Appendix
K.
Preliminary
Engineering
Report
Critical
design
and
implementation
issues
need
to
be
resolved
early
in
the
planning
phase
of
a
UV
disinfection
facility.
The
purpose
of
a
preliminary
engineering
report
is
1)
to
provide
conceptual
level
layouts
and
preliminary
cost
estimates
for
implementation
of
UV
disinfection
at
the
water
treatment
plant
(
WTP),
and
2)
to
recommend
an
implementation
plan
for
UV
installation
design
and
construction.
Specific
components
of
the
preliminary
engineering
analysis
are
listed
below:

 
Identification
of
UV
reactor
design
criteria
and
implementation
issues
 
Evaluation
of
UV
reactor
alternatives
and
potential
locations
for
the
proposed
installations
in
the
plant
treatment
train
 
Determination
of
the
hydraulic
characteristics
of
the
UV
reactor
and
incorporate
it
into
the
hydraulic
model
of
the
plant
 
Development
of
estimates
for
capital,
operational,
and
life­
cycle
costs
for
each
alternative
 
Comparison
of
feasible
alternatives
and
development
of
implementation
recommendations
This
appendix
presents
an
example
of
a
preliminary
engineering
report
(
PER)
for
retrofitting
a
UV
disinfection
facility
into
an
existing
WTP.
The
basic
elements
involved
in
the
planning
phase
of
the
UV
installation
are
discussed
in
this
report.
The
format
and
content
of
a
site­
specific
PER
should
be
coordinated
with
the
State.
This
example
report
is
based
largely
on
a
predesign
report
prepared
for
North
Shore
Water
Commission,
Wisconsin
(
Carollo
Engineers
2001).

Chapter
3
of
the
Guidance
Manual
presents
a
detailed
discussion
of
UV
installation
planning
and
design
principles.
A
flowchart
depicting
the
planning
and
design
process
is
included
in
Figure
3.1.
Table
K.
1
presents
a
correlation
between
the
flowchart
elements
discussed
in
Chapter
3
and
respective
sections
in
this
appendix.

Table
K.
1
Elements
of
the
Planning
and
Design
Process
(
Ref.
Figure
3.1)

Element
Chapter
3
Section
Appendix
K
Section(
s)
Define
UV
disinfection
goals
3.1.1
K.
1,
K.
2
Identify
potential
retrofit
locations
3.1.2
K.
5
Determine
design
parameters
3.1.3
K.
2.2
Evaluate
potential
UV
reactors
3.1.4
K.
3
Evaluate
operational
and
control
strategies
3.1.5
K.
4
Evaluate
hydraulic
profile
and
site
layouts
3.1.6
K.
5
Compare
retrofit
options
and
costs;
select
retrofit
locations
3.1.7
K.
6,
K.
7,
K.
8
UV
Disinfection
Guidance
Manual
K­
1
June
2003
Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
2
June
2003
K.
1
Background
of
Example
WTP
A
20
million
gallon
per
day
(
mgd)
surface
WTP
is
used
as
an
example
in
this
appendix.
The
water
treatment
processes
employed
are
coagulation
and
sedimentation
pretreatment,
granular
media
(
anthracite
and
sand)
filtration
followed
by
chlorine
disinfection.
Since
the
plant
was
put
into
service
in
the
1960s,
water
quality
regulations
have
become
more
stringent.
In
addition,
there
are
growing
concerns
over
chlorine­
resistant
pathogens
(
e.
g.,
Cryptosporidium)
and
production
of
chlorinated
disinfection
byproducts
(
e.
g.,
trihalomethanes,
haloacetic
acids).
In
order
to
upgrade
the
facility
to
meet
current
and
future
regulations
and
health
concerns,
several
research
studies
have
been
performed
involving
the
use
of
ozone,
membranes,
and
UV
disinfection.
From
the
results
of
those
studies,
it
was
concluded
that
the
most
feasible
and
cost
effective
solution
to
achieve
disinfection
of
chlorine­
resistant
pathogens
was
to
add
UV
disinfection
to
the
current
treatment
train.

The
following
are
the
general
performance
goals
of
the
UV
installation
for
the
example
WTP:

 
Provide
2­
log
Cryptosporidium
inactivation.

 
Provide
an
additional
disinfection
barrier
for
other
chlorine­
resistant
pathogens.

K.
2
UV
Disinfection
Criteria
This
section
includes
general
information
regarding
the
optimal
application
point
for
UV
disinfection
at
the
WTP
and
design
considerations
for
implementation.

K.
2.1
Application
Point
and
UV
Transmittance
One
of
the
important
parameters
controlling
UV
installation
design
is
the
UV
transmittance
(
UVT)
of
the
water
to
be
treated
(
section
3.1.3.1).
The
lower
the
UVT,
the
greater
the
UV
intensity
is
needed
to
provide
a
given
UV
dose
at
a
given
flowrate.
UVT
typically
varies
with
source
water,
seasonally,
and
through
the
treatment
processes.
Consequently,
a
thorough
UVT
analysis
was
completed
during
development
of
design
criteria.

