping
Pressure
Gage
Pressure
Gage
Influent
Quenching
Agent
T
 

M
sampling
F
P
S
water.
The
testing
should
be
conducted
after
steady­
state
conditions
are
achieved
for
desired
matrix
of
experimental
conditions
evaluating
variations
in
challenge
organism
concentration,
flowrate
UVT,
and
lamp
power/
output.
Samples
collected
from
the
influ
effluent
sample
ports
are
used
to
determine
the
inactivation
achieved
for
the
specific
reactor
condition
being
tested.
Operational
parameters,
such
as
UV
intensity,
flowrate,
UVT,
and
lam
power,
are
measured
during
the
test.

A
R
ted
with
equipment
and
data
are
used
to
determine
the
log
inactivation
credit
achie
the
UV
reactor.
Data
analysis
consists
of
four
steps:

1
collimated
beam
test
Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
12
June
2003
2.
Calculating
log
inactivation
from
the
biodosimetry
test
.
Determining
the
RED(
s)
from
the
results
of
steps
1
and
2
4.
Applying
safety
factors
to
determine
log
inactivation
credit
(
Tier
1
or
Tier
2
he
following
sections
describe
these
steps.
References
to
the
appropriate
sections
in
Append
4.5.1
eveloping
Challenge
Microorganism
Dose­
Response
Curve
ose­
response
curves
should
initially
be
generated
separately
for
each
collimated
beam
test
con
nt
er
he
following
sub­
sections
describe
how
to
calculate
the
log
inactivation
from
collimated
beam
t
.5.1.1
Calculate
Dose­
Response
Data
From
Collimated
Beam
Testing
he
log
inactivation
for
each
dose
delivered
by
the
collimated
beam
should
be
calculated
using
E
3
approach)

Ti
x
C
are
provided
for
further
details
and
examples.

D
Dd
ition
(
a
minimum
of
two
conditions
 
lowest
and
highest
UVT
 
is
recommended).
The
curves
should
predict
similar
dose­
response
relationships,
as
indicated
by
statistical
analyses.
If
statistically
similar,
the
data
can
be
combined
and
one
curve
generated
for
the
entire
dataset.
If
the
curves
are
statistically
different,
the
cause
of
the
difference
should
be
determined,
and
the
test
should
either
be
redone
or
the
different
dose­
response
curves
should
be
used
for
the
differe
test
conditions.
Differences
in
UV
dose­
response
could
occur
if
the
dose­
response
were
determined
with
different
batches
of
the
challenge
microorganism
or
if
coagulation
or
oth
water
quality
interferences
impacted
the
dose­
response.

T
est
data
and
generate
a
dose­
response
curve.
Section
C.
4.9.7.2
discusses
the
statistical
analysis
for
comparing
curves
and
combining
data.

4
Tq
uation
4.2:

  
 

  
 
=
N
N
on
Inactivati
Log
0
log
Equation
4.2
here
=
Challenge
microorganism
concentration
in
an
aliquot
of
sample
o
dose
aliquots
w
N
N0
=
Average
concentration
of
the
challenge
microorganism
in
the
zer
Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
13
June
2003
4.5.1.2
Fitting
Dose­
Response
Data
to
a
Curve
The
following
steps
describe
how
to
develop
a
dose­
response
curve:

1)
Plot
log
inactivation
achieved
as
a
function
of
UV
dose
in
the
collimated
beam
test
2)
Use
regression
analysis
to
derive
an
equation
that
best
fits
the
data
 
For
first­
order
kinetics,
a
linear
equation
should
fit
best:

B
on
Inactivati
Log
A
Dose
+
×
=

 
For
dose­
responses
showing
tailing
effects,
a
quadratic
equation
should
fit
best:

(
)
2
on
Inactivati
Log
D
on
Inactivati
Log
C
Dose
×
+
×
=

 
For
dose­
response
showing
shoulder
effects,
other
polynomial
equations
should
be
used.

3)
Evaluate
fit
of
equation
 
Equation
coefficients
should
be
significant
at
a
95
percent
confidence
level
(
section
C.
4.9.7.1
provides
an
example
that
uses
p­
statistics
to
evaluate
the
coefficients).

 
Confidence
intervals
for
the
fit
should
be
determined
at
an
80
percent
confidence
level.
(
The
Tier
1
approach
specifies
criteria
the
confidence
intervals
must
meet
and
the
Tier
2
approach
includes
an
uncertainty
term
for
the
confidence
intervals
in
the
safety
factor
calculation.)

 
The
differences
between
the
predicted
dose
and
measured
dose
at
a
given
log
inactivation
should
be
randomly
distributed
around
zero
and
not
dependent
on
dose.
In
other
words,
the
data
points
should
be
randomly
distributed
above
and
below
the
curve
(
section
C.
4.9.7.1
provides
an
example
of
this
evaluation).

4.5.2
Determining
Log
Inactivation
from
Biodosimetry
Testing
At
each
test
condition
 
flowrate,
UVT,
and
lamp
output
 
the
arithmetic
mean
and
standard
deviation
of
the
log
influent
and
effluent
challenge
microorganism
concentrations
should
be
calculated.
From
the
mean
concentrations,
log
inactivation
should
calculated
using
the
following
equation:

Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
14
June
2003
(
)
(

E
I
N
N
on
Inactivati
Log
log
log
 
=
)
Equation
4.3
where
log(
NI)
=
Mean
challenge
microorganism
log
concentration
of
the
reactor
influent
samples
log(
NE)
=
Mean
challenge
microorganism
log
concentration
of
the
reactor
effluent
samples
The
standard
deviation
is
used
in
the
safety
factor
calculation
for
Tier
2,
while
Tier
1
specifies
a
limit
for
the
standard
deviation.

4.5.3
Determining
the
RED
This
section
describes
how
to
calculate
RED
values
for
all
test
conditions
and
select
the
appropriate
RED
for
subsequent
log
inactivation
credit
determination.

4.5.3.1
Calculating
the
RED
Values
The
RED
is
calculated
by
inputting
the
biodosimetry
log
inactivation
values
for
each
test
condition
into
the
equation
describing
the
UV
dose­
response
curve
of
the
challenge
microorganism.

Example.
For
0.5
MGD
flow,
80
percent
UVT,
and
lamp
output
of
70
percent,
the
inactivation
calculated
from
Equation
4.3
was
4.0
log.
The
UV
dose­
response
equation
was
best
fit
with
the
equation:

0
.
6
5
.
15
 
×
=
on
Inactivati
Log
Dose
Inputting
4.0
log
into
the
above
equation
results
in
an
RED
of
56
mJ/
cm2.
This
calculation
should
be
repeated
for
each
test
condition
(
i.
e.,
flowrate,
UVT,
and
lamp
output
combination).

4.5.3.2
Selecting
the
Appropriate
RED
for
Log
Inactivation
Credit
Determination
Since
the
biodosimetry
test
is
conducted
at
various
flowrates,
UVT,
and
lamp
output
combinations,
the
validation
results
will
have
more
than
one
RED
value
for
each
setpoint.
Choosing
the
appropriate
RED
to
determine
log
inactivation
credit
depends
first
on
the
monitoring
approach
used
to
indicate
dose
delivery.
The
following
three
approaches
are
considered
in
this
text:

 
UV
intensity
setpoint
approach
­
the
UV
reactor
should
be
rated
at
the
lowest
inactivation
observed
for
each
set
point
condition
tested.

 
UV
intensity
and
UVT
setpoint
approach
­
the
UV
reactor
should
be
rated
at
the
inactivation
observed
with
UV
reactor
operation
under
setpoint
conditions.

Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
15
June
2003
 
Calculated
dose
approach
­
the
UV
reactor
should
be
rated
at
the
lowest
inactivation
observed
for
each
calculated
dose
setpoint
evaluated.

Section
C.
4.9.4
recommends
validation
conditions
for
each
of
the
above
approaches.
Section
C.
5
provides
examples
of
interpreting
validation
results
for
the
different
approaches.

4.5.3.3
Interpolating
RED
as
a
Function
of
Test
Conditions
The
RED
measured
during
validation
testing
can
be
interpolated
as
a
function
of
inverse
flowrate,
UVT,
or
UV
intensity
by
fitting
an
equation
to
the
data
being
interpolated
(
e.
g.,
RED
as
a
function
of
inverse
flowrate).
The
equation
should
not
be
used
for
extrapolation
(
i.
e.,
projecting
RED
outside
the
range
of
tested
conditions).
The
following
provides
guidelines
for
interpolation:

 
The
equation
should
pass
through
the
origin
(
0,0)
if
the
RED
is
interpolated
as
a
function
of
measured
intensity
or
inverse
flowrate
 
The
equation
coefficients
should
be
significant
at
a
95
percent
confidence
level
 
The
differences
between
the
values
measured
and
predicted
by
the
equation
should
be
randomly
distributed
around
zero
 
An
80
percent
confidence
interval
should
be
used
to
determine
the
uncertainty
of
the
equation
used
to
interpolate
the
RED
values.
For
Tier
1,
the
uncertainty
of
the
interpolation
should
be
10
percent
or
less
at
an
80
percent
confidence
level.
For
Tier
2
it
should
be
included
as
an
uncertainty
term
in
the
safety
factor
calculation
as
described
in
section
C.
4.10.2.3.

4.5.4
Determining
Inactivation
Credit
As
discussed
in
the
introduction
to
this
chapter,
there
are
two
approaches
described
for
determining
log
inactivation.

 
Tier
1
­
pre­
determined
safety
factor.

 
Tier
2
­
calculated
safety
factor
from
the
following
dose
delivery
monitoring
and
validation
bias
and
uncertainties:

­
RED
bias
­
Polychromatic
bias
(
for
MP
reactors)
­
Measured
RED
­
Interpolation
of
RED
as
a
function
of
flowrate,
UVT,
or
UV
intensity
­
Sensors
used
during
validation
(
UV
intensity,
UVT)
­
On­
line
and
reference
sensors
used
at
WTP
(
UV
intensity,
UVT)
­
Lamp
output
quantification
Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
16
June
2003
The
remainder
of
this
chapter
describes
how
to
determine
the
log
inactivation
credit
achieved
using
the
Tier
1
approach
and
the
criteria
that
should
be
met
in
order
to
use
this
approach.
Appendix
C
contains
a
detailed
description
of
the
basis
the
Tier
2
approach.

Tier
1
Log
Inactivation
Credit
Tables
4.1
and
4.2
present
the
RED
target
values
for
UV
reactors
using
LP/
LPHO
and
MP
lamps,
respectively.
The
values
in
these
tables
are
derived
by
multiplying
the
required
dose
values
by
the
Tier
1
safety
factors
(
see
Appendix
C
for
details).
The
values
in
Table
4.2
(
MP)
are
higher
than
in
Table
4.1
(
LP/
LPHO)
because
they
include
the
polychromatic
bias,
which
is
not
a
factor
in
monochromatic
(
LP/
LPHO)
reactors.

For
a
given
pathogen
and
level
of
log
inactivation
credit,
the
RED
measured
during
validation
should
be
greater
than
or
equal
to
the
corresponding
RED
target
listed
in
the
table.
Note,
validation
testing
with
multiple
setpoints
may
result
in
different
log
inactivation
credits
for
the
different
setpoints.

Example.
Using
an
LP
reactor
and
meeting
the
Tier
1
validation
criteria
(
see
section
4.6),
the
lowest
RED
measured
for
the
challenge
microorganism
during
validation
was
29
mJ/
cm2.
Consequently,
the
log
inactivation
credits
achieved
are
2.5
for
Cryptosporidium
and
2.5
for
Giarida.
No
inactivation
credit
is
achieved
for
viruses.

Table
4.1
Tier
1
RED
Targets
for
UV
Reactors
with
LP
or
LPHO
Lamps
RED
Target
(
mJ/
cm2)
Log
Inactivation
Credit
Cryptosporidium
Giardia
Virus
0.5
6.8
6.6
55
1.0
11
9.7
81
1.5
15
13
110
2.0
21
20
139
2.5
28
26
169
3.0
36
34
199
3.5
­
­
227
4.0
­
­
259
Table
4.2
Tier
1
RED
Targets
for
UV
Reactors
with
MP
Lamps
RED
Target
(
mJ/
cm2)
Log
Inactivation
Credit
Cryptosporidium
Giardia
Virus
0.5
7.7
7.5
63
1.0
12
11
94
1.5
17
15
128
2.0
24
23
161
2.5
32
30
195
3.0
42
40
231
3.5
­
­
263
4.0
­
­
300
Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
17
June
2003
4.6
Tier
1
Criteria
The
safety
factors
derived
for
the
Tier
1
approach
are
based
on
assumed
uncertainties
and
corrections
for
given
experimental
methods.
For
these
assumptions
to
be
practical,
and
thus
the
use
of
Tier
1
numbers
appropriate,
the
validation
conditions
should
meet
the
criteria
specified
in
this
section.
Note,
the
equipment
criteria
should
be
provided
by
the
UV
manufacturer
and
reviewed
by
a
third­
party
for
verification.

4.6.1
UV
Intensity
Sensors
 
UV
reactors
with
MP
lamps
should
be
equipped
with
one
sensor
per
lamp.
UV
reactors
with
LP
or
LPHO
lamps
should
be
equipped
with
at
least
one
sensor
per
bank
of
lamps.

 
UV
intensity
sensors
should
view
a
point
along
the
length
of
the
lamp
that
is
between
the
electrode
(
lamp
end)
and
within
25
percent
of
the
arc
length
away
from
the
electrode.

 
UV
intensity
sensors
should
have
a
spectral
response
that
peaks
between
250
and
280
nm.
When
mounted
on
the
UV
reactor
and
viewing
the
lamps
through
water,
the
measurement
of
UV
light
greater
than
300
nm
made
by
the
sensor
should
be
less
than
10
percent
of
the
total
measurement
made
by
the
sensor.
Conformance
to
these
criteria
can
be
demonstrated
using
UV
intensity
field
modeling.
Figure
4.4
presents
examples
of
two
sensors
where
both
have
the
appropriate
peaks,
but
one
has
too
much
UV
light
in
the
>
300
nm
range.

 
The
UV
intensity
sensors
used
during
validation
and
the
duty
and
reference
sensors
used
during
operation
of
the
UV
reactor
at
the
WTP
should
provide
National
Institute
of
Standards
and
Technology
(
NIST)­
traceable
measurements
with
an
uncertainty
of
±
15
percent
or
less
at
an
80
percent
confidence
level.

 
During
operation
of
the
UV
reactor
at
the
WTP,
measurements
made
by
the
duty
UV
intensity
sensor
should
be
checked
using
a
reference
UV
intensity
sensor.
If
the
duty
sensor
reads
higher
than
the
reference
sensor
(
i.
e.,
overestimating
dose
delivery),
or
substantially
lower,
it
should
be
recalibrated.
For
a
recommended
control
standard,
the
duty
sensor
should
not
read
less
than
the
reference
by
the
following
amount:

(
)
2
1
2
Duty
2
Ref
Ref
uty
D
100
1
I
I
 
+
 
 
×
 

  
 

 

  
 
 
Equation
4.4
where
IRef
=
Intensity
measured
by
the
reference
sensor
IDuty
=
Intensity
measured
by
the
duty
sensor
 Ref
=
Measurement
uncertainty
of
the
reference
sensor
(%)
 Duty
=
Measurement
uncertainty
of
the
duty
sensor
(%)

Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
18
June
2003
 
If
the
dose­
monitoring
strategy
uses
an
on­
line
UVT
monitor,
the
UV
absorbance
at
254
nm
(
A254)
calculated
from
the
measured
UVT
should
have
an
uncertainty
of
±
10
percent
or
less
at
an
80
percent
confidence
level.