K.
2.1.1
Point
of
Application
for
UV
Disinfection
In
keeping
with
the
content
of
this
guidance
manual,
the
UV
disinfection
alternatives
assessed
for
the
WTP
were
limited
to
applications
after
filtration.
Based
on
statistical
results
of
the
filtered
water
ultraviolet
absorbance
at
254
nanometers
(
A254)
data,
the
UV
reactors
are
sized
based
on
a
0.032/
cm
A254
(
93
percent
UVT;
10
mm
path
length;
light
at
254
nm),
which
is
the
99th
percentile
minimum
of
the
available
A254
data.

Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
3
June
2003
K.
2.1.2
Treatment
Chemical
Impact
on
Absorbance
Some
chemicals
used
in
water
treatment
absorb
UV
light
and
hence,
can
influence
the
design
absorbance
value,
as
discussed
in
Chapter
3.1.2
of
the
manual.
Ferric
iron
and
permanganate
are
two
of
these,
and
are
used
at
the
example
WTP.
Ferric
iron
strongly
absorbs
UV
light;
however,
post­
filtration
iron
levels
are
generally
low.
Permanganate
absorbs
UV
light,
but
at
permanganate
levels
of
less
than
1
mg/
L,
which
is
typically
the
case
post­
filtration,
the
impact
is
not
significant.
Therefore,
for
the
PER,
chemical
A254
is
not
considered
to
influence
the
UV
design
criteria.

K.
2.1.3
Power
Quality
Impact
on
Absorbance
As
stated
in
section
3.1.3.3,
the
sensitivity
of
UV
reactors
to
power
fluctuations
make
electrical
power
supply
a
critical
component
of
the
UV
installation
planning
and
design.
Preliminary
pilot
testing
of
UV
reactors
over
the
course
of
a
year
at
this
site
did
not
indicate
any
problems
with
existing
water
utility
power
quality
for
the
UV
reactor's
operational
continuity.
Therefore,
for
this
PER,
power
quality
is
not
considered
to
negatively
impact
the
UV
installation
design.

K.
2.2
Inactivation
Goals
and
UV
Dose
The
goal
of
UV
disinfection
at
the
example
WTP
is
to
provide
inactivation
of
chlorineresistant
pathogens.
By
using
UV
disinfection,
the
general
goal
of
improving
public
health
protection
will
be
met
and
compliance
with
the
Long
Term
2
Enhanced
Surface
Water
Treatment
Rule
(
LT2ESWTR)
Cryptosporidium
inactivation
requirements
may
be
achieved,
if
needed
(
40
CFR
141.702).
(
Source
water
sampling
for
"
Bin"
determination
has
not
yet
been
completed
at
this
example
WTP,
so
the
level
of
additional
credit
needed
in
the
future
is
unknown.)
UV
disinfection
credit
will
also
be
available
for
Giardia
and
virus
inactivation.
This
additional
UV
disinfection
credit
will
most
likely
reduce
chlorine
disinfection
requirements,
and
hence,
reduce
disinfection
by­
product
formation.

The
desired
UV
dose
(
or
validated
reduction
equivalent
dose
[
RED])
depends
on
the
disinfection
strategy
of
the
individual
UV
installation.
The
State,
utility,
and
designer
must
decide
the
log
inactivation
requirements
for
a
target
pathogen.
Once
this
information
is
known,
the
UV
dose
can
be
established
(
section
3.1.1).
For
this
PER
example,
a
UV
dose
of
40
mJ/
cm2
is
recommended
to
achieve
2­
log
Cryptosporidium
inactivation
and
is
used
for
UV
disinfection
pre­
design
purposes.

A
12­
month
pilot
study
was
conducted
to
assess
the
long­
term
disinfection
efficiency
and
operation
and
maintenance
(
O&
M)
issues.
The
study
results
indicated
that
lamp
fouling
and
power
quality
issues
should
not
be
a
concern
for
the
facility
(
Mackey
et
al.
2001).

Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
4
June
2003
K.
3
UV
Installation
Equipment
General
information
on
UV
reactors
and
the
types
of
reactor
configurations
used
for
water
treatment
is
provided
in
this
section.

K.
3.1
UV
Lamp
Types
For
the
flowrates
associated
with
the
WTP
applications
in
this
example,
the
number
of
lamps
needed
for
a
low­
pressure
(
LP)
reactor
would
be
excessive,
so
consideration
is
limited
to
low
pressure
high
output
(
LPHO)
and
medium
pressure
(
MP)
lamps.
The
general
relative
characteristics
of
each
of
these
lamp
types
are
listed
in
Table
K.
2.
The
ratio
of
number
of
lamps
needed
to
achieve
equivalent
RED
for
LPHO
lamps
as
compared
to
MP
lamps
is
on
the
order
of
6:
1.

Table
K.
2
Relative
Characteristics
of
LPHO
and
MP
Lamps
LPHO
MP
Lamp
Power
Output
Low
High
Power
Efficiency
High
Low
Number
of
Lamps
Needed
High
Low
Operating
Temperature
(
°
C)
130
 
200
600
­
900
Typical
Lamp
Life
(
hours)
8,000
­
12,000
3,000
­
8,000
K.
3.2
UV
Reactor
Configuration
UV
installations
can
be
designed
around
open­
channel
or
closed­
vessel
configurations.
In
keeping
with
the
content
of
this
guidance
manual
and
the
general
trend
of
the
drinking
water
industry,
the
conceptual
designs
developed
herein
are
limited
to
MP
and
LPHO
closed­
vessel
reactors.

K.
4
UV
Reactor
Description
This
section
contains
information
on
the
UV
reactor
design
criteria
for
disinfection
at
the
WTP.

K.
4.1
General
UV
Reactor
Description
Each
UV
reactor
for
the
WTP
should
include
appropriate
control
and
electrical
cabinets
and
an
off­
line
chemical
cleaning
(
OCC)
system
or
an
on­
line
mechanical
cleaning
(
OMC)
system
for
the
lamps.
The
cleaning
systems
should
allow
for
the
removal
of
organic
and
inorganic
foulants
that
have
accumulated
on
the
surfaces
of
the
lamp
sleeves.

Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
5
June
2003
K.
4.2
Process
Control
For
disinfection
of
drinking
water,
the
ability
of
the
UV
reactor
to
deliver
the
design
RED
of
40
mJ/
cm2
depends
on
the
flowrate,
feed
water
UVT,
and
UV
intensity.
UV
intensity
is
subject
to
lamp
aging,
lamp
sleeve
cleanliness,
and
water
quality
(
mainly
water
UVT).
The
UV
reactor
should
be
designed
to
deliver
the
appropriate
dose
of
UV
light
to
the
process
flow
based
on
predetermined
maximum
flowrate
and
minimum
water
quality
parameter
setpoints
with
an
appropriate
factor
of
safety
(
see
Chapter
3).

UV
intensity
sensors
in
each
UV
reactor
should
provide
continuous
performance
verification
of
the
reactor
during
operation.
In
case
of
lamp
failure,
the
UV
reactor
programmable
logic
controller
(
PLC)
should
be
programmed
to
either
replace
one
row
of
lamps
with
another
row
that
was
off,
or
turn
the
reactor
off
after
replacing
it
with
a
stand­
by
reactor.
(
Note
that
Alternative
1,
described
in
section
K.
5.1,
involves
placing
a
UV
reactor
on
each
filter
effluent
pipe,
therefore
stand­
by
reactors
for
individual
filter
installations
are
not
provided).
The
failed
lamp
can
then
be
replaced
with
minimal
interruption
of
UV
reactor
operation.

In
case
of
incorrect
operation
of
lamps
or
low
level
of
UV
intensity,
the
PLC
should
display
a
warning
to
indicate
to
the
operator
that
cleaning
of
the
reactor
should
be
performed.
The
operator
initiates
the
cleaning
of
any
reactor
through
the
local
human
machine
interface
(
HMI).
After
selection,
the
UV
reactor
PLC
turns
on
the
stand­
by
reactor.
Then,
the
PLC
closes
the
inlet
and
outlet
valves
and
isolates
the
reactor
to
be
cleaned.

K.
4.3
Expected
UV
Reactor
Maintenance
Although
maintenance
methods
are
installation
and
site­
specific,
some
general
maintenance
tasks
have
been
developed
and
are
briefly
described
in
this
section.
As
the
UV
reactor
represents
a
critical
disinfection
process,
preventative
maintenance
should
be
carried
out
on
a
routine
basis
to
ensure
that
UV
reactors
reliably
provide
the
specified
dose
(
40
mJ/
cm2).
Inadequate
cleaning
is
a
common
cause
of
underdosing
in
UV
reactors.
The
lamp
sleeves
should
be
cleaned
regularly
by
OMC
or
periodic
OCC,
and
manually
cleaned
periodically
to
supplement
automatic
cleaning.
The
cleaning
frequency
is
dependent
on
the
water
quality.
Chemical
cleaning
is
most
commonly
done
with
dilute
citric
or
phosphoric
acid.

The
effective
life
of
the
UV
lamps
depends
on
the
minimum
UV
dose.
The
UV
lamps
should
be
replaced
either
at
the
end
of
their
expected
lifetime
or
following
failure.
Generally,
UV
lamps
are
replaced
when
the
intensity
has
dropped
to
70
percent
of
the
original
new­
lamp
intensity
(
following
cleaning
of
the
chamber).
This
typically
occurs
after
about
8,000
to
12,000
hours
(
approximately
300
to
500
days)
of
operation
for
LPHO
lamps
and
about
3,000
to
8,000
hours
(
approximately
100
to
300
days)
for
MP
lamps.
The
front
panel
of
the
enclosures
indicates
the
cumulative
hours
each
lamp
has
operated.
The
lamp
run
time
display
will
facilitate
monitoring
of
lamp
replacement
needs.

Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
6
June
2003
K.
4.4
Power
Needs
UV
reactor
power
needs
to
vary
depending
on
the
type
of
equipment
that
is
installed
and
UVT
of
the
water
being
disinfected.
The
LPHO
reactors
used
for
this
pre­
design
have
power
requirements
of
approximately
20
kW
for
treating
20
mgd.
The
MP
UV
reactors
require
about
130
kW
for
treating
20
mgd.
As
indicated
in
Table
K.
3
through
K.
5,
additional
power
would
be
necessary
to
allow
for
future
expansion
of
the
UV
facilities.