Figure
4.4
Examples
of
UV
Intensity
Sensor
Spectral
Response
Ranges
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
200
250
300
350
400
Wavelength
(
nm)
Spectral
Response
Relative
to
254
nm
Unfiltered
Sensor
Filtered
Sensor
250
280
Filtered
Sensor.
Detected
UV
light
with
a
0
cm
sensor­
to­
lamp
water
layer.
Detected
UV
>
300
nm
is
0.7%
of
total
UV
light
detected.
Unfiltered
Sensor.
Detected
UV
light
with
a
0
cm
sensor­
to­
lamp
water
layer.
Detected
UV
>
300
nm
is
41%
of
total
UV
light
detected.
0
5
10
15
20
25
200
250
300
350
400
Wavelength
(
nm)
Detected
Irradiance
0
10
20
30
40
50
200
250
300
350
400
Wavelength
(
nm)
Detected
Irradiance
Filtered
Sensor.
Detected
UV
light
with
a
20
cm
sensor­
to­
lamp
water
layer.
Detected
UV
>
300
nm
is
5%
of
total
UV
light
detected.
Unfiltered
Sensor.
Detected
UV
light
with
a
20
cm
sensor­
to­
lamp
water
layer.
Detected
UV
>
300
nm
is
85%
of
total
UV
light
detected.
0.0000
0.0005
0.0010
0.0015
0.0020
0.0025
200
250
300
350
400
Wavelength
(
nm)
Detected
Irradiance
0.0000
0.0100
0.0200
0.0300
0.0400
200
250
300
350
400
Wavelength
(
nm)
Detected
Irradiance
Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
19
June
2003
4.6.2
UV
Lamp
Output
 
The
standard
deviation
of
the
UV
output
of
LP
or
LPHO
lamps
should
be
15
percent
or
less
of
the
mean
output.
The
standard
deviation
should
be
determined
using
either
life
test
or
field
test
data
on
aged
lamps.

4.6.3
Flow
Measurements
 
The
flow
measurements
made
during
validation
and
during
operation
of
the
UV
reactor
at
the
WTP
should
have
an
uncertainty
of
±
5
percent
or
less
at
an
80
percent
confidence
level.

4.6.4
Collimated
Beam
Apparatus
 
The
calculated
dose
delivered
by
the
collimated
beam
apparatus
should
have
a
measurement
uncertainty
of
±
15
percent
or
less
at
an
80
percent
confidence
level.

4.6.5
Challenge
Microorganism
Dose­
Response
 
Over
the
range
of
doses
within
one
log
unit
of
the
log
inactivation
demonstrated
during
validation,
the
UV
sensitivity
of
the
challenge
microorganism
should
be
less
than
or
equal
to
25
mJ/
cm2
per
log
inactivation
(
the
dose­
response
of
a
resistant
strain
of
MS2).
For
example,
if
the
challenge
microorganism
log
inactivation
measured
by
the
UV
reactor
ranges
between
1.5
and
3.5
log,
the
dose­
response
of
the
challenge
microorganism
should
be
less
than
or
equal
to
25
mJ/
cm2
per
log
inactivation
between
0.5
and
4.5
log
inactivation.

 
If
the
dose­
response
of
the
challenge
microorganism
has
a
shoulder,
that
shoulder
should
not
occur
over
a
dose
range
greater
than
50
percent
of
the
RED
demonstrated
during
validation.
The
shoulder
is
defined
by
extrapolating
the
exponential
reduction
region
of
the
dose­
response
curve
to
the
dose­
axis
(
see
Figure
4.5).

 
If
the
dose­
response
demonstrates
tailing,
the
tailing
should
not
occur
until
one
log
reduction
greater
than
the
log
reduction
demonstrated
during
validation.

 
A
plot
of
dose
versus
log
inactivation
for
the
collimated
beam
test
should
have
an
80
percent
confidence
interval
of
10
percent
or
less
at
the
log
inactivation
demonstrated
by
the
UV
reactor.

Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
20
June
2003
Figure
4.5
Dose­
Response
With
a
Shoulder
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
0
25
50
75
UV
Dose
(
mJ/
cm2)
Log
Inactivation
100
This
range
should
be
less
than
1/
2
the
RED
demonstrated
during
validation.
In
this
case,
the
RED
should
be
>
20
mJ/
cm2.

4.6.6
Medium
Pressure
Lamps
 
During
validation,
the
UVT
of
the
water
at
254
nm
should
be
greater
than
the
values
specified
in
Figure
4.6
for
a
given
sensor­
to­
lamp
water
layer
and
UV­
absorbing
chemical.
The
sensor­
to­
lamp
water
layer
is
defined
as
the
distance
traveled
through
water
by
UV
light
passing
from
the
lamp
to
the
sensor.
The
values
in
Figure
4.6
were
taken
from
Figure
C.
7
of
Appendix
C
for
a
polychromatic
bias
of
1.2.

Proposal
Draft
4.
Overview
of
Validation
Testing
UV
Disinfection
Guidance
Manual
4­
21
June
2003
Figure
4.6
Criteria
for
the
Minimum
UVT
of
MP
Reactors
under
Tier
1
75
80
85
90
95
100
0
5
10
15
20
25
30
Sensor­
to­
lamp
wate
r
layer
(
cm)
Mi
nimum
UVT
at
254
nm
(%)

Coffee
Lignin
Sulphonate
NOM
4.6.7
Biodosimetry
Sampling
 
Five
influent
and
five
effluent
samples
should
be
collected
for
each
test
condition
and
evaluated
as
described
in
section
C.
4.9.5.

 
The
standard
deviation
of
the
challenge
microorganism
concentration
measured
with
the
influent
and
effluent
samples
should
be
less
than
or
equal
to
0.20
log
units.

Proposal
Draft
5.
Start­
Up
and
Operation
of
UV
Installations
This
chapter
describes
the
start­
up
activities
and
routine
operational
issues
associated
with
a
UV
disinfection
facility.
The
start­
up
discussion
focuses
on
the
functional
and
performance
testing
that
should
be
conducted
during
the
start­
up
process.
The
remainder
of
the
chapter
describes
the
requirements
and
recommendations
for
operation,
maintenance,
monitoring,
and
reporting
for
UV
installations.
The
organization
of
this
chapter
is
presented
below
by
the
key
question
each
section
addresses.

 
What
is
included
in
final
UV
installation
inspection?
................................
Section
5.1.1
 
What
testing
should
be
completed
during
start­
up?..................
Sections
5.1.2
and
5.1.3
 
What
items
should
be
included
in
the
operations
and
maintenance
manual?..................................................................................
Section
5.1.4
 
What
are
the
operational
requirements
and
recommended
tasks?
........................................................................................
Sections
5.2.1
and
5.2.2
 
What
are
the
routine
start­
up
and
shutdown
procedures?
..........................
Section
5.2.3
 
What
maintenance
tasks
are
recommended?
.............................................
Section
5.3.1
 
What
spare
parts
are
recommended
to
be
kept
on
hand?............................
Section
5.3.3
 
What
monitoring
is
required
for
regulatory
compliance?...........................
Section
5.4.1
 
What
additional
monitoring
is
recommended?
..........................................
Section
5.4.2
 
What
should
be
reported
to
the
State?
.......................................................
Section
5.4.3
 
How
do
you
determine
the
operational
requirements
from
validation
testing?
........................................................................................
Section
5.5
 
What
should
be
done
if
there
is:
..................................................................
Section
5.6
­
Low
UV
intensity?
­
High
UV
absorbance?
­
Rapid
flow
increase/
high
flow?
­
Unreliable
UV
intensity
sensor
readings?
­
Power
loss?

 
What
staffing
issues
are
associated
with
operation,
maintenance,
and
monitoring
of
UV
installations?
......................................
Section
5.7
Given
the
wide
range
of
UV
installations
and
UV
reactors
available,
this
document
cannot
address
or
anticipate
all
scenarios.
The
guidelines
provided
in
this
manual
are
a
compilation
of
industry
experience
and
manufacturers'
recommendations.
Therefore,
they
may
UV
Disinfection
Guidance
Manual
5­
1
June
2003
Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
2
June
2003
differ
from
those
provided
by
specific
manufacturers
for
their
equipment.
In
these
situations,
the
manufacturer's
standards
should
be
followed.

The
general
process
to
be
followed
for
the
start­
up
and
routine
operation
of
a
UV
installation
is
shown
in
Figure
5.1.
A
detailed
description
of
each
activity
is
given
in
the
remainder
of
this
chapter.

Figure
5.1
Start­
up
and
Operation
Flowchart
Routine
Operations
Sections
5.2.1
to
5.2.4
Report
to
State
Section
5.4.3
Completion
of
Construction
and
Inspection
Final
Inspection
Section
5.1.1
Maintenance
Section
5.3
Monitoring
and
Recording
Sections
5.4.1
to
5.4.2
Determination
of
Operational
Requirements
Section
5.5
Appendix
C
Staffing
Issues
Section
5.7
Operational
Challenges
Section
5.6
Routine
Operation
Start­
Up
Activities
Functional
Testing
Section
5.1.2
Performance
Testing
Section
5.1.3
Operations
and
Maintenance
Development
Section
5.1.4
Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
3
June
2003
5.1
Start­
up
of
UV
Installation
For
the
purposes
of
this
manual,
the
start­
up
of
the
UV
installation
is
considered
as
the
transition
from
the
construction
phase
to
the
operation
phase.
Start­
up
activities
include
final
inspection
of
the
UV
reactors
and
ancillary
equipment,
functional
testing,
performance
testing,
operations
and
maintenance
(
O&
M)
manual
development.
Functional
testing
confirms
the
mechanical,
instrumentation
and
controls,
and
hydraulic
conditions
of
the
UV
installation
to
ensure
they
meet
the
requirements
of
the
contract
documents.
It
also
verifies
that
the
operational
conditions
are
consistent
with
the
validated
conditions.
Performance
testing
verifies
that
the
UV
reactors
are
operating
in
accordance
with
the
contract
documents.
In
addition,
an
O&
M
manual
should
be
developed
during
UV
installation
start­
up.

A
start­
up
plan
should
be
developed
in
collaboration
with
the
UV
installation
designer,
plant
operations
staff,
and
the
UV
manufacturer.
The
designer
will
be
most
familiar
with
the
layout
of
the
reactors,
piping,
and
how
to
integrate
the
UV
installation
with
the
other
treatment
processes.
The
operations
staff
will
be
able
to
identify
potential
impacts
on
routine
plant
operations.
The
manufacturer
will
be
most
familiar
with
operation
of
the
UV
reactors.
The
startup
plan
should
include
a
pre­
start
checklist,
a
procedure
for
checking
equipment
installation
and
calibration
(
functional
testing),
a
procedure
for
verifying
system
operation,
and
a
procedure
for
checking
alarm
settings
and
system
controls
(
performance
testing).

5.1.1
Final
Inspection
As
the
first
step
in
the
start­
up
process,
a
detailed
inspection
of
the
UV
installation
should
be
completed.
The
inspection
should
include
a
visual
assessment
to
ensure
that
all
components
meet
the
technical
specifications
and
that
the
UV
installation
was
completed
in
accordance
with
the
construction
documents.
The
configuration
of
the
piping
and
UV
reactors
should
meet
the
constraints
established
during
validation
testing
(
see
section
4.3.1).
If
on­
site
validation
will
be
performed,
the
availability
of
the
necessary
features
(
e.
g.,
feed
and
sample
ports,
mixing
systems,
drains)
should
be
confirmed.
In
addition,
leak
testing
should
be
performed,
and
then
all
UV
installation
components
and
associated
valves
and
piping
should
be
thoroughly
cleaned
and
disinfected
(
State
requirements
may
apply).

5.1.2
Functional
Testing
Functional
testing
consists
of
a
series
of
short
duration
tests
that
assess
the
ability
of
each
component
of
the
system
to
function
in
accordance
with
the
specifications
detailed
in
the
contract
documents.
Some
of
the
evaluations
are
conducted
by
monitoring
performance
during
normal
operations.
However,
the
majority
of
functional
testing
is
completed
through
simulations
of
specific
operating
conditions
and
monitoring
the
UV
reactor
operation
and
response.
Functional
testing
entails
flooding
and
energizing
the
UV
reactors
to
confirm
the
operation
of
the
following
items:

 
UV
lamps
and
UV
intensity
sensors
 
Operating
sequence
and
control
logic
for
the
reactor
Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
4
June
2003
 
Ancillary
equipment,
including
UV
transmittance
(
UVT)
monitors,
flowmeters,
and
control
valves
 
Electrical
system
components,
including
ballasts,
uninterruptible
or
standby
power
supplies,
and
the
ballast
cooling
system
It
is
strongly
recommended
that
the
UV
manufacturer
inspect
the
UV
installation
prior
to
energizing
the
UV
reactors
and
be
present
when
the
UV
reactors
are
first
energized.
Manufacturers
may
require
the
presence
of
one
of
their
representatives
during
these
activities
as
a
condition
of
their
equipment
warranty.

5.1.2.1
Verification
of
Mechanical
Operation
UV
reactors
may
incorporate
mechanical
elements
such
as
valves
and
on­
line
mechanical
cleaning
(
OMC).
During
functional
testing,
the
satisfactory
operation
of
these
mechanical
components
should
be
confirmed.
The
procedures
used
to
confirm
valve
operation
for
a
UV
installation
are
not
different
from
those
for
other
applications
that
use
valves
for
isolation
or
flow
control
and,
therefore,
are
not
described
here.
The
OMC
system,
if
provided,
should
be
checked
for
proper
operation.
Specifically,
the
following
items
should
be
verified:

 
Smooth
movement
of
the
wiper
with
no
jamming
or
binding
of
the
wiper
on
the
sleeve
 
Extension
of
wiper
stroke
to
the
full
length
of
the
sleeve
with
no
impact
at
the
end
of
travel
that
could
damage
or
break
the
sleeve
 
Proper
operation
of
the
wiper
drive
mechanism
and
motor
with
no
slipping
or
binding
5.1.2.2
Verification
of
Monitoring
Equipment
The
monitoring
equipment
is
important
for
UV
reactor
operation,
and
its
proper
operation
should
be
verified
during
functional
testing.

Flowmeter
Accurate
measurement
of
the
flow
is
essential
to
ensure
that
the
UV
reactors
are
operating
within
the
validated
conditions.
Not
all
utilities
will
install
dedicated
flowmeters.
For
those
facilities
that
rely
on
flow
measurement
using
an
existing,
common
flowmeter
(
e.
g.,
raw
water
flowmeter),
the
functionality
of
the
flowmeter
should
be
verified
in
conjunction
with
its
intended
use
with
the
UV
installation.
Specifically,
the
accuracy
and
operating
range
of
the
flowmeter
should
be
verified
and
the
availability
of
the
necessary
output
signals
from
the
meter
should
be
confirmed.
If
pressure
gauges
are
used
to
monitor
the
flow
split
between
UV
reactors,
the
calibration
and
installation
of
the
pressure
gauges
should
be
verified
as
well.

The
uncertainty
associated
with
the
existing
flowmeter
should
be
determined
to
ensure
that
the
appropriate
validation
constraints
were
used.
It
is
recommended
that
the
original
Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
5
June
2003
certification
of
calibration
be
reviewed
in
conjunction
with
the
equipment
specifications
to
establish
the
measurement
uncertainty
for
the
existing
flowmeter.
There
are
three
methods
to
verify
the
flowmeter
operation:
flow
verificators,
a
time­
discharge
test,
and
a
clamp­
on
flowmeter.
The
flow
verificators
assess
the
physical
condition
of
the
installed
equipment
relative
to
its
condition
at
the
time
of
factory
calibration
to
confirm
that
the
original
uncertainty
can
be
maintained.
For
example,
verification
of
a
magnetic
flowmeter
would
consist
of
an
insulation
test
of
the
entire
flowmeter
system
and
cable;
testing
of
the
sensor
magnetic
properties;
testing
of
signal
converter
gain,
linearity
and
zero
point;
testing
of
digital
output;
and
testing
of
analog
output.
A
time­
discharge
test
compares
the
flowrate
measured
by
the
flowmeter
against
the
value
calculated
by
measuring
the
volume
of
water
discharged
over
a
predetermined
amount
of
time
(
using
a
bucket,
clearwell,
or
tank
of
known
volume).
A
temporary
or
clamp­
on
flowmeter
can
be
used
to
assess
the
accuracy
of
the
existing
flowmeter.
It
is
important
to
consider
the
uncertainty
of
the
reference
flowmeter
when
using
this
approach.

If
a
new
flowmeter
is
used
to
measure
the
flow
through
the
reactor,
the
flowmeter
manufacturer
should
provide
a
certification
of
calibration
at
the
time
of
equipment
delivery.
It
is
also
recommended
that
the
manufacturer
inspect
the
UV
installation
and
confirm
that
it
was
completed
in
accordance
with
their
recommendations
to
ensure
the
certified
accuracy
of
the
flowmeter
is
achieved.
The
flowmeter
measurement
uncertainty
should
be
equal
to
or
better
than
that
used
during
validation.