K.
5
Site
Plans
and
Facility
Layouts
The
preferred
process
location
for
a
UV
installation
at
a
WTP
is
downstream
of
the
filters
and
upstream
of
the
high­
service
pumps
(
section
3.1.2).
At
the
example
WTP,
there
are
three
viable
alternatives
for
the
UV
installation
downstream
of
the
filters:

 
Alternative
1
 
Filter
Gallery
 
Alternative
2
 
Existing
Chemical
Room
 
Alternative
3
 
New
Building
There
are
eight
granular
media
filters
at
the
WTP.
Alternative
1
involves
placing
one
UV
reactor
on
the
effluent
pipe
of
each
filter
in
the
filter
gallery
between
the
filters
and
clearwell.
Alternative
2
is
to
construct
the
UV
installation
in
a
chemical
room
in
the
WTP
between
the
low
service
pumps
and
the
reservoir.
Alternative
3
involves
constructing
a
new
building
between
the
low
service
pumps
and
the
reservoir
to
house
the
UV
reactors.
Figure
K.
1
presents
a
portion
of
the
plants
hydraulic
profile
and
indicates
the
vertical
locations
of
the
three
viable
alternatives.

The
construction
requirements
and
preliminary
drawings
for
each
alternative
are
illustrated
and
described
in
the
following
section,
along
with
preliminary
design
criteria.
Costs
for
the
three
alternatives
are
also
compared.
The
preliminary
site­
specific
design
criteria
are
provided
for
example
purposes
only.
Application­
specific
design
criteria
should
be
provided
by
the
UV
manufacturer
for
each
individual
UV
disinfection
implementation
project
on
a
case­
bycase
basis.

K.
5.1
Alternative
1
­
Filter
Gallery
In
Alternative
1,
one
3
mgd
UV
reactor
is
installed
on
the
discharge
pipe
of
each
of
the
eight
filters,
as
shown
in
Figure
K.
2.
The
UV
reactors
would
be
installed
below
the
hydraulic
grade­
line
(
HGL)
of
the
existing
clearwell
to
ensure
constant
submergence
(
section
3.1.6.1).
Flow
through
the
UV
reactors
is
by
gravity
from
the
filters
into
the
clearwell.
During
filter
backwashing
and
filter­
to­
waste
cycles,
valves
located
at
the
influent
of
each
UV
reactor
can
be
closed
to
keep
the
reactor
flooded
while
it
is
taken
off­
line.

Compared
to
the
other
alternatives,
construction
of
Alternative
1
is
the
simplest.
Construction
would
include
lowering
the
level
in
the
clearwell
to
below
the
filter
discharge
Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
7
June
2003
pipes,
then
taking
each
filter
off­
line
individually
to
install
the
new
piping,
valves,
and
UV
reactor.
This
would
preclude
significant
disruption
of
plant
operation
during
construction.

Figure
K.
1
Portion
of
the
WTP
Hydraulic
Profile
and
Alternative
Locations
for
UV
Implementation
(
Carollo
Engineers
2001)

Low
Service
Pump
Filters
Reservoir
Clearwell
Elev.
72.0'

Elev.
56.5'

Elev.
69.0'
Alternative
2
Alternative
3
Alternative
1
Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
8
June
2003
Figure
K.
2
Alternative
1
­
Filter
Gallery
(
Carollo
Engineers
2001)

Note:
UV
reactor
is
installed
on
discharge
pipe
of
each
filter.
Valve
on
UV
reactor
influent
is
closed
to
maintain
water
in
the
UV
reactor
during
filter
backwashing.

Some
concerns
associated
with
installing
UV
reactors
in
the
filter
gallery
are
space
constraints,
climate
control,
and
impact
on
filtration.
In
the
design,
care
would
need
to
be
taken
to
allow
enough
space
for
maintenance
and
construction
of
the
UV
installation.
Although
the
UV
installation
in
this
example
will
fit
into
the
existing
filter
gallery
of
the
WTP,
generally
there
is
little
room
to
work
in
these
locations.
In
addition,
depending
on
the
location
of
the
WTP,
humidity
and
pipe
sweating
in
this
space
might
be
a
concern.
However,
there
are
protective
climate
controlled
enclosures
available
for
these
conditions,
though
they
add
to
cost
and
maintenance
needs.
A
detailed
hydraulic
analysis
of
the
filters
would
need
to
be
completed
prior
to
designing
a
UV
installation
on
the
filter
effluent
piping.
There
must
be
adequate
head
available
from
the
filters
to
the
clearwell
to
allow
for
the
addition
of
a
UV
reactor
that
will
not
Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
9
June
2003
adversely
affect
the
filter
performance.
Furthermore,
in
the
event
of
a
UV
reactor
shutdown,
the
filter
associated
with
that
UV
reactor
shutdown
would
also
need
to
be
removed
from
service.

The
preliminary
design
criteria
for
Alternative
1
are
provided
in
Table
K.
3.
Due
to
the
large
size
of
the
LPHO
reactors,
they
will
not
fit
into
the
filter
gallery.
Therefore,
only
the
MP
reactors
are
considered
for
Alternative
1.