On­
line
UVT
Monitor
An
on­
line
UVT
monitor
may
be
included
as
part
of
the
UV
reactor,
especially
if
a
UV
intensity
and
UVT
setpoint
or
calculated
dose
control
strategy
(
section
3.1.4.2)
is
used.
The
online
UVT
monitor
should
be
calibrated
and
its
operation
verified.
Calibration
can
be
completed
using
a
buffer
solution
of
known
UVT
and
may
be
operation
may
be
verified
by
collecting
and
analyzing
grab
samples,
using
a
bench
top
spectrophotometer.

5.1.2.3
Verification
of
Instrumentation
and
Control
Systems
The
amount
of
testing
needed
for
the
instrumentation
and
control
systems
is
proportional
to
the
complexity
of
the
control
strategy
that
is
used.
Testing
should
include
verification
of
monitoring
equipment
(
including
calibration
of
all
instruments),
tuning
of
control
loops,
checking
operation
functions,
and
verifying
all
final
control
actions.
As
described
below,
the
UV
reactors
should
be
run
through
a
series
of
simulations
that
represent
the
possible
operating
scenarios
in
order
to
confirm
that
the
appropriate
UV
reactor
response
occurs.
Typically,
the
packaged
UV
reactor
control
panel
contains
all
of
the
components
to
control
and
operate
the
UV
reactor.
The
panel
should
provide
the
operating
status,
diagnostic
information,
and
operator
interface
capability.
It
should
also
include
lamp
status
indicators
and
programmable
logic
controllers
(
PLC)
and
may
include
ballasts,
and
lamp
starters.
The
PLCs
are
typically
used
to
control
the
operation
of
a
UV
reactor
based
on
certain
input
signals.
A
manufacturer
representative
should
be
present
during
the
simulations
to
assist
in
troubleshooting
and
addressing
any
issues
that
may
result
from
the
packaged
UV
reactor
controls.

Simulations
should
be
used
to
confirm
the
operation
of
the
UV
reactors
and
the
operation
of
all
ancillary
equipment
and
instrumentation,
including
valves,
flowmeters,
and
UVT
monitors.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
6
June
2003
As
applicable,
specific
operating
conditions
that
should
be
simulated
include
the
following
conditions:

 
Cold
start
of
the
UV
reactors
 
Cool
down
and
restart
of
the
UV
reactors
 
Sequencing
of
the
UV
reactors
in
multiple
reactor
installations
 
Adjustment
of
lamp
intensity
in
response
to
varying
water
quality
or
flowrate
 
Shutdown
of
the
UV
reactors
 
Operation
of
the
UV
reactors
during
line
power
failure
(
when
backup
or
uninterruptible
power
supplies
(
UPS)
are
available)

 
Manual
override,
safety
interlocks,
and
report
generation
During
these
simulations,
the
utility
should
record
the
amount
of
off­
specification
time
and
discharge
volume
(
i.
e.,
operation
outside
of
validated
conditions)
associated
with
each
action.
This
is
necessary
to
assess
the
potential
effect
of
the
conditions
associated
with
these
actions
on
the
utility's
ability
to
meet
its
disinfection
goals
and
comply
with
the
State­
established
limitations
for
off­
specification
operation.
In
addition
to
simulating
possible
operating
conditions,
each
of
the
alarm
conditions
and
monitoring
functions
incorporated
in
the
design
should
be
verified.
Possible
monitoring
functions
and
alarm
conditions
are
discussed
in
section
3.3.3.8
and
may
include
the
following
conditions:

 
Low
UV
dose
and
UV
intensity
 
Low
UVT
 
Low
and
high
flowrate
 
Lamp
age
 
Lamp
or
ballast
failure
 
Low
liquid
level
in
the
UV
reactor
 
High
temperature
 
OMC
system
failure
5.1.2.4
Verification
of
Flow
Distribution
and
Headloss
If
each
reactor
is
not
equipped
with
a
dedicated
flowmeter,
then
it
will
be
necessary
to
verify
the
flow
split
between
reactors
over
the
entire
operating
flow
range.
This
flow
split
and
the
total
plant
flow
should
be
used
to
estimate
the
flow
through
each
UV
reactor
and
confirm
Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
7
June
2003
operation
is
within
the
validated
conditions.
Clamp­
on
type
flowmeters
or
differential
pressure
readings
across
each
parallel
reactor
are
alternatives
for
field
verification
of
the
flow
split.

The
allowable
difference
in
flow
among
reactors
(
flow
split
differential)
is
established
during
validation
and
should
be
accounted
for
in
the
validation
protocol
safety
factor
(
section
4.2.1
and
section
F.
5).
If
the
actual
flow
split
differential
is
greater
than
assumed
in
validation,
then
steps
should
be
taken
to
improve
the
flow
split.
The
Tier
1
recommendations
for
validation
of
UV
reactors
(
section
4.6)
necessitates
a
flow
split
differential
of
10
percent
or
less.
If
this
is
not
observed
during
functional
testing,
then
a
Tier
2
analysis
for
validation
safety
factor
needs
to
be
completed
for
the
UV
reactor.
Appendix
C
provides
details
about
the
Tier
1
and
2
analysis
and
Appendix
F
provides
details
about
the
development
of
the
safety
factor.

The
headloss
should
be
measured
for
each
reactor
and
compared
to
the
headloss
specified
in
the
contract
documents
(
if
applicable).
Pressure
transducers
or
pressure
gauges
can
be
used
to
measure
the
headloss.

5.1.3
Performance
Testing
Performance
testing
is
intended
to
assess
the
operating
performance
of
the
UV
reactor
as
a
whole,
as
well
as
the
individual
performance
of
its
components.
While
functional
testing
is
primarily
completed
through
simulations
of
specific
operating
conditions,
performance
testing
is
generally
accomplished
through
extensive
monitoring
of
reactor
performance
during
the
early
stages
of
continuous
operation.
It
is
important
to
note
that
performance
testing
is
not
intended
to
validate
disinfection
performance,
which
is
completed
during
validation
testing
(
as
described
in
Chapter
4).
However,
performance
testing
can
be
used
to
confirm
that
the
actual
operating
conditions
are
within
the
constraints
established
during
validation
testing.
Performance
testing
focuses
on
the
accuracy,
reliability,
and
repeatability
of
UV
reactor
operation,
whereas
validation
is
used
to
measure
the
effectiveness
of
the
UV
reactor
at
delivering
the
UV
doses
required
for
target
pathogen
inactivation
credit.

When
UV
lamps
are
first
energized,
they
go
through
a
stabilizing
period
called
"
burn­
in."
For
some
UV
lamp
designs,
the
initial
lamp
output
may
significantly
exceed
the
design
value.
During
burn­
in,
the
lamp
output
may
rapidly
decrease
to
a
value
more
consistent
with
the
design.
Following
burn­
in,
lamp
output
becomes
relatively
stable
until
the
end
of
lamp
life
is
approached.
Typically,
new
UV
lamps
will
not
have
undergone
burn­
in
prior
to
installation.
Because
performance
testing
should
compare
actual
operating
conditions
to
validated
conditions,
it
is
important
that
the
lamps
be
in
the
same
condition
as
they
were
during
validation
testing.
Therefore,
UV
lamps
should
be
burned­
in
prior
to
performance
testing,
which
typically
takes
100
hours
of
continuous
operation.
The
actual
required
burn­
in
time
should
be
discussed
with
the
manufacturer
and
confirmed
through
documented
operating
experience
at
other
UV
installations.

The
duration
of
performance
testing
and
the
extent
of
monitoring
will
be
project­
specific
and
should
be
established
by
the
utility
and
designer
based
on
the
objectives
of
the
performance
testing.
Performance
testing
may
range
in
duration
from
as
little
as
48
hours
of
uninterrupted
operation
to
greater
than
four
months
of
demonstrative
operation.
Similarly,
the
scope
of
the
testing
may
range
from
an
increased
monitoring
frequency
to
confirm
performance
to
an
Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
8
June
2003
extensive
testing
protocol
to
fully
optimize
reactor
performance
and
establish
long­
term
operating
procedures.
During
performance
testing,
treated
water
may
be
sent
to
the
distribution
system
if
upstream
treatment
has
not
changed
and
meets
existing
regulations.
However
this
should
be
confirmed
with
the
State.

Performance
testing
may
include
the
following
items:

 
Operation
of
each
UV
reactor
in
automatic
mode
and
demonstration
that
actual
operating
conditions
are
within
the
constraints
established
during
validation
testing
 
Demonstration
of
UV
reactor
start­
up
and
switchover
sequences
that
result
from
water
quality
and/
or
flowrate
changes
 
Observation
of
operation,
including
periods
of
off­
specification
operation,
due
to
power
quality
problems,
and
other
alarm
conditions
 
Measurement
of
electrical
service
voltage,
current,
and
power
consumption
with
different
flow
and
water
quality
combinations
to
optimize
energy
use
within
the
constraints
established
during
validation
 
Assessment
of
the
effectiveness
of
the
cleaning
system
by
inspecting
sleeve
clarity
and
condition
at
regular
intervals
throughout
the
test
period
 
Confirmation
that
the
programmed
cleaning
frequency
correlates
with
the
actual
frequency
of
cleaning
 
Verification
of
UV
intensity
sensor
operation
 
Confirmation
of
duty
sensor
accuracy
using
reference
sensors
(
see
section
5.3.2.2)

 
Observation
of
ballast
temperature
and
cooling
system
performance
 
Verification
of
the
accuracy
and
repeatability
of
the
on­
line
UVT
monitor
through
the
collection
of
grab
samples
and
analysis
using
a
bench­
top
spectrophotometer
(
if
applicable)

 
Confirmation
of
backup
generator
and/
or
UPS
power
transfer
to
the
UV
reactor.
This
may
necessitate
simulation
of
line
power
failure
to
trigger
the
backup
power
supply.
It
is
recommended
that
the
backup
power
supply
be
tested
for
a
minimum
of
two
separate
one­
hour
periods.

The
performance
testing
should
be
tailored
to
the
specific
UV
installation.
An
example
monitoring
program
for
a
4­
week
performance
test
is
shown
in
Table
5.1.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
9
June
2003
Table
5.1
Example
Monitoring
During
a
Four
Week
Performance
Test
Frequency
Task
Notes
Confirm
the
operating
setpoint(
s)
Monitor
reactor
operation
to
confirm
compliance
with
the
setpoint(
s)
established
during
validation.

Continuous
Develop
energy
efficient
operation
Monitor
the
power
consumption.
Test
the
automatic
operation
and
power
consumption
under
the
flow
and
water
quality
variations
to
determine
if
energy
efficiency
improvements
can
be
made
within
the
validation
constraints.

Check
the
on­
line
UVT
monitor
calibration
Check
the
on­
line
UVT
monitor
against
a
bench­
top
spectrophotometer
to
determine
if
the
on­
line
unit
is
in
calibration.
Weekly
Check
UV
intensity
sensor
calibration
Check
the
duty
sensor
against
a
reference
sensor,
using
the
recommended
protocol
(
section
5.3.2.2)
to
determine
whether
the
duty
sensor
is
in
calibration.

Switch
to
standby
reactor
Monitor
the
time
it
takes
to
switch
to
a
standby
reactor
to
determine
if
there
will
be
off­
specification
operation
during
switchover.
Twice
during
testing
period
Switch
to
standby
power
or
UPS
Monitor
the
time
it
takes
to
switch
to
the
standby
power
supply
to
determine
if
there
will
be
off­
specification
operation
because
of
power
transfer.

After
4
weeks,
100
OMC
cycles
or
one
Off­
line
chemical
clean
(
OCC)
Inspect
lamp
sleeves
for
fouling
Remove
a
sleeve
from
the
reactor
and
inspect
as
recommended
in
section
5.3.2.3.

Any
off­
specification
time
and
flow
should
be
recorded
during
all
performance
tests,
and
these
results
should
be
evaluated
to
ensure
that
off­
specification
requirements
are
met.
During
performance
testing,
any
component
that
is
not
operating
properly
should
be
corrected
and
retested
to
ensure
satisfactory
operation.
This
may
necessitate
manufacturer
involvement,
especially
if
specifications
in
the
contract
documents
were
not
met.
Following
performance
testing,
ongoing
monitoring
and
recording
of
reactor
operation
should
continue
at
a
reduced
frequency
as
discussed
in
section
5.4
and
as
required
by
the
State.

5.1.4
Operations
and
Maintenance
Manual
The
O&
M
manual
should
be
site­
specific
and
based
on
as­
built
drawings,
manufacturer's
shop
drawings,
operating
procedures,
recommended
maintenance
tasks,
and
results
from
the
performance
testing.
If
possible,
the
O&
M
manual
should
be
developed
prior
to
routine
operations.
At
a
minimum,
O&
M
manuals
should
include
the
following
items:

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
10
June
2003
 
Federal
and
State
regulatory
requirements
and
guidelines
 
Overall
treatment
objectives
 
Role
of
the
UV
installation
in
the
overall
disinfection
strategy
 
Relationship
to
adjoining
unit
processes
 
UV
reactor
design
criteria
 
UV
reactor
validation
criteria
 
General
description
of
UV
installation
 
Controls
and
monitoring
 
Standard
operating
procedures
 
Start­
up
procedures
 
Shutdown
procedures
(
manual
and
automatic)

 
Safety
issues
 
Emergency
procedures
and
contingency
plan
 
Alarm
response
plans
 
Preventative
maintenance
needs
and
procedures
 
Equipment
calibration
needs
and
procedures
 
Troubleshooting
guide
 
Equipment
component
summary
 
Spare
parts
inventory
 
Contact
information
for
equipment
manufacturers
and
technical
services
5.2
Operation
of
UV
Installations
The
operation
of
UV
installations
will
vary
based
on
the
UV
manufacturer,
the
UV
reactor
configuration,
and
the
dose
control
strategy.
This
section
discusses
the
required
and
recommended
operational
and
routine
start­
up
and
shutdown
procedures
that
are
common
to
all
UV
reactors.
The
operational
tasks
presented
in
this
section
are
general
in
nature,
and
the
specific
operational
procedures
for
the
installed
UV
reactors
should
be
developed
with
assistance
from
the
manufacturer
and
UV
installation
designer.
Examples
of
how
to
determine
the
operational
requirements
are
presented
in
section
5.5.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
11
June
2003
5.2.1
Operational
Requirements
To
receive
inactivation
credit,
the
UV
reactors
are
required
to
operate
within
the
validated
limits
(
40
CFR
141,
Subpart
W,
Appendix
D).
When
a
UV
reactor
is
operating
outside
of
these
limits,
the
UV
reactor
is
operating
off­
specification
as
described
previously.
Unfiltered
systems
that
use
UV
disinfection
to
meet
the
Cryptosporidium
treatment
requirement
of
the
Long­
Term
2
Enhanced
Surface
Water
Treatment
Rule
(
LT2ESWTR)
must
demonstrate
that
at
least
95
percent
of
the
water
delivered
to
the
public
during
each
month
is
treated
by
UV
reactors
operating
within
validated
limits
(
40
CFR
141.721(
c)(
2)).
In
other
words,
the
UV
reactors
cannot
operate
off­
specification
for
more
than
5
percent
of
the
water
delivered
to
the
public.

The
LT2ESWTR
does
not
establish
an
off­
specification
requirement
for
filtered
systems;
however,
States
may
adopt
a
5
percent
off­
specification
or
more
stringent
requirement.
Although
the
specific
criteria
limiting
off­
specification
water
are
defined
by
the
State,
the
United
States
Environmental
Protection
Agency
(
EPA)
recommends
that
the
UV
reactors
be
operated
to
minimize
off­
specification
water.
The
UV
reactors
must
operate
under
the
validated
conditions
that
are
determined
based
on
validation
testing
(
section
5.5)
(
40
CFR
141,
Subpart
W,
Appendix
D).
The
specific
monitoring
requirements
associated
with
off­
specification
are
described
in
section
5.4.

5.2.2
Recommended
Operational
Tasks
UV
reactors
typically
use
automatic
control
systems
and
do
not
need
significant
operational
attention.
This
section
outlines
the
general
operational
tasks
that
are
recommended
(
Table
5.2).
Site­
specific
operational
tasks
should
be
determined
by
the
manufacturer,
UV
installation
designer,
and
facility
operators,
and
should
be
described
in
the
O&
M
manual
(
section
5.1.4).
Recommended
maintenance
tasks
are
discussed
in
section
5.3.1.

Table
5.2
Recommended
Operational
Tasks
for
the
UV
Reactor
Frequency
Recommended
Tasks
Daily
 
Perform
overall
visual
inspection
of
the
all
UV
reactors.
 
Ensure
system
control
is
on
automatic
mode
(
if
applicable).
 
Check
control
panel
display
for
status
of
system
components
and
alarm
status
and
history.
 
Ensure
all
on­
line
analyzers,
flowmeters,
and
data
recording
equipment
are
operating
normally.
 
Review
24­
hour
monitoring
data
to
ensure
that
the
reactor
has
been
operating
within
validated
limits
during
that
period.
Weekly
 
Initiate
manual
operation
of
wipers
(
if
provided)
to
ensure
proper
operation.

Monthly
 
Check
lamp
run
time
values.
Consider
changing
lamps
if
operating
hours
exceed
design
life
or
UV
intensity
is
low.

Semiannually
 
Check
ballast
cooling
fans
for
unusual
noise.
 
Check
operation
of
automatic
and
manual
valves.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
12
June
2003
5.2.3
Start­
up
and
Shutdown
of
UV
Reactors
UV
reactors
may
be
turned
on
and
off
regularly
in
response
to
varying
flowrate
and
water
quality.
This
section
describes
the
routine
start­
up
procedures,
shutdown
procedures,
and
winterization
of
the
UV
reactors.
The
routine
start­
up
and
shutdown
procedures
shown
are
not
all
inclusive.
Utilities
should
modify
these
procedures
based
on
the
specific
manufacturer's
recommendations
and
operating
requirements
for
their
system.

5.2.3.1
Routine
Start­
up
The
following
start­
up
procedure
serves
as
an
example
procedure.
The
UV
reactors
should
be
operating
within
validated
conditions
once
the
start­
up
sequence
is
complete.

1.
Follow
site­
specific
procedures
for
removal
of
lockouts
and
tag­
outs
of
the
power
supply
and
control
panel.

2.
Ensure
all
lamp
and
ground
connections
are
properly
made.
Verify
that
all
incoming
power
conductors,
including
ground
conductors
are
properly
terminated.

3.
Ensure
that
the
lamp
ends
and
all
other
reactor
ports
are
covered
and/
or
sealed
to
eliminate
the
potential
for
operator
exposure
to
UV
light.

4.
Ensure
the
breakers
are
turned
on,
and
all
electrical
cabinets
and
equipment
are
clear
and
closed.

5.
Initiate
the
UV
reactors'
start­
up
sequence.

6.
Initiate
water
flow
(
if
it
is
not
automatically
done
in
UV
reactor
controls)
to
the
reactor
and
gradually
increase
the
flow
until
the
minimum
flow
required
for
lamp
cooling
is
reached.
The
water
exiting
the
reactor
is
not
disinfected
and
is
considered
off­
specification.

7.
Verify
that
all
air
is
purged
from
reactors
(
i.
e.,
reactor
completely
full).
Check
the
top
of
the
reactor
for
heat
buildup,
which
indicates
an
air
pocket.

8.
Check
the
UV
reactor
control
panel
to
ensure
that
all
of
the
lamps
are
on
and
all
of
the
monitoring
parameters
are
being
displayed.

9.
Check
and
resolve
any
system
alarms
being
displayed.

10.
Ensure
all
of
the
on­
line
analyzers
(
UV
intensity
sensors
and
UVT
monitors,
if
applicable)
and
flowmeters
are
operating
as
intended.

11.
After
lamp
warm­
up
period,
increase
flow
to
the
minimum
validated
flow
(
if
flow
is
not
automatically
adjusted
with
UV
reactor
control
sequence).

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
13
June
2003
12.
Verify
correct
flow
split
between
parallel
UV
reactors
using
flowmeters
and/
or
differential
pressure
gauges.

13.
Verify
that
the
UV
reactor
is
operating
within
validated
limits.

5.2.3.2
Routine
Shutdown
UV
reactors
will
need
to
be
shut
down
periodically
for
maintenance
or
to
accommodate
water
quality
or
flow
changes.
The
main
steps
involved
in
shutting
reactors
down
are
as
follows:

1.
Throttle
the
effluent
valve
(
if
not
part
of
the
control
sequence)
to
reduce
flow
through
the
reactor
to
the
minimum
required
for
cooling.
If
complete
closure
of
the
effluent
valve
can
be
accomplished
without
overheating
the
lamps,
it
is
recommended.

2.
De­
energize
the
reactors.

3.
Close
effluent
valve
if
not
completed
in
Step
1.
The
water
exiting
the
de­
energizing
reactor
is
considered
off­
specification.

4.
If
maintenance
is
being
performed,
the
following
steps
should
be
followed.
If
the
UV
reactor
is
to
be
placed
on
standby,
the
following
steps
are
not
necessary.

5.
Follow
lock
out
and
tag­
out
procedures
for
the
facility.

6.
Drain
the
reactor
if
necessary
for
the
specific
maintenance
task.

7.
Inspect
and
repair
or
replace
any
necessary
equipment.

After
an
extended
shutdown
period
(
greater
than
30
days),
the
operator
should
perform
a
cleaning
and
then
inspect
the
lamp
sleeves
for
fouling.
Additional
cleaning
may
be
necessary
prior
to
start­
up.

5.2.3.3
Winterization
In
most
drinking
water
applications,
the
UV
reactors
will
probably
be
located
within
a
building.
However,
in
some
instances,
the
reactors
may
be
located
in
unheated
concrete
vaults.
When
it
is
necessary
to
shut
down
a
UV
reactor
for
an
extended
period
of
time
and
freeze
damage
is
possible,
the
UV
reactors
should
be
winterized
in
accordance
with
the
manufacturer's
recommendations.

5.3
Maintenance
of
UV
Reactors
There
are
no
specific
regulatory
requirements
for
maintenance
of
a
UV
reactor.
However,
the
UV
reactors
need
to
be
maintained
to
ensure
that
disinfection
requirements
are
met.
Poor
maintenance
may
cause
the
UV
reactors
to
be
operating
off­
specification.
As
part
of
Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
14
June
2003
the
maintenance
tasks,
UV
reactor
components
will
need
to
be
replaced;
therefore,
an
inventory
of
spare
parts
is
necessary.
These
tasks
are
described
in
this
section.

5.3.1
Summary
of
Recommended
Maintenance
Tasks
Table
5.3
summarizes
the
recommended
maintenance
tasks
and
refers
to
the
general
guidelines
for
those
tasks
that
are
discussed
in
section
5.3.2.
Before
any
maintenance
is
performed,
the
main
electrical
supply
to
the
UV
reactors
should
be
disconnected,
lockout
and
tag­
out
protocol
should
be
followed,
and
the
operator
should
wait
at
least
5
minutes
(
or
as
recommended
by
the
manufacturer)
for
the
lamps
to
cool
down
and
energy
to
dissipate.

Table
5.3.
Recommended
Maintenance
Tasks
Frequency
Task
General
Guideline
Section
Reference
Action
Weekly
Check
on­
line
UVT
monitor
calibration
section
5.3.2.5
Calibrate
UVT
monitor
when
manufacturer's
guaranteed
measurement
uncertainty
is
exceeded.

Monthly
Check
reactor
housing,
sleeves,
and
wiper
seals
for
leaks
Replace
housing,
sleeve,
or
wiper
seals
if
damaged
or
leaking.

Monthly
UV
intensity
sensor
calibration
check
protocol
section
5.3.2.2
Check
the
sensor
calibration
at
the
lamp
power
utilized
during
routine
operating
conditions
(
e.
g.,
the
majority
of
operation).
A
sensor
is
out
of
calibration
when
it
fails
the
criteria
shown
in
section
5.3.2.2
When
UV
intensity
sensor
fails
calibration
check
Replace
duty
sensor
with
calibrated
backup
sensor
section
5.3.2.2
 
Check
the
reference
sensor
with
second
reference
sensor
or
two
other
duty
sensors
to
ensure
the
first
reference
sensor
is
calibrated.
 
If
reference
sensor
is
properly
calibrated,
replace
the
duty
sensor
with
calibrated
sensor,
and
send
the
duty
sensor
that
failed
calibration
to
the
manufacturer.
 
Check
the
replaced
sensor
one
hour
later.
Monthly
(
OCC)
Semi­
annually
(
OMC)
Check
cleaning
efficiency
section
5.3.2.4
 
Record
UV
intensity
sensor
reading.
 
Extract
one
sleeve
per
reactor
(
or
bank
of
lamps
for
low
pressure
high
output
(
LPHO)
reactors)
for
inspection.
 
Check
remaining
sleeves
if
fouling
is
observed
on
the
first
sleeve.
 
Manually
clean
sleeve(
s)
if
fouling
is
seen
on
the
sleeves.
 
Record
UV
intensity
sensor
reading
and
compare
to
original
reading
after
cleaning.
 
Replace
sleeve
if
UV
intensity
is
not
restored
to
validated
level.
Semi­
annually
(
OMC)
Check
cleaning
fluid
reservoir
(
if
provided)
section
5.3.2.4
Replenish
solution
if
the
reservoir
level
is
low.
Drain
and
replace
solution
if
the
solution
is
discolored.

Annually
Calibrate
reference
sensor
section
5.3.2.2
Send
the
reference
sensor
to
the
manufacturer
for
calibration.

Annually
Test­
trip
GFI
section
5.3.2.8
Maintain
ground
fault
interrupt
(
GFI)
breakers
in
accordance
with
the
manufacturer's
recommendations.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
15
June
2003
Table
5.3.
Recommended
Maintenance
Tasks
(
continued)

Frequency
Task
General
Guideline
Section
Reference
Action
Manufacturer's
recommended
frequency
Check
flowmeter
calibration
section
5.3.2.6
Calibrate
flowmeter
when
manufacturer's
guaranteed
measurement
uncertainty
is
exceeded.

Lamp/
manufacturer
specific
Replace
lamp
section
5.3.2.1
Replace
lamps
when
any
one
of
the
following
conditions
occur:
 
Initiation
of
low
UV
intensity
alarm
(
UV
intensity
equal
to
or
less
than
set
point
value)
after
verifying
that
this
condition
is
caused
by
low
lamp
output.
 
Initiation
of
lamp
failure
alarm.
When
lamps
are
replaced
Properly
dispose
of
lamps
section
5.3.2.1
Send
spent
lamps
to
a
mercury
recycling
facility
or
back
to
the
manufacturer.

Sleeve/
Manufacturer
specific
Replace
sleeve
section
5.3.2.3
Replace
sleeve
every
3
to
5
years
or
when
damage,
cracks,
or
excessive
fouling
significantly
decreases
UV
intensity
of
an
otherwise
acceptable
lamp
to
the
minimum
validated
intensity
level.
The
replacement
frequency
should
be
adjusted
based
on
operational
experience.
Pressure
gauge
manufacturer
specific
Check
operation
of
the
pressure
gauges
that
are
used
to
confirm
flow
split
(
if
applicable)
section
5.3.2.6
Replace
the
pressure
gauge
if
deemed
faulty
by
manufacturer's
evaluation
procedure.

Manufacturer
specific
Clean
UVT
monitor
Clean
according
to
manufacturer's
recommended
procedure.

Manufacturer
specific
Inspect
OMC
drive
mechanism
Inspect
and
maintain
OMC
drive
routinely
as
recommended
by
the
manufacturer.
Manufacturer
specific
Inspect
ballast
cooling
fan
Check
the
ballast
cooling
fans
for
dust
buildup
and
damage.
Replace
if
necessary.

5.3.2
General
Guidelines
for
UV
Reactor
Maintenance
This
section
describes
general
guidelines
for
UV
reactor
components
that
relate
to
maintenance
tasks.
Specific
operations,
maintenance,
and
monitoring
tasks
are
described
individually
in
later
sections.
These
latter
sections
also
refer
back
to
this
section
as
a
reminder
of
the
general
recommendations.

5.3.2.1
UV
Lamp
Characteristics
UV
lamp
output
decreases
over
time,
and
UV
lamps
will
need
to
be
replaced
periodically
to
maintain
sufficient
UV
intensity
(
i.
e.,
the
validated
UV
intensity
setpoint).
Replacement
lamps
should
be
identical
to
those
used
during
reactor
validation
with
respect
to
arc
length,
lamp
envelope
material
and
dimensions,
amount
of
mercury,
and
spectral
output.
If
the
lamps
supplied
are
not
equal
to
the
lamps
used
during
validation,
the
UV
reactor
is
not
operating
as
validated
and
is
considered
off­
specification.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
16
June
2003
If
the
mercury
content
or
power
rating
changes,
the
different
lamp
should
be
assessed
by
comparing
UV
intensity
sensor
readings,
after
burn­
in,
to
the
lamps
that
were
validated
to
determine
if
the
new
lamps
are
equal
to
the
validated
lamps.
If
the
sensor
reading
is
equal
to
or
greater
than
that
of
the
validated
lamps
after
burn­
in,
the
different
lamps
are
acceptable
and
comparable
to
the
validated
lamps.
However,
if
a
utility
replaces
the
lamps
with
higher
power
lamps
to
receive
higher
log
inactivation
credit,
validation
testing
should
be
performed
to
confirm
performance.
Lamp
manufacturers
should
also
provide
documentation
of
lamp
output
decay
characteristics,
guaranteed
life,
and
lamp
burn­
in
period.
This
information
will
help
the
utility
determine
the
lamp
replacement
frequency.
It
should
be
noted
that
different
lamps
might
have
different
aging
characteristics,
which
may
affect
operations
and
maintenance
costs.

The
frequency
of
UV
lamp
replacement
can
be
based
on
a
utility­
determined
schedule,
lamp
operating
hours,
or
the
UV
intensity
reduction
as
measured
by
the
UV
intensity
sensor
(
after
sleeve
and
sensor
window
cleaning);
lamp
replacement
recommendations
are
discussed
in
section
5.3.2.1.
During
replacement,
the
lamps
and
sleeves
should
be
handled
in
accordance
with
manufacturer
recommendations,
using
clean
cotton,
powder­
free
latex,
or
vinyl
gloves
because
fingerprints
can
cause
damage
to
the
lamps
or
sleeves
during
operation.

Lamp
manufacturers
are
required
to
determine
whether
their
products
exhibit
the
toxicity
characteristic
for
mercury
and
whether
their
lamp
is
regulated
as
a
universal
hazardous
waste
under
Subtitle
C
of
Resource
Conservation
and
Recovery
Act
(
RCRA)
[
40
CFR
Part
260,
261,
264
and
273].
Currently,
most
UV
lamps
exceed
these
toxicity
characteristics
and
require
regulated
disposal.
As
such,
these
lamps
should
be
sent
to
a
mercury
recycling
facility
where
the
mercury
is
recovered
and
lamp
components
are
recycled.
Some
UV
reactors
and
lamp
manufacturers
will
accept
spent
or
broken
lamps
for
recycling
or
proper
disposal
(
Dinkloh
2001;
Lienberger
2002;
Gump
2002).
Utilities
should
contact
their
lamp
manufacturer
to
determine
if
they
accept
spent
lamps
or
should
contact
their
State
for
a
list
of
local
mercury
recycling
facilities.

5.3.2.2
UV
Intensity
Sensors
Well
performing
UV
intensity
sensors
are
necessary
to
assess
whether
the
validated
UV
intensity
is
being
achieved.
Sensor
calibration,
rotation,
and
placement
affect
operation.
This
section
describes
these
effects
and
provides
recommendations
to
minimize
them.

There
are
two
types
of
sensors
used
for
UV
reactor
operation:
duty
and
reference
sensors.
Duty
sensors
are
on­
line
sensors
and
continuously
monitor
UV
intensity,
while
the
reference
sensors
are
off­
line
sensors
used
to
assess
the
duty
sensor
performance.
Therefore,
the
reference
sensor
specifications
should
exactly
match
those
of
the
duty
sensors,
so
that
a
valid
comparison
can
be
completed.
Both
duty
and
reference
sensors
are
described
in
this
section.