Table
K.
3
Preliminary
Design
Criteria
 
Alternative
1
 
MP
UV
Reactors
Description
Unit
Criteria
Current
Future
Treatment
plant
capacities
Flowrate
mgd
20
40
Water
quality
UVT
In
a
10mm
quartz
cell
@
254
nm
%
UVT
93
93
Ultraviolet
reactors
Type
of
reactors:
medium­
pressure
Number
of
reactors
No
8
16
Number
of
banks
per
reactor
No.
2
2
Number
of
lamps
per
bank
No.
4
4
Total
number
of
lamps
per
reactor
No.
8
8
Input
power
per
lamp
W
2000
2000
Total
operating
electric
load
kW
128
256
Total
installed
electric
load
kW
128
256
Headloss
through
reactor
(
at
current
and
future
flowrates)
Inches
12
36
Approximate
dimensions
of
each
UV
reactor
Length
Inches
22
50
Width
Inches
36
36
Height
Inches
26
26
Flanges
diameter
Inches
12
12
As
stated
previously,
the
eight
UV
reactors
listed
in
Table
K.
3
are
designed
for
a
maximum
capacity
of
3
mgd
each.
This
design
is
based
on
the
assumption
that
one
UV
reactor
would
be
taken
off­
line
periodically
during
a
filter
backwash
cycle.
The
WTP
must
be
able
to
treat
20
mgd
with
one
filter
out
of
service,
so
the
remaining
seven
UV
reactors
would
need
to
be
able
to
disinfect
the
maximum
plant
flow.
This
arrangement
also
provides
reactor
redundancy.
If
one
UV
reactor
were
taken
out
of
service,
the
associated
filter
would
also
be
taken
out
of
service.

In
this
example,
future
plant
expansions
needed
to
be
taken
into
account.
For
the
present
analysis,
provisions
are
made
so
that
future
expansion
of
the
UV
installation
to
an
ultimate
flow
of
40
mgd
will
be
possible.
If
the
filter
capacities
can
be
expanded
to
40
mgd,
the
UV
installation
expansion
will
necessitate
placing
two
UV
reactors
(
16
total)
in
series
along
each
filter
effluent
pipe.
During
the
construction
of
the
initial
design
for
20
mgd,
adequate
space
and
Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
10
June
2003
mechanical
layouts
would
be
provided
for
the
addition
of
a
second
UV
reactor.
For
example,
a
section
of
pipe
located
at
the
outlet
of
the
UV
reactor
for
the
20
mgd
design
could
be
designed
for
easy
removal
and
installation
of
a
second
UV
reactor.

K.
5.2
Alternative
2
­
Existing
Chemical
Room
Alternative
2
consists
of
installing
three
10
mgd
reactors
(
2
operational
+
1
stand­
by)
in
an
existing
chemical
room
in
the
WTP,
as
shown
in
Figure
K.
3.
This
alternative
necessitates
placing
the
UV
reactors
above
the
existing
HGL
of
the
plant.
The
low­
lift
clearwell
pumps
provide
the
head
through
the
UV
reactors.
(
It
is
generally
more
advantageous
to
place
the
UV
reactors
below
the
HGL.
However,
due
to
the
space
constraints
at
the
example
WTP,
and
to
provide
an
example
of
issues
that
may
arise
during
design,
this
option
is
discussed.)

Figure
K.
3
Alternative
2
­
Existing
Chemical
Room
with
UV
Reactors
(
Carollo
Engineers
2001)

In
theory,
an
outlet
weir
structure
would
be
a
viable
option
to
raise
the
HGL
of
the
plant
to
ensure
constant
submergence
of
the
UV
reactors
(
section
3.1.6.1).
In
this
case,
there
is
not
enough
space
for
such
a
structure.
To
ensure
constant
submergence
of
the
UV
reactors,
a
vertical
pipe
at
the
outlet
header
would
maintain
the
water
level
at
an
elevation
above
the
top
of
Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
11
June
2003
the
UV
reactors.
Air
vacuum
valves
would
be
installed
on
the
inlet
and
the
effluent
vertical
pipe
to
counteract
siphon
effects
on
the
UV
reactors.
The
discharge
from
the
effluent
header
would
then
flow
by
gravity
to
the
reservoirs.
The
hydraulic
design
of
the
inlet
and
outlet
channels
provides
a
continuous
equal
flow
split
between
the
reactors
(
section
3.3.1.2).

The
construction
needs
of
Alternative
2
are
more
difficult
than
Alternative
1.
A
36­
inch
finished
water
pipe
from
the
low
head
pumps
would
need
to
be
taken
out­
of­
service
long
enough
to
cut
the
pipe
and
tie­
in
a
new
section
with
fittings
and
valves
for
connection
of
the
new
UV
reactors.
The
existing
equipment
would
need
to
be
moved
to
alternate
locations
to
accommodate
the
new
large
piping
and
UV
reactors,
and
the
floor
would
need
to
be
cut
to
provide
clearance
for
the
piping
to
and
from
the
lower
floor.