Duty
UV
Intensity
Sensor
Calibration
Prior
to
installation,
manufacturers
calibrate
the
UV
intensity
sensors.
However,
over
time
the
sensor
may
drift
out
of
calibration.
Because
these
sensors
are
vital
to
assessing
the
UV
disinfection
performance,
the
calibration
of
each
sensor
should
be
checked
at
least
monthly
Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
17
June
2003
against
the
reference
sensor.
To
assess
the
calibration,
the
following
sensor
calibration
check
protocol
should
be
followed:

1.
Measure
the
UV
intensity
with
the
duty
sensor,
and
record
the
measurement
result.

2.
Replace
the
duty
sensor
with
the
reference
sensor
in
the
same
location
(
i.
e.,
port)
as
the
duty
sensor
used
in
Step
1.

3.
Measure
the
UV
intensity
with
the
reference
sensor
and
record
the
measurement
result.

4.
Determine
if
Equation
5.1
holds
true
for
the
two
UV
intensity
sensor
readings:

(
)
2
1
2
Duty
2
Ref
Ref
uty
D
100
*
1
I
I
 
+
 
 

 

  
 

 

  
 
 
Equation
5.1
where
IRef
=
Intensity
measured
with
the
reference
sensor
(
mW/
cm2)
IDuty
=
Intensity
measured
with
the
on­
line
sensor
(
mW/
cm2)
 
Duty
=
Measurement
uncertainty
of
the
on­
line
UV
intensity
sensor
(%)
as
provided
by
the
UV
manufacturer
in
the
validation
report
 
Reference
=
Measurement
uncertainty
of
the
reference
UV
intensity
sensor
(%)
as
provided
by
the
UV
manufacturer
in
the
validation
report
5.
Replace
the
duty
sensor
with
another
calibrated
duty
sensor
if
the
relationship
Equation
5.1
does
not
hold
true.

The
calibration
of
the
UV
intensity
sensor
is
sensitive
to
the
power
level
of
the
UV
lamps
(
Swaim
et
al.
2002).
To
most
effectively
compare
the
duty
sensor
to
the
reference
sensor,
the
power
level
should
be
set
at
the
level
typically
used
during
routine
operation
(
e.
g.,
the
majority
of
operation).

UV
Intensity
Sensor
Rotation
Some
UV
intensity
sensors
are
sensitive
to
their
rotational
alignment
within
the
sensor
port
and
will
have
different
readings
at
different
rotations.
This
may
be
due
to
the
UV
intensity
sensor
configuration
(
e.
g.,
acceptance
angle).
Section
A.
3.5
discusses
UV
intensity
sensors
configurations
in
more
detail.
The
sensors
should
be
rotated
until
the
lowest
UV
intensity
reading
is
obtained
for
routine
monitoring
purposes.
Alternatively,
UV
reactors
may
be
designed
so
the
UV
intensity
sensors
are
keyed
in
the
same
rotational
position
at
all
times.
This
may
not
be
an
issue
for
all
UV
intensity
sensors.

Measuring
Lamp
Output
Variability
UV
lamp
output
differs
for
each
lamp,
depending
on
lamp
age
and
lot.
As
discussed
in
section
2.4.6,
a
sensor
measures
the
UV
intensity
at
its
location
in
the
UV
reactor
and
cannot
assess
lamp
output
variability
unless
there
is
one
sensor
per
lamp.
Many
low
pressure
(
LP)
or
LPHO
reactors
have
one
sensor
to
monitor
a
bank
of
lamps,
and
some
MP
reactors
use
one
UV
Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
18
June
2003
intensity
sensor
to
monitor
more
than
one
lamp
in
the
reactor.
The
effect
of
variable
lamp
output
is
accounted
for
in
the
validation
protocol
safety
factor
as
discussed
in
section
F.
3.
For
routine
operation,
the
oldest
lamp
should
be
placed
in
the
position
closest
to
the
UV
intensity
sensor
if
one
sensor
monitors
multiple
lamps.

Reference
UV
Intensity
Sensor
The
reference
sensor
should
be
calibrated
at
least
once
per
year
at
a
qualified
facility
(
e.
g.,
manufacturer)
to
ensure
that
it
is
calibrated
properly
for
the
regular
duty
sensor
calibration
checks.
The
reference
sensor
should
not
be
exposed
to
UV
light
for
longer
than
it
takes
to
perform
the
reference
sensor
measurement.
When
not
in
use,
the
reference
sensor
should
be
stored
under
conditions
that
will
maintain
its
integrity
and
accuracy
as
recommended
by
the
manufacturer.
If
the
reference
sensor
is
found
to
be
out
of
calibration,
the
calibration
interval
should
be
shortened.
One
indicator
that
the
reference
sensor
itself
may
be
out
of
calibration
is
if
it
shows
that
all
on­
line
sensors
are
out
of
calibration.
Some
utilities
may
choose
to
have
multiple
reference
sensors
to
help
determine
if
one
reference
sensor
is
out
of
calibration,
as
a
replacement
reference
sensor,
or
to
allow
multiple
duty
sensors
to
be
checked
simultaneously.

5.3.2.3
Lamp
Sleeves
Lamp
sleeves
degrade
over
time
due
to
solarization
(
section
2.4.4)
and
internal
sleeve
fouling,
resulting
in
cloudiness
and
a
loss
of
UV
transmittance.
Abrasion
of
the
sleeve
surface
during
handling
or
mechanical
cleaning
may
also
be
a
contributing
factor
to
the
loss
of
UV
transmittance.
Sleeve
transmittance
loss
is
reflected
in
the
UV
intensity
sensor
reading
and,
therefore,
does
not
need
to
be
monitored.
However,
a
low
UV
intensity
sensor
reading
may
be
from
sleeve
transmittance
loss
and
should
be
considered
when
troubleshooting
the
cause
of
this
problem
(
as
discussed
in
section
5.6.1).
Sleeves
will
need
to
be
replaced
in
the
case
of
UV
transmittance
loss
or
other
damage.

Sleeves
should
be
replaced
every
3
to
5
years
or
when
damage,
cracks
or
excessive
fouling
diminishes
UV
intensity
to
the
minimum
validated
intensity
level,
whichever
occurs
first.
This
replacement
frequency
should
be
increased
or
decreased
based
on
operational
experience.
Replacement
sleeves
should
be
identical
to
the
sleeves
used
during
validation,
meet
the
design
and
UV
manufacturer's
material
and
construction
specifications,
and
be
certified
as
described
in
section
F.
6.3.
The
sleeves
should
be
handled
in
accordance
with
manufacturer
recommendations,
using
clean
cotton,
powder­
free
latex,
or
vinyl
gloves
because
fingerprints
can
cause
damage
to
the
sleeves
during
operation.
When
the
sleeves
are
replaced,
the
manufacturer's
procedure
should
be
closely
followed
because
the
lamp
sleeve
can
crack
and
break
from
overtightening
of
the
compression
nuts
that
hold
it
in
place.

5.3.2.4
Fouling
As
discussed
in
Chapters
2
and
3,
the
lamp
sleeves
and
UV
intensity
sensors/
windows
may
foul
over
time,
depending
on
the
water
quality,
lamp
type,
and
cleaning
regime.
This
section
describes
possible
cleaning
techniques
and
provides
some
specific
recommendations
for
addressing
fouling
issues.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
19
June
2003
Sleeve
and
UV
Intensity
Sensor
Surface/
Window
Fouling
There
are
two
types
of
sleeve
cleaning
techniques
as
discussed
in
section
2.4.5.
The
first
type
is
an
OMC
system,
which
typically
utilizes
an
automatic
mechanical
wiper
(
e.
g.,
O­
ring,
brush)
to
wipe
the
surface
of
the
sleeve
at
a
prescribed
frequency.
Some
OMC
systems
have
Orings
with
cleaning
fluid
enclosed
in
them
to
enhance
cleaning.
The
second
type
is
an
OCC,
which
is
also
a
called
flush
and
rinse
system.
OCC
systems
are
off­
line,
manual
systems
that
pump
cleaning
solution
(
typically
an
acid)
into
the
reactor
and
circulate
the
solution
for
a
period
of
time.
Helsinki
Water
uses
an
OCC
system;
a
description
of
their
cleaning
regime
is
discussed
in
Appendix
O.
Also,
OCC
systems
clean
the
sensor
wetted
surface/
window;
however,
OMC
systems
may
not,
depending
on
the
UV
reactor.

The
frequency
of
cleaning
is
site­
specific.
An
appropriate
sleeve
cleaning
frequency
(
manual
or
automatic)
can
be
determined
based
on
the
rate
of
fouling
during
the
start­
up
period,
which
can
be
assessed
by
monitoring
the
UV
intensity
sensor
measurement.
For
routine
operation,
the
cleaning
frequency
should
be
increased
or
decreased
based
on
the
amount
of
fouling
left
on
the
sleeves
after
the
cleaning
cycle
and
the
loss
of
UV
intensity
prior
to
cleaning.

Sleeves
should
initially
be
inspected
for
fouling
every
six
months
if
OMC
is
employed
and
every
month
if
OCC
is
used.
This
frequency
should
be
adjusted
after
2
years
of
operating
data
are
available.
A
decrease
in
UV
intensity
may
indicate
sleeve
fouling,
and
sleeves
should
be
inspected
if
fouling
is
the
suspected
cause
of
the
UV
intensity
drop.
In
addition,
the
sensor
windows
(
if
applicable)
should
be
inspected
for
fouling
and
supplemental
cleaning
should
be
conducted
if
necessary,
according
to
the
manufacturers
recommendation.

For
sleeve
inspection,
one
sleeve
per
reactor
(
or
bank
of
lamps
for
LP
or
LPHO
reactors)
should
be
inspected.
The
sleeves
should
be
handled
in
the
same
manner
as
described
for
UV
lamps.
If
damage
or
fouling
is
observed,
the
remaining
sleeves
should
be
inspected.
External
fouling
can
be
difficult
to
identify.
Sleeve
discoloration
is
more
easily
seen
by
laying
the
sleeve
on
a
clean,
white,
lint­
free
cloth
along
side
of
a
new
sleeve.
If
streaks
are
observed,
this
may
indicate
that
the
OMC
wiper
material
may
be
worn
or
damaged
or
not
aligned
properly;
therefore,
the
wiper
should
also
be
inspected.
If
fouling
is
observed,
the
cleaning
frequency
should
be
increased,
and
supplemental
manual
cleaning
should
be
conducted
as
necessary.

If
manual
cleaning
(
i.
e.,
beyond
routine
OCC
or
OMC
cleaning)
of
lamp
sleeves
is
necessary,
this
should
be
done
according
to
manufacturer
recommendations
and
procedures.
Abrasive
cleaners
or
pads
that
might
scratch
the
lamp
sleeve
should
not
be
used.
In
addition,
the
inside
of
the
sleeve
should
be
dry
prior
to
re­
installation
because
water
or
cleaning
solutions
could
cause
a
coating
to
form
during
operation.
One
method
of
drying
the
sleeve
is
to
use
isopropyl
alcohol
and
a
lint­
free
cloth;
however,
there
should
not
be
any
alcohol
left
inside
the
sleeve
after
this
procedure.
As
noted
earlier,
when
the
sleeves
are
re­
installed
after
inspection,
the
manufacturer's
procedure
should
be
closely
followed
to
avoid
over­
tightening
of
the
compression
nuts.

If
OMC
cleaning
is
used,
the
OMC
wipers
should
be
checked
for
deformation
or
degradation
at
the
same
time
the
sleeves
are
checked.
If
the
OMC
cleaning
uses
a
cleaning
solution,
the
cleaning
solution
reservoir
should
be
checked
every
six
months
to
determine
Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
20
June
2003
whether
more
solution
should
be
added.
In
addition
the
solution
should
be
replaced
if
it
is
discolored
or
if
the
OMC
system
is
not
effectively
cleaning
the
sleeve.

Fouling
During
Periods
of
Standby
When
the
UV
reactors
are
out­
of­
service
and
full
of
water,
the
sleeves
may
become
fouled
(
Toivanen
2000).
The
rate
of
fouling
is
site­
specific
and
depends
on
the
influent
water
quality.
UV
reactors
equipped
with
OMC
should
continue
to
clean
the
sleeves
even
though
the
UV
reactor
is
off­
line.
This
should
prevent
fouling
of
the
sleeves.
For
UV
reactors
that
do
not
include
OMC,
the
utility
should
consider
draining
the
UV
reactor
if
it
is
off­
line
for
more
than
one
week.
However,
this
period
could
be
shorter
or
longer,
depending
on
the
water
quality.
After
an
extended
shutdown
period
of
greater
than
30
days,
the
operator
should
perform
a
cleaning
(
OCC
or
OMC)
and
then
inspect
the
lamp
sleeves
for
fouling.
Additional
cleaning
may
be
necessary
prior
to
start­
up
after
extended
periods
of
standby.

5.3.2.5
On­
line
UVT
Monitor
Calibration
On­
line
UVT
measurements
should
be
compared
to
those
obtained
using
a
bench­
top
spectrophotometer
every
week.
The
grab
samples
that
are
used
to
check
calibration
should
be
collected
from
a
location
close
to
the
on­
line
UVT
monitor
sampling
point.
The
frequency
may
be
decreased
or
increased
based
on
the
performance
demonstrated
over
a
one­
year
period.
For
example,
the
frequency
could
be
reduced
to
once
per
month
if
the
UVT
monitor
was
consistently
within
the
calibration
specification
for
over
a
month
during
the
first
year
of
monitoring.

5.3.2.6
Flowmeter
Calibration
The
flowmeter
calibration
should
be
checked
at
the
frequency
recommended
by
the
manufacturer.
Techniques
for
verifying
calibration
are
discussed
in
section
5.1.2.2.

Some
UV
installations
will
not
have
dedicated
flowmeters
and
may
use
a
combination
of
an
upstream
flowmeter
and
differential
pressure
gauges
to
verify
flow
split
as
described
in
section
3.3.1.2.
If
differential
pressure
is
used
to
verify
the
flow
split,
the
calibration
of
the
main
flowmeter
should
be
checked
at
the
manufacturer's
recommended
frequency
and
the
accuracy
of
the
pressure
gauges
should
be
periodically
verified
using
a
reference
gauge
or
redundant
gauge
to
confirm
measurement
consistency
between
the
gauges.

5.3.2.7
UV
Reactor
Temperature
UV
lamps
operate
at
high
temperatures
(
as
discussed
in
section
2.4.2)
and
need
water
flow
to
maintain
them
at
their
optimal
temperature
and
to
prevent
overheating.
Another
concern
related
to
overheating
is
the
formation
of
air
pockets
in
the
UV
reactor.
Air
pockets
can
cause
the
UV
reactor
temperature
to
increase
and
may
alter
the
flow
pattern
in
the
UV
reactor.
UV
lamps
can
break
if
their
threshold
temperature
is
exceeded,
which
is
discussed
in
more
detail
in
section
N.
2.1.2.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
21
June
2003
The
water
temperature
should
be
monitored.
If
the
water
temperature
exceeds
manufacturer
recommendations,
the
UV
reactor
should
be
shut
down.
Water
level
monitoring
or
reactor
temperature
monitoring
are
typically
included
in
the
packaged
control
systems
for
the
UV
reactor.
The
water
level
monitoring
should
detect
any
air
pockets
in
the
UV
reactor.
During
start­
up
and
whenever
necessary,
air
should
be
bled
from
the
UV
reactors.
The
UV
reactor
surface
can
become
hot
during
operation
if
air
pockets
or
stagnant
water
are
present
in
the
UV
reactor.
As
a
result,
nothing
unrelated
to
reactor
equipment
should
be
in
external
contact
with
the
reactor
while
in
service.

5.3.2.8
Electrical
Concerns
UV
reactors
operate
at
high
voltages.
Before
any
maintenance
on
the
UV
reactor
is
performed,
the
main
electrical
supply
to
the
UV
reactors
should
be
disconnected
and
the
operator
should
wait
at
least
5
minutes
for
the
lamps
to
cool
down
and
energy
to
dissipate.
Lockout,
tagout
procedures
and
all
applicable
codes
should
be
followed.
The
UV
reactors
should
not
be
operated
if
any
of
the
control
panel
doors
are
open,
and
water
should
not
be
sprayed
around
the
electrical
equipment.

Typically,
power
to
the
UV
reactors
are
provided
via
a
distribution
transformer,
a
circuit
breaker,
a
disconnect
switch
at
the
UV
reactor,
and
related
wires
and
conduits.
If
maintenance
is
necessary
on
the
control
panel,
the
main
electrical
supply
should
be
disconnected.
The
power
to
the
lamps
is
typically
delivered
through
individual
GFI
circuit
breakers
and
ballasts.
Maintenance
of
the
GFI
breakers
is
important
because
they
are
safety
devices
that
protect
the
operators
when
they
are
working
around
the
powered
equipment.
The
GFI
breakers
should
be
test­
tripped
at
least
once
per
year
and
should
be
maintained
in
accordance
with
the
manufacturer's
recommendations.
Ballast
output
should
be
monitored
through
the
UV
reactor's
control
panel.
Irregularities
or
instabilities
in
ballast
output
may
indicate
a
problem
with
the
electrical
feed
or
the
ballast
itself.

The
ballasts,
typically
connected
between
the
GFI
breakers
and
the
lamps,
are
electrical
components
that
regulate
the
line
power
to
match
the
input
requirement
of
the
lamps.
Three
types
of
ballasts
are
typically
used
with
UV
reactors
for
converting
power:
electronic
ballasts,
electromagnetic
ballasts,
and
transformers.
Electromagnetic
ballasts
and
transformers
are
very
similar
in
that
both
contain
a
specially
wound
coil
of
wire
that
is
used
to
control
the
current
to
the
lamp.
Typically
inductors
or
capacitors
are
used
to
allow
step
adjustment
of
the
lamp
output.
Electronic
ballasts,
sometimes
referred
to
as
solid­
state
ballasts,
contain
semiconductors
and
other
electronic
components
that
allow
the
ballast
to
behave
like
a
switching
power
supply.
Electronic
ballast
technology
allows
nearly
continuous
adjustment
of
lamp
output.

Power
regulation,
particularly
with
electromagnetic
ballasts
and
transformers,
will
result
in
significant
heat
build­
up
within
the
ballast
enclosure.
If
the
excess
heat
is
not
dissipated,
it
can
damage
the
ballast
electronics
and
cause
failure.
A
cooling
system
is
normally
provided
with
LPHO
and
medium
pressure
(
MP)
reactors
to
maintain
the
ballast
temperature
below
the
maximum
specified
limit.
LP
reactors
typically
do
not
need
ballast
cooling.
The
ballast
cooling
system
should
be
inspected
and
maintained
as
recommended
by
the
manufacturer.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
22
June
2003
Power
use
depends
on
the
specific
UV
reactor
and
how
it
adjusts
to
changes
in
water
quality
and
flow.
Power
use
should
be
monitored
as
operational
adjustments
are
made
for
changes
in
flow,
UV
intensity,
UVT,
lamp
aging
and
output,
and
other
factors.
This
information
can
be
used
to
determine
the
most
energy
efficient
operating
strategies.
For
example,
some
UV
reactors
can
both
increase
lamp
output
and
energize
additional
lamps
to
respond
to
a
low
UV
intensity
reading.
The
power
use
under
these
two
strategies
can
be
compared
to
determine
which
is
more
energy
efficient.

5.3.3
Spare
Parts
The
actual
life
of
a
component
is
a
function
of
many
variables,
including
operating
conditions,
maintenance
practices,
the
quality
of
the
materials
of
construction,
and
fabrication
practices.
As
a
consequence,
it
is
impossible
to
predict
the
actual
life
of
a
component.
To
overcome
the
operational
impacts
of
this
uncertainty,
an
adequate
inventory
of
critical
spare
parts
should
be
maintained
to
ensure
reliable
and
consistent
performance
of
the
UV
reactors
and
minimize
the
delivery
of
off­
specification
water.

All
UV
components
have
a
design
life
and
a
guaranteed
life.
The
design
life
represents
the
expected
duration
of
operation.
The
guaranteed
life
incorporates
the
risk,
assumed
by
the
manufacturer,
to
account
for
the
uncertainties
associated
with
the
quality
of
materials
used,
production,
and
operating
conditions.
Generally,
guarantees
are
conditional
in
nature
and
are
valid
under
certain
operating
conditions.
For
example,
guaranteed
lamp
life
is
normally
linked
to
the
lamp
power
setting
or
the
number
of
on/
off
cycles
per
24­
hour
period.
If
equipment
failure
occurs
during
the
warranty
period
and
if
all
of
the
warranty
conditions
are
satisfied,
the
manufacturer
will
typically
replace
the
component
and
charge
the
owner
a
prorated
fee
for
the
use
of
the
replaced
component.

Table
5.4
provides
typical
design
and
guaranteed
lives
for
major
UV
reactor
components.
These
represent
current
industry
trends
and
are
likely
to
change
as
more
operation
and
maintenance
information
becomes
available
and
technological
advances
occur.
Manufacturers
should
be
contacted
directly
for
details
specific
to
their
equipment.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
23
June
2003
Table
5.4.
Design
and
Guaranteed
Lives
of
Major
UV
Components
(
Based
on
Manufacturers'
Input)

Component
Design
Life1
Guaranteed
Life
2
Low
pressure
lamps
(
LP
and
LPHO)
12,000
hours
8,000
­
12,000
hours
MP
lamps
10,000
hours
4,000
­
8,000
hours
Sleeve
8
to
10
years
1
to
3
years
UV
Intensity
Sensor
3
to
10
years
1
year
UVT
monitor
3
to
5
years
1
year
Cleaning
systems
3
to
5
years
1
to
3
years
Ballasts
10
to
15
years
1
to
3
years
1
Expected
duration
of
operation
2
Accounts
for
variability
of
material
quality,
production,
and
operating
conditions.

The
following
is
a
suggested
minimum
inventory
of
spare
parts,
expressed
as
a
percentage
of
the
installed
number.
A
full
list
of
spare
parts
will
vary
depending
on
the
specific
equipment
installed
and
should
be
coordinated
with
the
UV
manufacturer.
The
number
of
spare
parts
needed
depends
on
the
guaranteed
life
of
the
specific
equipment.
For
example,
a
higher
percentage
of
MP
lamps
may
be
necessary
compared
to
LP
lamps
because
the
guaranteed
lamp
life
is
less
for
MP
lamps,
and
therefore
they
need
to
be
replaced
more
frequently.

 
UV
lamps­
10
percent
with
a
minimum
of
two
lamps
 
Sleeves­
5
percent
with
a
minimum
of
one
sleeve
 
O­
ring
Seals­
5
percent
with
a
minimum
of
two
seals
 
OMC
wipers­
5
percent
with
a
minimum
of
two
wipers
 
OMC
wiper
drive
mechanisms­
2
percent
with
a
minimum
of
one
drive
 
Ballasts­
5
percent
with
a
minimum
of
one
unit
 
Ballast
cooling
fan­
1
unit
 
Duty
UV
intensity
sensor­
minimum
of
2
units
(
adjust
number
based
on
operating
experience)

 
Reference
UV
intensity
sensor­
2
units
 
On­
line
UVT
monitor­
1
unit
(
if
used
for
control
strategy)

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
24
June
2003
5.4
Monitoring,
Recording,
and
Reporting
of
UV
Installation
Operation
Operation
of
the
UV
reactors
should
be
monitored
to
ensure
the
reactors
are
operating
within
validated
limits,
to
diagnose
operating
problems,
to
determine
when
maintenance
is
necessary,
and
to
maintain
safe
operation.
This
section
discusses
the
required
and
recommended
monitoring,
recording,
and
reporting
activities
for
UV
installations.

5.4.1
Monitoring
and
Recording
Frequency
for
Compliance
Parameters
Utilities
must
monitor
each
reactor
to
determine
whether
it
is
operating
within
validated
conditions.
They
also
must
determine
the
percentage
of
flow
that
was
treated
within
validated
limits
(
40
CFR
141,
Subpart
W,
Appendix
D).
The
flow
is
off­
specification
when
a
reactor
is
operating
outside
of
validated
limits.
The
monitoring
parameters
depend
on
the
control
strategy
used
and
the
validation
results.
Table
5.5
presents
the
monitoring
parameters
for
each
control
strategy,
the
criteria
for
when
off­
specification
occurs,
and
examples
of
off­
specification
operating
conditions.

Table
5.5
Off­
Specification
Operations
for
Each
Control
Strategy
Control
Strategy
Parameters
Monitored
Off­
Specification
Examples
UV
intensity
setpoint
UV
intensity,
flowrate,
lamp
status
Anytime
these
values
are
outside
of
the
validated
limits
for
these
parameters
1)
UV
intensity
below
setpoint
2)
Flowrate
outside
validated
limits
3)
UV
lamp
failure
4)
UV
intensity
sensor
failure
UVT
and
UV
intensity
setpoints
UV
intensity,
flowrate,
UVT,
lamp
status
Anytime
these
values
are
outside
of
the
validated
limits
for
these
parameters
1)
UV
intensity
below
setpoint
2)
Flowrate
outside
validated
limits
3)
UV
lamp
failure
4)
UV
intensity
sensor
failure
5)
UVT
below
setpoint
Calculated
dose
Calculated
dose,
flowrate,
UVT,
lamp
status
Anytime
the
calculated
dose
is
below
the
validated
setpoint
(
if
validation
certifies
that
the
calculated
dose
can
be
used
to
control
the
UV
reactor
 
see
section
F.
2)
1
1)
Calculated
dose
below
setpoint
2)
Flowrate
outside
validated
limits
3)
UV
lamp
failure
4)
UV
intensity
sensor
failure
5)
UVT
below
setpoint
1
If
validation
deems
that
the
calculated
dose
control
is
not
acceptable,
the
UV
reactor
should
use
the
UVT
and
UV
intensity
setpoint
control
strategy.

It
is
recommended
that
the
required
monitoring
parameters
be
continuously
monitored
for
each
UV
reactor
and
that
these
values
be
recorded
at
least
once
every
four
hours.
These
fourhour
records
should
be
used
to
determine
the
percentage
of
flow
that
is
off­
specification.
Very
small
systems
(
e.
g.,
systems
serving
less
than
500
people)
that
are
unable
to
record
reactor
status
every
4
hours
(
e.
g.,
manual
recording
is
practiced)
can
consider
a
reduced
recording
frequency;
however,
the
frequency
should
not
be
less
than
once
per
day
and
should
be
approved
by
the
State.
The
monitoring
guidelines
are
summarized
in
Table
5.6.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
25
June
2003
Table
5.6
Monitoring
Parameters
and
Recording
Frequency
Parameter
General
Guideline
Section
Reference
(
if
applicable)
Recommended
Recording
Frequency
Notes
UV
intensity
Every
4
hours
The
UV
intensity
must
be
above
the
validated
setpoint
UVT1
Every
4
hours
The
UVT
must
be
above
the
validated
setpoint.
If
not
required
to
be
monitored,
this
information
will
assist
in
determining
if
low
UV
intensity
readings
are
because
of
low
UVT
Calculated
dose1
Every
4
hours
The
calculated
dose
must
be
above
the
validated
setpoint
Lamp
status
Every
4
hours
The
lamps
should
be
energized
if
water
is
flowing
through
the
UV
reactor
Calibration
of
UV
intensity
sensors
section
5.3.2.2
Monthly
The
UV
intensity
sensor
calibration
must
be
checked,
using
sensor
calibration
check
protocol
1
Only
required
if
necessary
for
the
control
strategy
(
Table
5.11)

5.4.2
Monitoring
and
Recording
for
Other
Operational
Parameters
In
order
to
minimize
operational
problems,
facilitate
regulatory
compliance,
and
evaluate
UV
reactor
performance,
it
is
recommended
that
additional
parameters,
beyond
those
needed
for
regulatory
compliance,
be
monitored.
Table
5.7
presents
these
additional
parameters
recommended
for
monitoring
and
the
recommended
recording
frequency.
These
recommended
parameters
and
their
monitoring
frequency
should
be
adjusted
based
on
site­
specific
operating
experience.
For
example,
if
sleeve
fouling
is
a
maintenance
issue
and
supplemental
reactor
cleaning
is
frequent
(
e.
g.,
monthly),
then
the
fouling
parameters
should
be
monitored
on
a
daily
basis
as
opposed
to
weekly
as
shown
in
the
table
below.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
26
June
2003
Table
5.7
Recommended
Monitoring
Parameters
and
Recording
Frequency
Parameter
General
Guideline
Section
Reference
(
if
applicable)
Monitoring
Frequency
Recording
Frequency
Notes
Power
draw
section
5.3.2.8
Continuous
Every
4
hours
This
information
can
be
used
to
determine
the
most
energy
efficient
operation
strategies
Water
Temperature
section
5.3.2.7
Continuous
Daily
Monitor
to
ensure
the
high
temperature
limit
is
not
exceeded
(
usually
part
of
packaged
UV
control
system)
UV
lamp
on/
off
cycles
section
5.3.2.1
Continuous
Weekly
(
Total
cycles
in
a
week)
Monitor
to
assess
status
of
the
UV
lamps
since
the
of
on/
off
cycles
can
help
assess
lamp
aging
Turbidity
Daily
Weekly
Monitor
if
chemicals
(
e.
g.,
lime)
are
added
post­
filtration
or
prior
to
UV
disinfection
(
monitoring
may
not
be
necessary
for
many
UV
reactors)
pH,
iron,
calcium,
alkalinity,
hardness
section
5.3.2.4
Weekly
(
reduce
if
fouling
is
not
prevalent)
Weekly
Monitor
to
help
assess
fouling
issues
if
necessary
UVT
monitor
calibration
section
5.3.2.5
Weekly
(
reduce
if
appropriate
based
on
operational
experience)
Weekly
Information
can
assist
in
planning
scheduled
maintenance
and
O&
M
budget
Age
of
the
following
equipment:
 
Lamp
 
Ballast
 
Sleeve
 
UV
intensity
sensor
Monthly
Monthly
Information
can
assist
in
planning
scheduled
maintenance
and
O&
M
budget
Calibration
of
flowmeter
section
5.3.2.6
Monthly
Monthly
Information
can
assist
in
planning
scheduled
maintenance
and
O&
M
budget
All
data
related
to
UV
reactor
operation
should
be
gathered,
compiled,
and
stored
for
easy
retrieval.
The
recorded
data
should
be
stored
for
at
least
two
years.
Appendix
M
provides
example
logs
for
many
of
the
parameters
listed
in
Table
5.13.

5.4.3
Reporting
to
the
State
Monthly
reports
must
be
prepared
and
submitted
to
the
State.
The
report
must
include
the
percentage
of
off­
specification
flow,
which
should
be
based
on
at
least
4­
hour
records
for
each
reactor.
The
State
may
have
additional
reporting
requirements.
In
addition,
the
percentage
of
the
UV
intensity
sensors
that
were
checked
for
calibration
must
be
reported
monthly;
all
sensors
should
be
checked
every
month.
An
example
monthly
monitoring
form
is
shown
in
the
Appendix
M.
The
State
should
be
contacted
to
determine
the
specific
content
of
the
monthly
reports
and
to
coordinate
with
other
reporting
requirements.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
27
June
2003
5.5
Determination
of
Validated
Operational
Parameters
For
each
UV
reactor,
the
operating
conditions
associated
with
a
given
level
of
inactivation
credit
must
be
defined
based
on
validation
testing
results
(
40
CFR
141,
Subpart
W,
Appendix
D).
The
validation
testing
and
resultant
data
that
are
used
to
determine
these
operating
conditions
will
vary
with
different
control
strategies.
A
detailed
discussion
of
the
three
common
control
strategies
is
presented
in
section
3.3.2.
A
brief
description
of
each
of
the
control
strategies
is
shown
in
Table
5.8.

Table
5.8
UV
Reactor
Control
Strategies
Control
Strategy
Dose
Delivery
Monitoring
and
Control
Based
On
UV
Intensity
Setpoint
UV
intensity
sensor
measurement
UV
Intensity
and
UVT
setpoints
UV
intensity
sensor
and
UVT
measurement
Calculated
Dose
The
calculated
UV
dose1
1
The
UV
reactor
calculates
a
UV
dose
using
the
UV
intensity
sensor
measurement,
the
UVT
of
the
water,
and
the
flowrate.

This
section
provides
example
operational
requirements
based
on
the
validation
examples
described
in
section
C.
5
of
the
validation
protocol.
Each
example
describes
how
the
operating
requirements
are
determined
based
on
the
control
and
operation
strategy
used
and
the
validation
results.