Other
issues
with
Alternative
2
are
that
space
constraints
in
the
chemical
room,
possible
structural
upgrades
of
the
building,
and
raising
the
HGL
of
the
plant.
To
allow
for
adequate
space
for
maintenance,
piping,
and
instrumentation,
the
existing
chemical
equipment
in
the
room
would
need
to
be
removed
and
reinstalled
elsewhere
in
the
plant.
Since
this
is
a
second
level
room,
a
detailed
structural
analysis
would
need
to
be
completed
to
ensure
the
floor
is
able
to
withstand
the
load
of
the
UV
installation.
If
structural
upgrades
are
needed,
they
could
prove
to
be
expensive
and
difficult
to
design
and
construct.
Installing
the
UV
reactors
in
the
second
level
room
above
the
HGL
would
significantly
increase
the
total
dynamic
head
(
TDH)
placed
on
the
low­
lift
clearwell
pumps.
Therefore,
pump
upgrades
would
be
necessary
to
overcome
the
additional
headloss
of
the
UV
installation.

K.
5.3
Alternative
3
­
New
Building
Alternative
3
includes
constructing
a
new
building
to
house
three
10
mgd
UV
reactors
(
2
operational
for
20
mgd
+
1
standby)
and
related
equipment.
The
building
layout
and
UV
design
for
Alternative
3
is
presented
in
Figures
K.
4
and
K.
5.
The
36­
inch
finished
water
line
from
the
low­
service
clearwell
pumps
would
be
modified
to
provide
flow
to
the
UV
reactors
in
the
new
building.
The
new
facility
would
include
a
two­
level
structure
to
house
mechanical
and
electrical
equipment
and
large
diameter
piping
to
convey
the
filtered
water
through
the
UV
reactors.
The
UV
reactors
would
be
installed
in
the
basement
of
the
new
building
below
the
HGL
of
the
plant
to
ensure
constant
submergence.
The
hydraulic
design
of
the
inlet
and
outlet
channels
would
provide
an
equal
flow
split
between
the
reactors,
and
the
discharge
would
flow
under
pressure
to
the
reservoirs.

The
UV
building
design
and
mechanical
piping
shown
in
Figures
K.
4
and
K.
5
are
for
preliminary
design
and
cost
estimates
only.
If
this
option
were
selected,
the
building
size
and
configuration
for
this
alternative
would
need
to
be
evaluated
in
more
detail
during
the
design
phase
and
adjusted
as
necessary,
depending
on
the
final
UV
reactors
selected
to
be
used.

Construction
for
Alternative
3
would
be
the
most
involved
of
the
three
options
because
it
would
include
excavation
and
construction
of
a
new
building.
Besides
the
building
construction,
the
project
would
involve
tying
into
the
existing
36­
inch
finished
water
pipe
in
two
locations
below
grade,
and
modifying
site
amenities
such
as
pavement
and
landscaping.
In
addition,
the
new
building
would
also
need
new
power,
control,
and
security
systems
as
well
as
plumbing,
HVAC,
etc.

Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
12
June
2003
Figure
K.
4
Alternative
3
­
New
Building
with
UV
Reactors
 
Plan
View
(
Carollo
Engineers
2001)

Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
13
June
2003
Figure
K.
5
Alternative
3
­
New
Building
with
UV
Reactors
 
Section
View
(
Carollo
Engineers
2001)

The
preliminary
design
criteria
for
Alternatives
2
and
3
using
the
MP
UV
reactors
are
presented
in
Table
K.
4.

Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
14
June
2003
Table
K.
4
Preliminary
Design
Criteria
 
Alternatives
2
and
3
 
MP
UV
Reactors
Description
Unit
Criteria
Current
Future
Treatment
plant
capacities
Flowrate
mgd
20
40
Water
quality
UVT
in
a
10mm
quartz
cell
at
254
nm
%
UVT
93
93
Ultraviolet
reactors
Type
of
reactors:
medium­
pressure
Number
of
reactors
No.
(
Duty
+
Standby)
2+
1
2
+
1
Number
of
banks
per
reactor
No.
2
2
Number
of
lamps
per
bank
No.
8
8
Total
number
of
lamps
per
reactor
No.
16
16
Input
power
per
lamp
W
4000
4500
Total
operating
electric
load
kW
128
144
Total
installed
electric
load
kW
192
216
Headloss
through
reactor
(
at
current
&
future
flows)
Inches
10
48
Approximate
dimensions
of
each
UV
reactor
Length
Inches
48
48
Width
Inches
49
49
Height
Inches
41
41
Flanges
diameter
Inches
30
30
The
three
MP
UV
reactors
selected
have
design
capacities
of
10
mgd
each.
Three
10
mgd
UV
reactors
for
the
20
mgd
design
provide
one
stand­
by
reactor
in
the
event
of
a
malfunction,
cleaning,
or
maintenance
of
one
UV
reactor.

Note
that
for
expansion
of
the
UV
installation
using
the
MP
reactors
given
in
Table
K.
4,
the
size
and
number
of
UV
reactors
remains
constant.
In
order
to
provide
extra
lamp
intensity
to
meet
dose
requirements
at
the
ultimate
flow,
4000
W
lamps
would
replaced
with
4500
W
lamps.
(
Note
that
re­
validation
of
the
reactors
with
the
4500
W
lamps
would
be
required
(
40
CFR141.729
(
d)).

The
preliminary
design
criteria
using
the
LPHO
reactors
for
Alternatives
2
and
3
is
provided
in
Table
K.
5.

Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
15
June
2003
Table
K.
5
Preliminary
Design
Criteria
 
Alternatives
2
and
3
 
LPHO
UV
Reactors
Criteria
Description
Unit
C
urrent
Future
Treatment
plant
design
capacities
Plant
flowrate
mgd
20
40
Water
quality
UVT
in
a
10mm
quartz
cell
at
254
nm
%
UVT
93
93
Ultraviolet
reactors
Type
of
reactors:
Low­
Pressure
High­
Output
Number
of
reactors
No
(
Duty
+
Standby)
2+
1
2+
1
Number
of
rows
per
reactor(
1)
No.
5
9
Number
of
rows
with
lamps
installed
No.
4
8
Number
of
lamps
per
row
No.
12
12
Total
number
of
lamps
per
reactor
No.
48
96
Input
power
per
lamp
W
200
200
Total
operating
electric
load
kW
19.2
38.4
Total
installed
electric
load
kW
28.8
57.6
Headloss
through
reactor
(
at
current
&
future
flows)
Inches
24
35
Approximate
dimensions
of
each
UV
reactor
Length(
2)
Inches
110
144
Width
Inches
51
51
Height
Inches
100
100
Flanges
diameter
Inches
32
32
1
Preliminary
design
assumes
one
spare
row
in
addition
to
current
flow
demand
requirements
for
installation
of
lamps
in
the
future.
2
Length
varies
depending
on
the
number
of
rows
installed.

The
expansion
from
20
mgd
to
40
mgd
using
the
LPHO
UV
reactors
would
be
accomplished
by
adding
additional
rows
of
lamps
to
the
reactor.
The
UV
manufacturers
would
oversize
the
reactor
and
additional
rows
of
lamps
could
be
inserted
as
needed
for
increasing
flow
capacities.
However,
UV
reactor
validation
would
need
to
be
done
both
with
and
without
the
additional
rows
for
the
maximum
and
ultimate
flow
conditions.

Alternatively,
the
UV
installation
could
be
sized
to
allow
additional
UV
reactors
to
be
installed
for
expansion.
Initially,
three
10
mgd
reactors
would
be
installed
for
a
capacity
of
20
mgd
(
2
operational
and
1
standby).
Space
would
be
provided
to
install
two
additional
reactors
in
the
future
for
a
capacity
of
40
mgd
(
4
operational
and
1
standby).

These
examples
of
UV
installation
expansion
alternatives
provide
various
options
to
the
designer.
To
confidently
design
for
future
UV
installation
expansions,
it
will
be
critical
to
have
an
accurate
flow
projection
and
adequate
space
for
the
UV
installation
expansion.
In
addition
to
the
UV
reactors
needed
for
an
expansion,
mechanical
piping,
controls,
instrumentation,
and
wiring
would
need
to
be
considered
during
the
preliminary
engineering
phase.
Furthermore,
the
designer
should
work
closely
with
the
UV
manufacturer
to
decide
on
an
expansion
plan
that
has
been
proven
to
work
effectively
and
efficiently
for
the
specific
UV
installation
design.

Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
16
June
2003
K.
6
Preliminary
Capital
and
O&
M
Cost
Estimates
The
preliminary
capital,
operational,
and
maintenance
costs
for
each
alternative
are
summarized
in
this
section.
Estimated
costs
presented
for
Alternative
1
are
based
solely
on
the
MP
design.
The
LPHO
UV
reactors
used
for
comparison
here
would
not
fit
into
the
filter
gallery
and
so
was
not
considered
for
that
alternative.

K.
6.1
Capital
Cost
Estimate
Summary
The
estimated
capital
improvement
costs
for
each
alternative
are
summarized
in
Table
K.
6.
Total
project
cost
includes
UV
reactors,
construction
cost,
engineering
services,
and
a
20
percent
estimating
contingency.

The
cost
estimates
presented
for
Alternative
1
in
Table
K.
6
are
based
on
using
eight
MP
UV
reactors.
The
costs
presented
for
Alternatives
2
and
3
were
developed
around
using
three
MP
and
LPHO
UV
reactors.
The
equipment
cost
for
installing
the
MP
UV
reactors
for
all
three
Alternatives
is
higher
than
installing
LPHO
UV
reactors
for
Alternatives
2
and
3.
However,
due
to
the
relatively
simple
construction
details
associated
with
Alternative
1,
the
total
project
cost
is
considerably
lower
than
Alternatives
2
and
3,
which
require
significant
construction
provisions.
A
comparison
of
these
alternatives
is
provided
in
section
K.
7.

Table
K.
6
Preliminary
Capital
Cost
Estimates
UV
Reactor
Cost
Total
Project
Cost
Annualized
Capital
Cost1
Alternative
1­
Filter
Gallery
(
MP)
$
531,000
$
1,900,000
$
166,000
Alternative
2­
Chemical
Room
(
MP)
$
556,000
$
2,400,000
$
209,000
Alternative
2­
Chemical
Room
(
LPHO)
$
450,000
$
2,300,000
$
201,000
Alternative
3­
New
Building
(
MP)
$
556,000
$
2,800,000
$
244,000
Alternative
3­
New
Building
(
LPHO)
$
450,000
$
2,700,000
$
235,000
1
Annualized
costs
calculated
at
6
percent
interest
for
20
years.