Example
1.
UV
Intensity
Setpoint
Control
­
Single
Operational
Setpoint
for
all
Conditions
(
Section
C.
5.1)

The
simplest
operational
strategy
uses
one
single
UV
intensity
setpoint
for
all
flows.
In
this
example,
a
LPHO
reactor
that
uses
the
UV
intensity
setpoint
control
strategy
was
validated
at
flows
between
100
and
500
gallons
per
minute
(
gpm)
and
a
UVT
range
of
84
to
98
percent.
This
reactor
passed
the
criteria
for
2­
log
inactivation
of
Cryptosporidium
with
an
intensity
sensor
setpoint
of
5
mW/
cm2.
The
validation
testing
verified
that
the
UV
intensity
setpoint
control
strategy
is
appropriate
for
this
reactor
Based
on
this
validation,
this
reactor
must
operate
at
a
minimum
UV
intensity
sensor
setpoint
of
5.0
mW/
cm2
and
a
flow
range
between
100
and
500
gpm
to
claim
2
log
Cryptosporidium
credit.
The
UV
intensity
setpoint
approach
accounts
for
the
UVT
in
the
UV
intensity
measurement.
Therefore,
the
intensity
setpoint
of
5.0
mW/
cm2
can
be
used
for
any
UVT.
Although
this
is
a
simple
and
straightforward
operating
strategy,
single
setpoint
operation
will
not
be
as
energy
efficient
as
using
a
variable
setpoint
approach,
which
is
described
in
Example
2.

Example
2.
UV
Intensity
Setpoint
Control
­
Variable
Setpoint
Operation
for
Different
Flow
Conditions
(
Section
C.
5.2)

The
variable
UV
intensity
setpoint
approach
has
a
different
UV
intensity
setpoint
at
different
flowrates.
This
operation
promotes
more
energy
efficient
operation
compared
to
the
Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
28
June
2003
single
setpoint
approach
because
the
UV
intensity
setpoint
can
be
decreased
at
lower
flows.
For
this
example,
a
LPHO
reactor
that
uses
the
UV
intensity
setpoint
control
strategy
was
validated
under
the
conditions
shown
in
Table
5.9
and
passed
the
criteria
for
3­
log
inactivation
of
Cryptosporidium
at
each
condition.

Table
5.9
Example
Validation
Data
for
Variable
Setpoint
Operation
Flow
(
mgd)
UVT
(%)
UV
Intensity
(
mW/
cm2)
0.90
70
6.1
1.2
75
7.5
1.7
83
10
2.4
92
14
The
UV
intensity
measurements
recorded
during
validation
verified
that
the
UV
intensity
setpoint
approach
is
appropriate
for
this
reactor.
Because
of
the
data
collected,
this
UV
reactor
can
be
operated
at
a
different
setpoint
for
each
flow
range.
These
intensity
setpoints
could
be
used
in
three
ways.

1.
A
single
setpoint
as
described
in
Example
1.
For
example,
a
setpoint
of
14
mW/
cm2
could
be
used
at
between
0.90
and
2.4
mgd
with
any
UVT.

2.
Each
intensity
setpoint
could
be
used
over
a
given
flow
range
as
shown
in
Table
5.10.
The
higher
UV
intensity
measurement
from
each
flow
range
should
be
used
as
the
UV
intensity
setpoint
to
be
conservative.

Table
5.10
UV
Intensity
Setpoint
for
Different
Flow
Ranges
Minimum
Flow
(
mgd)
Maximum
Flow
(
mgd)
UV
Intensity
(
mW/
cm2)
0.90
1.2
7.5
1.2
1.7
10
1.7
2.4
14
3.
The
intensity
setpoints
could
be
interpolated
as
a
function
of
flowrate.
Figure
5.2
presents
an
equation
based
on
interpolation.
For
example,
for
a
flowrate
of
2
mgd,
interpolation
indicates
that
a
setpoint
of
12
mW/
cm2
is
needed
to
achieve
3­
log
inactivation.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
29
June
2003
Figure
5.2
Example
2
 
Interpolation
of
Validation
Data
to
Determine
UV
Intensity
Setpoints
y
=
5.1979x
+
1.3063
0
5
10
15
0
1
2
3
Flowrate
(
mgd)
Intensity
setpoint
(
mW/
cm2)
5.20x
+
1.31
Example
3.
UV
Intensity
Setpoint
Control
­
Variable
Setpoint
Operation
for
Different
Flow
Conditions
and
Inactivation
Goals
(
Section
C.
5.3)

For
this
example,
a
UV
manufacturer
has
completed
a
matrix
of
tests
at
different
flowrates,
UVT,
and
lamp
power
to
develop
a
relationship
between
UV
intensity
readings,
log
inactivation
credit,
and
flow.
Table
5.11
shows
the
results
of
the
validation
tests.

Table
5.11
Example
Validation
Data
for
Variable
Setpoint
Operation
Flow
(
mgd)
UV
Intensity
(
mW/
cm2)
Cryptosporidium
Log
Credit
5
5.1
3.0
5
3.3
2.5
5
1.8
1.0
10
9.1
3.0
10
5.6
2.5
10
2.6
1.0
20
15
3.0
20
11
2.0
20
5.6
1.0
The
UV
intensity
measurements
recorded
during
validation
verified
that
the
UV
intensity
setpoint
approach
is
valid
for
this
reactor.
In
contrast
to
Example
2,
this
reactor
was
validated
for
three
different
levels
of
Cryptosporidium
inactivation
credit.
For
a
utility
that
only
is
required
to
achieve
a
2.0­
log
inactivation,
using
this
reactor
would
reduce
energy
costs
compared
to
a
reactor
that
had
only
been
validated
for
3.0­
log
Cryptosporidium
inactivation.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
30
June
2003
These
intensity
setpoints
could
be
used
in
three
ways
to
achieve
2.0­
log
Cryptosporidium
inactivation
credit
with
this
reactor,

1.
A
single
setpoint
as
described
in
Example
1.
For
example,
a
setpoint
of
11
mW/
cm2
could
be
used
at
or
between
5
and
20
mgd
with
any
UVT.

2.
Each
intensity
setpoint
could
be
used
over
a
given
flow
range
as
shown
in
Table
5.12.
The
higher
UV
intensity
measurement
from
each
flow
range
should
be
used
as
the
UV
intensity
setpoint
to
be
conservative.

Table
5.12
UV
Intensity
Setpoint
for
Different
Flow
Ranges
Minimum
Flow
(
mgd)
Maximum
Flow
(
mgd)
UV
Intensity
(
mW/
cm2)
5
10
5.6
10
20
11
3.
The
intensity
setpoints
could
be
interpolated
as
a
function
of
flowrate.
Figure
5.3
presents
an
equation
based
on
interpolation
for
three
different
levels
of
Cryptosporidium
inactivation.
For
example,
for
a
flowrate
of
12
mgd,
interpolation
indicates
that
a
setpoint
of
3.8
mW/
cm2
is
needed
to
achieve
2­
log
inactivation.

Figure
5.3
Example
3
 
Interpolation
of
Validation
Data
to
Determine
UV
Intensity
Setpoints
at
Different
Flows
and
Cryptosporidium
Inactivation
y
=
0.001x2
+
0.517x
+
0.872
y
=
0.005x2
+
0.176x
+
0.968
y
=
0.003x2
+
0.346x
+
0.920
0
2
4
6
8
10
12
14
0
5
10
15
20
25
Flowrate
(
gpm)
Intensity
setpoint
(
mW/
cm2)
2.0
log
Crypto
2.5
log
Crypto
3.0
log
Crypto
Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
31
June
2003
Example
4.
UV
Intensity
and
UVT
Setpoint
Control
Strategy
­
Single
Operational
Setpoint
for
all
Conditions
(
Section
C.
5.4)

This
example
uses
single
operational
setpoint
as
the
operating
strategy,
which
is
the
same
as
example
1.
However,
this
example
uses
both
a
UV
intensity
and
a
UVT
setpoint
to
control
the
reactor
operation.
In
this
example,
a
MP
reactor
that
uses
the
UV
intensity
and
UVT
setpoint
control
strategy
was
validated
at
flows
between
0.1
and
0.5
mgd
and
a
UVT
range
of
75
to
98
percent.
This
reactor
passed
the
criteria
for
3­
log
Cryptosporidium
inactivation
credit
with
a
UV
intensity
sensor
setpoint
of
41
mW/
cm2
and
a
UVT
setpoint
of
85
percent.

Therefore,
to
claim
3­
log
Cryptosporidium,
this
reactor
must
operate
under
the
following
conditions:

 
Maintain
minimum
UV
intensity
sensor
setpoint
of
41.0
mW/
cm2.

 
Operate
within
a
flow
range
of
0.1
mgd
and
0.5
mgd.

 
Operate
within
a
UVT
range
of
85
to
98
percent.

Example
5.
Calculated
Dose
Setpoint
Control
­
Variable
Setpoint
Operation
for
Different
Flow
Conditions
and
Inactivation
Goals
(
Section
C.
5.5)

The
calculated
dose
control
strategy
uses
UVT,
UV
intensity,
and
flow
measurements
to
estimate
a
UV
dose.
For
this
example,
a
UV
manufacturer
has
completed
a
matrix
of
tests
at
different
flowrates,
UVT,
and
lamp
power
to
develop
a
relationship
between
calculated
dose,
log
inactivation,
and
flow.
A
MP
reactor
that
uses
the
calculated
dose
control
strategy
was
validated
at
flows
between
10
to
40
mgd
and
a
UVT
range
of
75
to
98
percent.
Table
5.13
shows
the
results
of
the
validation
tests.

Table
5.13
Dose
Setpoints
for
Various
Log
Inactivation
of
Cryptosporidium
Cryptosporidium
Log
Inactivation
Calculated
Dose
Setpoint
(
mJ/
cm2)
UVT
Range
(%)

1.0
14
75
­
98
1.5
18
75
­
98
2.0
23
75
­
98
2.5
28
75
­
98
3.0
30
79
­
98
The
validation
tests
as
described
in
section
C.
5.5
verified
that
the
calculated
dose
approach
is
valid
for
this
reactor
and
that
the
calculated
dose
setpoints
could
be
used
for
the
ranges
of
flows
tested
(
10
 
40
mgd).
In
addition,
this
reactor
can
be
utilized
by
utilities
that
need
different
levels
of
Cryptosporidium
inactivation
credit.
For
a
utility
that
only
is
required
to
achieve
a
2.0
log
inactivation
credit,
using
this
reactor
would
reduce
energy
costs
compared
to
a
reactor
that
had
only
been
validated
for
3­
log
Cryptosporidium
inactivation
credit.
Therefore,
this
reactor
could
be
operated
at
any
flow
between
10
and
40
mgd,
the
UVT
range
specified
in
Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
32
June
2003
Table
5.8,
and
at
the
specified
calculated
dose
in
Table
5.13
to
achieve
a
specific
level
of
Cryptosporidium
inactivation
credit.
For
example,
a
reactor
must
operate
at
a
minimum
calculated
dose
of
28
mJ/
cm2
and
a
flow
range
between
10
and
40
mgd
and
UVT
between
75
and
98
percent
to
achieve
2.5­
log
Cryptosporidium
inactivation
credit.

5.6
Operational
Challenges
An
excursion
from
validated
limits
can
be
caused
by
low
UV
intensity,
low
UVT,
high
or
low
flowrate,
poor
UV
intensity
sensor
performance,
power
quality
problems,
or
a
combination
of
these
conditions.
These
conditions
will
need
to
be
resolved
quickly
to
ensure
regulatory
compliance
because
they
can
result
in
prolonged
off­
specification
operation.
This
section
discusses
some
of
the
potential
operational
challenges
and
suggested
corrective
measures.

5.6.1
Low
UV
Intensity
or
Low
Calculated
UV
Dose
Although
the
UV
intensity
and
calculated
dose
control
strategies
are
different,
approaches
for
addressing
either
a
low
UV
intensity
or
low
calculated
dose
are
typically
the
same.
This
is
because
the
UV
intensity
setpoint
control
strategy
uses
UV
intensity
as
an
indicator
for
UV
dose;
therefore,
the
causes
of
a
low
UV
intensity
in
a
UV
intensity
control
strategy
and
a
low
calculated
dose
in
a
calculated
dose
control
strategy
are
similar.

The
output
of
the
UV
lamps,
UV
transmittance
of
the
sleeves,
status
of
the
UV
intensity
sensor,
and
fouling
of
both
lamp
sleeves
and
sensor
windows
affect
UV
intensity
sensor
readings.
In
the
UV
intensity
setpoint
control
strategy,
UV
intensity
sensors
are
placed
far
enough
from
the
UV
lamp
to
be
affected
by
UVT.
In
the
UV
intensity
and
UVT
setpoint
or
calculated
dose
setpoint
control
strategy,
the
UV
intensity
sensors
are
close
to
the
lamps
and
should
not
be
affected
by
UVT
changes.

If
one
or
more
UV
intensity
sensors
reads
below
the
required
setpoint,
the
cause
could
be
low
UV
lamp
output.
If
the
UV
lamp
life
is
greater
than
the
design
life,
the
lamp
should
be
replaced.
If
the
UV
intensity
is
still
low,
sensor
accuracy
should
be
determined
by
replacing
the
duty
sensor
with
the
reference
sensor.
If
the
duty
and
reference
sensor
agree
within
the
required
uncertainty
(
from
validation),
the
cause
of
the
low
intensity
reading
may
be
due
to
UV
intensity
sensor
surface
or
sensor
window
fouling
or
sleeve
UV
transmittance
loss.
Potential
corrective
measures
include
cleaning
of
fouled
surfaces
and
replacement
of
defective
sleeves.

Figure
5.4
presents
a
decision
tree
for
evaluating
low
UV
intensity
problems.
If
the
above
strategies
cannot
be
implemented
or
are
not
successful
in
reducing
the
low
UV
intensity,
the
UV
manufacturer
or
UV
installation
designer
should
be
contacted
to
investigate
the
problem
further.
The
utility
should
activate
any
backup
disinfection
or
consider
shutting
down
the
water
treatment
plant
(
WTP)
until
the
UV
intensity
is
within
the
validated
limits.
Anytime
that
the
UV
intensity
is
lower
than
the
validated
limit,
it
should
be
recorded
as
off­
specification
even
if
this
does
not
occur
at
precisely
the
time
(
e.
g.,
4­
hour
interval)
when
the
4­
hour
recording
is
completed.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
33
June
2003
Figure
5.4
Low
UV
Intensity
or
Low
Calculated
UV
Dose
Decision
Chart
UV
Intensity
or
calculated
dose
is
below
validated
limits
Is
the
UVT
low
or
below
validated
limits?

Replace
duty
sensor(
s)
with
reference
sensor(
s)

Is
UV
intensity
or
calculated
dose
still
low?
Evaluate
and
repair
faulty
sensor
Is
UV
intensity
or
calculated
dose
still
low?
Replace
lamp
Is
UV
intensity
or
calculated
dose
still
low?
Replace
sleeve
Check
other
lamps
and/
or
sleeves
in
other
reactors
to
see
if
they
need
to
be
replaced
No
No
See
evaluation
of
low
UVT
(
Figure
5.4)
Yes
Is
the
UV
lamp
age
beyond
the
design
life?

Clean
sleeve
and
or
sensor
surface/
window
Inspect
other
reactors
and
sensors
for
fouling
Take
out
quartz
sleeve
and/
or
sensor
window
and
inspect
for
fouling
Are
the
sleeves
or
sensor
surface
/
windows
fouled?

Is
the
quartz
sleeves
age
beyond
the
design
life?

Contact
UV
manufacturer
and/
or
UV
facility
designer
to
evaluate
problem
further
and
activate
backup
disinfection
or
consider
WTP
shutdown
Yes
Yes
No
Yes
Yes
No
Yes
Yes
No
Adjust
UV
system
operation
to
compensate
for
low
UV
intensity
if
not
done
automatically
by
the
control
system.
Is
the
UV
intensity
or
calculated
dose
still
low?

Yes
Continue
operation
No
No
Is
the
UV
intensity
or
calculated
dose
still
low?

Yes
Continue
operation
No
No
Continue
operation
Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
34
June
2003
5.6.2
Low
UV
Transmittance
This
evaluation
of
low
UVT
assumes
either
that
the
low
intensity
evaluation
has
been
completed
and
the
cause
of
the
low
UV
intensity
was
low
UVT
or
that
the
operational
staff
has
observed
low
UVT.
Some
UV
reactors
may
increase
lamp
output
or
number
of
lamps
in
service
to
accommodate
a
decrease
in
UVT.
If
the
system
does
not
sufficiently
compensate,
or
if
the
UV
reactor
cannot
adjust
lamp
output,
the
UV
intensity
may
go
below
the
validated
limits.
The
steps
for
evaluating
low
UVT
are
described
below.

The
first
step
is
to
evaluate
the
UVT
monitor
function.
If
UVT
is
monitored
using
an
online
instrument,
the
utility
should
verify
the
low
reading
with
a
bench­
top
spectrophotometer.
If
the
second
measurement
differs
significantly
from
the
on­
line
instrument
response,
appropriate
repair
and
calibration
of
the
on­
line
instrument
is
necessary.

If
UVT
is
determined
using
grab
samples,
a
duplicate
sample
should
be
obtained
and
analyzed.
If
the
UVT
of
the
duplicate
sample
remains
low,
the
spectrophotometer
response
should
be
checked
using
a
phthalate
standard
(
EPA
ICR
UV254
method
or
Standard
Method
5910).
If
the
spectrophotometer
response
is
determined
to
be
inaccurate,
the
spectrophotometer
monitor
should
be
calibrated
or
repaired.

If
the
low
UVT
is
determined
to
be
real
and
not
due
to
a
faulty
instruments,
it
should
be
compared
to
the
validated
UVT
set
point.
If
UVT
is
below
the
validated
UVT
set
point,
the
following
operational
changes
should
be
considered:

 
Vary
source
water
blending
ratio
(
if
available)
to
increase
UVT.

 
Evaluate
whether
the
coagulation
process
has
been
optimized
for
natural
organic
matter
(
NOM)
removal
and
whether
the
coagulant
dose
should
be
increased.
Poor
coagulation
caused
by
coagulant
under­
dosing
can
lead
to
increased
NOM
concentration
and
an
associated
decrease
in
UVT.

 
Increase
oxidant
dose
prior
to
the
UV
installation
if
possible.
However,
this
strategy
may
increase
disinfection
byproduct
(
DBP)
formation,
which
must
also
be
evaluated
if
this
option
is
used.

 
Investigate
potential
upstream
chemical
interferences
that
may
be
from
a
process
failure
or
upset.
For
example,
if
the
ozone
quenching
system
failed,
the
UVT
would
decrease.

If
the
above
strategies
cannot
be
implemented
or
are
not
successful
in
reducing
the
low
UVT,
the
UV
manufacturer
or
UV
installation
designer
should
be
contacted
to
investigate
the
problem
further.
The
utility
may
consider
shutting
down
the
WTP
or
activating
any
backup
disinfection
capacity
until
the
UVT
is
within
the
validated
limits.
A
decision
tree
that
summarizes
the
approach
for
troubleshooting
low
UVT
is
shown
on
Figure
5.5.
Anytime
the
UVT
is
lower
than
the
validated
limit,
it
should
be
recorded
as
off­
specification
even
if
it
does
not
occur
at
precisely
the
time
(
e.
g.,
4­
hour
interval)
that
recording
is
completed.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
35
June
2003
Figure
5.5
High
UV
Absorbance
Decision
Chart
Resample
and
Reanalyze
Low
UVT
Is
on­
line
UVT
monitor
used?
Check
grab
UVT
sample
with
bench­
top
spectrophotometer
Yes
No
Is
repeat
UVT
similar
and
low?
Evaluate
sampling
and
analytical
procedures
No
Check
spectrophotometer
with
phthalate
standard
Yes
Is
spectrophotometer
response
acceptable?
Is
grab
UVT
sample
low
after
repeat
sampling?

Recalibrate
or
repair
bench­
top
spectrophotometer
No
Is
the
UV
system
operating
off­
specification
because
of
low
UVT?
Recalibrate
or
repair
on­
line
UVT
monitor
No
Yes
Yes
Continue
operation
No
Can
UVT
be
increased
through
WTP
operation
changes?
Yes
Is
UVT
below
the
validation
limit?

Is
UV
intensity
or
calculate
dose
below
validation
limits?
Continue
operation
Yes
No
No
Contact
manufacturer
or
UV
system
designer
to
investigate
this
issue
further.
Consider
WTP
shutdown
Yes
See
low
UV
intensity
and
calculated
dose
decision
tree
(
Figure
5.3)
Yes
No
Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
36
June
2003
5.6.3
Rapid
Flow
Increase
or
High
Flow
It
may
be
possible
to
compensate
for
increased
flow
(
depending
on
validation
data)
by
completing
one
or
more
of
the
following
actions:

 
Increasing
the
output
of
the
UV
lamps
 
Using
additional
lamps
or
banks
of
lamps
 
Using
additional
UV
reactors
The
success
of
these
strategies
depends
on
the
magnitude
of
the
flowrate
increase
and
the
type
and
configuration
of
the
UV
reactors.
These
changes
should
occur
automatically
for
reactors
that
are
controlled
using
PLCs.

If
the
measured
flowrate
is
higher
than
the
validated
limits
and
cannot
be
reduced,
the
flowmeter
and/
or
differential
pressure
meter
(
if
used)
should
be
evaluated
to
determine
if
it
is
functioning
properly.
Instrument
error
can
be
assessed
by
comparing
signals
from
individual
flowmeters
or
differential
pressure
devices
to
anticipated
values
based
on
facility
flowrate
and
historic
operating
data.
Alternatively,
a
calibrated
clamp­
on
type
flowmeter
may
be
used
to
verify
flowrates.
If
the
flowmeter
is
not
operating
properly,
it
should
be
repaired
or
replaced.
If
flow
monitoring
devices
appear
to
be
functioning
properly,
valve
position
or
blockage
may
be
the
cause
of
unequal
flow
distribution
and
should
be
evaluated.

If
the
flow
is
below
the
validated
limits,
one
UV
reactor
should
be
taken
off­
line,
which
will
transfer
that
flow
to
the
other
energized
reactors.
This
change
in
operation
should
result
in
the
UV
reactors
being
within
the
validated
flow
range.
Anytime
the
flow
is
lower
or
higher
than
the
validated
limit,
it
should
be
recorded
as
off­
specification
even
if
it
does
not
occur
at
precisely
the
time
(
e.
g.,
4­
hour
interval)
that
the
recording
is
completed.

5.6.4
Unreliable
UV
Intensity
Sensor
Readings
Consistent
UV
intensity
sensor
readings
are
important
to
ensure
that
the
UV
reactors
are
operating
within
the
validated
limits.
Unreliable
UV
intensity
sensor
readings
can
be
described
by
one
or
more
of
the
following
behaviors:

 
Calibration
checks
outside
of
uncertainty
specified
in
the
validation
testing
 
Random
fluctuations
of
greater
than
25
percent
 
Biased
readings
(
UV
intensity
sensor
reading
is
offset
from
the
reference
sensor
readings
by
a
certain
value)

Unreliable
UV
intensity
sensor
readings
can
be
due
to
UV
intensity
sensor
malfunction,
condensation
in
the
sensor
or
between
the
sensor
and
sensor
window,
lamp
malfunction,
poor
grounding,
degradation
of
sensor
electronics,
or
electronic
short
circuits.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
37
June
2003
The
UV
intensity
sensor
and
lamp
electrical
cables
should
be
secured,
and
a
reference
or
standby
sensor
should
be
compared
to
the
duty
sensor
reading.
If
the
duty
sensor
is
found
to
be
defective
or
out
of
calibration,
it
should
be
sent
to
the
manufacturer
for
repair,
and
the
standby
sensor
used
in
its
place.

5.6.5
Power
Quality
Problems
UV
lamps
can
potentially
lose
their
arc
if
a
voltage
sag,
power
quality
anomaly,
or
a
power
interruption
occurs.
Voltage
sags
as
little
as
10
to
15
percent
from
the
nominal
voltage
for
as
few
as
2
to
5
cycles
can
cause
a
UV
lamp
to
lose
its
arc.
LP
lamps
generally
can
return
to
full
operating
status
within
15
seconds
after
power
is
restored.
LPHO
and
MP
lamps
will
need
to
be
re­
struck,
which
generally
requires
between
4
and
10
minutes
to
get
to
full
lamp
power,
to
restart.
LPHO
and
MP
lamps
are
affected
differently
from
power
losses
as
discussed
in
more
detail
in
section
3.1.3.3.

The
corrective
actions
for
short­
term
power
failures
(
e.
g.,
voltage
sag)
are
different
for
LPHO
and
MP
reactors.
LPHO
lamps
need
to
warm­
up
before
the
arc
can
be
struck,
and
MP
lamps
need
to
be
cooled
before
the
arc
can
be
struck.
Standby
MP
reactors
(
i.
e.,
not
in
operation
when
voltage
sag
occurred)
should
be
energized
instead
of
"
warm"
reactors
because
they
will
take
less
time
to
restore
operation
to
within
validated
limits
because
the
UV
lamps
do
not
have
to
cool
down
before
re­
striking.
However,
installations
using
LPHO
reactors
should
energize
their
"
warm"
reactors
(
i.
e.,
the
reactors
on­
line
when
the
voltage
sag
occurred)
instead
of
standby
LPHO
reactors
because
the
UV
lamp
warm­
up
time
will
be
less
compared
to
a
cold
LPHO
reactor.

For
long­
term
power
failure
(
e.
g.,
>
5
minutes)
without
a
UPS
system,
the
UV
reactors
should
be
powered
by
the
backup
generator
until
power
is
restored.
When
power
is
restored,
the
shift
from
the
backup
generator
will
likely
cause
the
UV
lamps
to
lose
their
arc
again.

Given
the
restrictions
on
operation
outside
of
validated
limits
(
section
1.3.1.3),
the
utility
should
stop
water
flow
through
the
UV
reactors
when
the
lamps
are
not
operating.
Also,
utilities
should
consider
installing
a
UPS
if
power
quality
problems
are
frequent
because
a
standby
generator
alone
may
not
adequately
alleviate
frequent,
off­
specification
flows
due
to
power
quality
problems.
A
UPS
system
delivers
consistent,
continuous
power
even
when
power
problems
occur.

5.7
Staffing
Issues
In
order
to
provide
consistent
and
reliable
operation
of
UV
reactors,
the
utility
needs
to
have
appropriate
staffing,
training,
and
safety
measures
in
place.
This
section
discusses
these
issues.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
38
June
2003
5.7.1
Staffing
Levels
During
start­
up
operation,
a
UV
reactor
will
need
more
operator
attention
to
assist
with
functional
and
performance
testing
and
to
establish
site­
specific
O&
M
procedures
(
described
in
section
5.1.4).
However,
a
typical
UV
installation
needs
little
operator
attention
during
normal
operation,
depending
on
the
level
of
automation.
Generally,
UV
installations
use
PLCs
to
monitor
operating
parameters,
control
the
UV
reactor,
and
generate
alarms.
Increased
automation
(
e.
g.,
remote
monitoring
capability)
may
be
incorporated
to
further
reduce
operator
requirements.
Table
5.14
describes
how
various
site­
specific
factors
affect
staffing
needs
for
a
UV
installation.

Table
5.14
Factors
Impacting
Staffing
Needs
Factor
Impact
on
Staffing
Type
of
UV
reactor
LP
and
LPHO
reactors
may
need
more
maintenance
compared
to
a
MP
reactor
because
they
have
more
lamps
and
usually
employ
OCC
cleaning.
However,
MP
lamps
will
probably
need
to
be
replaced
more
often
than
LP
lamps.

Instrumentation
and
control
strategy
More
automated
control
strategies
will
result
in
lower
staffing
levels
due
to
enhanced
remote
operation
and
monitoring
capability.

Water
quality
Sleeve
fouling
and
cleaning
frequency
is
affected
by
water
quality
and
the
design
of
the
UV
reactor.
These
in
turn
impact
the
staffing
needs
for
manual
cleaning
for
OCC
systems
and
for
maintaining
the
OMC
system.

5.7.2
Training
Training
is
necessary
for
all
personnel
who
are
associated
with
the
UV
installation,
including
operators,
maintenance
workers,
instrumentation
technicians,
electricians,
laboratory
staff,
custodial
staff,
engineers,
and
administrators.
The
training
program
should
incorporate
any
State
requirements
and
should
emphasize
both
normal
and
emergency
operating
procedures,
safety
issues,
process
control
and
alarm
conditions,
validated
operation,
and
response
to
deviations.

The
UV
manufacturer
and
UV
installation
designer
should
provide
training
on
the
UV
reactors,
UV
installation
design,
and
operation
and
maintenance
activities.
It
is
recommended
that
training
include
both
classroom
instruction
and
field
training.
In
addition,
actively
involving
the
operating
staff
during
start­
up
will
provide
another
opportunity
to
reinforce
classroom
instructions.
Continued
training
should
be
provided
when
new
employees
are
hired
or
when
a
process
or
control
alteration
is
made.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
39
June
2003
5.7.3
Safety
Issues
The
Office
of
Safety
and
Health
Administration
(
OSHA)
issues
regulations
and
guidance
to
support
operator
safety
in
the
workplace.
There
may
also
be
specific
safety
requirements
imposed
by
the
State.
In
addition
to
the
standards
and
procedures
established
for
WTP
operations,
the
following
safety
issues
pertain
specifically
to
UV
reactors:

 
UV
light
exposure
 
Electrical
safety
 
Burns
from
hot
lamps
or
equipment
 
Abrasions
or
cuts
from
broken
lamps
 
Potential
exposure
to
mercury
from
broken
lamps
­
Over­
exposure
to
UV
light
can
cause
eye
injury
and
skin
damage.

Threshold
Limit
Values
(
TLVs)
are
issued
biannually
by
the
American
Conference
of
Governmental
Industrial
Hygienists
(
ACGIH).
The
TLVs
for
UV
radiation
apply
to
occupational
exposure
to
UV
incident
on
the
skin
or
eye.
The
recommended
TLVs
depend
on
the
lamp
wavelengths
emitted
and
the
irradiance
(
mW/
cm2);
the
utility
can
determine
the
appropriate
TLV
for
their
UV
reactors,
using
the
TLVs
for
Chemical
Substances
and
Physical
Agents
and
Biological
Exposure
Indices
(
ACGIH
2002).
These
values
are
not
enforceable
standards
but
should
be
considered
when
establishing
operational
procedures.
To
limit
or
prevent
operator
exposure
to
the
UV
light,
UV
reactors
should
have
interlocks
that
deactivate
the
lamps
when
reactors
are
accessed.
Viewing
ports,
if
provided,
should
be
fitted
with
UV
filtering
windows,
or
operators
should
wear
a
UV
resistant
face
shield
when
looking
at
lamps
or
the
reaction
chamber.
In
addition,
warning
signs
should
be
placed
to
minimize
the
danger
of
exposure.

To
reduce
the
risk
of
electrical
shock,
the
main
electrical
supply
to
the
UV
reactors
should
be
disconnected
and
the
operator
should
wait
at
least
5
minutes
for
the
lamps
to
cool
down
and
energy
to
dissipate
before
any
maintenance
is
performed.
All
safety
and
operation
precautions
required
by
the
National
Electric
Code
(
NEC),
OSHA,
local
electric
codes,
and
the
UV
manufacturer
should
be
followed
and
include
the
following
precautions:

 
Proper
grounding
 
Lockout,
tag­
out
procedures
 
Use
of
proper
electrical
insulators
 
Installation
of
safety
cut­
off
switches
The
ballasts
and
the
reactor
chamber
can
also
become
hot
during
operation.
The
temperature
of
these
components
should
be
assessed
prior
to
contact.

Proposal
Draft
5.
Start­
up
and
Operation
of
UV
Installations
UV
Disinfection
Guidance
Manual
5­
40
June
2003
Broken
lamps
pose
two
potential
safety
hazards.
The
lamps
and
sleeves
are
constructed
of
quartz
tubing,
which
can
fracture
and
cause
serious
cuts
or
injury.
In
addition,
broken
lamps
may
release
mercury.
Operators
should
be
trained
in
proper
mercury
cleanup
and
disposal
procedures
to
prevent
mercury
inhalation
or
absorption
through
the
skin.
Appendix
N
discusses
potential
mercury
cleanup
procedures.

Proposal
Draft