K.
6.2
Preliminary
Operating
and
Maintenance
Costs
Table
K.
7
presents
a
summary
of
the
estimated
O&
M
costs
and
total
annualized
costs
for
each
alternative
(
four
MP
reactors
in
service
for
Alternative
1,
two
MP
and
LPHO
UV
reactors
are
in
service
for
Alternatives
2
and
3).
Detailed
O&
M
costs
for
an
average
flow
of
10
mgd
are
provided
for
each
alternative
and
UV
reactors
in
Tables
K.
8
and
K.
9.

Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
17
June
2003
Table
K.
7
Estimated
UV
Disinfection
Costs
Annul
O&
M
Cost1
Annualized
Capital
Cost2
Total
Annual
Cost
Annual
Difference3
Alternative
1
$
72,000
$
166,000
$
238,000
$
1,000
Alternative
2
(
MP)
$
111,000
$
209,000
$
320,000
$
83,000
Alternative
2
(
LPHO)
$
36,000
$
201,000
$
237,000
­
0­
Alternative
3
(
MP)
111,000
$
244,000
$
355,000
$
118,000
Alternative
3
(
LPHO)
$
36,000
$
235,000
$
271,000
$
34,000
1.
Costs
for
UV
intensity
sensor
calibrations,
lamp
sleeve,
ballast
and
sensor
replacement
are
not
included.
2.
Annualized
costs
calculated
at
6
percent
interest
for
20
years.
3.
Relative
to
the
least
expensive
alternative
(
Alternative
2­
LPHO).

Table
K.
8
MP
UV
Installation
O&
M
Cost
Estimates
for
Alternatives
1,
2
and
3
O&
M
Costs
Alt.
1
Alts.
2
&
3
Average
Plant
Flow
10
mgd
1
­
Power
Consumption
Annual
power
consumption
of
lamps
in
kWh
530,155
883,5921
Price
of
electricity
($/
kWh)
0.10
0.10
Annual
Expenses
($)
53,015
88,359
2
­
Consumables
Lamp
replacement
#
operating
32
$/
Lamp
500
(#
1)
600(#
2&
3)
17,500
21,000
#
replaced
/
yr
35
Annual
Expenses
($)
17,500
21,000
3
­
Labor
Lamps
#
replaced
/
yr
35
Time
(
hr)
8.8
15
min
/
lamp
Cleaning
1
clngs
/
yr
/
reactor
3
Time
(
hr)
9.0
3
hrs
/
cleaning
Total
Time
(
hr)
17.8
$/
hr
65
65
Annual
Expenses
($)
1,153
1,153
TOTAL
COSTS
1
­
Power
Consumption
53,015
88,359
2
­
Consumables
17,500
21,000
3
­
Labor
1,153
1,153
4
­
Chemicals
100
100
Total
Annual
Costs
71,770
110,610
COSTS
PER
MG
TREATED
Costs
per
MG
Treated
$/
MG
20.00
30.00
1
Alternative
1
utilizes
eight
smaller
UV
reactors
while
Alternatives
2
and
3
utilize
2
large
reactors.
At
an
average
flow
of
10
mgd,
the
large
UV
reactors
operate
at
the
lowest
possible
setting,
which
is
higher
than
required
at
10mgd.

Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
18
June
2003
Table
K.
9
LPHO
UV
Installation
O&
M
Cost
Estimates
for
Alternatives
2
and
3
O&
M
Costs
Average
Plant
Flow
10
mgd
1
­
Power
Consumption
Annual
power
consumption
of
lamps
in
kWh
168,303
Price
of
electricity
($/
kWh)
0.10
Annual
Expenses
($)
16,830
2
­
Consumables
Lamp
replacement
#
operating
96
$/
Lamp
150
12,600
#
replaced
/
yr
84
Annual
Expenses
($)
12,600
3
­
Labor
Lamps
#
replaced
/
yr
84
Time
(
hr)
21.0
15
min
/
lamp
Cleaning
1
clngs
/
yr
/
reactor
24
Time
(
hr)
72.0
3
hrs
/
cleaning
Total
Time
(
hr)
93.0
$/
hr
65
Annual
Expenses
($)
6,045
TOTAL
COSTS
1
­
Power
Consumption
16,830
2
­
Consumables
12,600
3
­
Labor
6,045
4
­
Chemicals
600
Total
Annual
Costs
36,070
COSTS
PER
MG
TREATED
$/
MG
10.00
Power
consumption
represents
the
majority
of
the
operational
cost.
The
power
cost
used
for
calculation
of
the
annual
O&
M
costs
was
$
0.10/
kWh.
As
expected,
the
MP
reactors
have
considerably
higher
power
costs
associated
with
their
operation
than
the
LPHO
reactors
(
Tables
K.
8
and
K.
9).

Lamp
replacement
costs
are
also
significant.
Cost
of
lamp
replacement
is
based
on
an
estimated
lamp
life
of
10,000
hours
and
$
150/
lamp
equipment
cost
for
the
LPHO
reactors.
The
MP
estimated
lamp
life
is
8,000
hours
with
a
lamp
replacement
cost
of
$
500/
lamp
for
Alternative
1
and
a
lamp
cost
of
$
600/
lamp
for
Alternatives
2
and
3.

Estimated
labor
needs
range
from
18
hours
for
the
MP
UV
installation
to
93
hours
per
year
for
the
LPHO
UV
installation.
Labor
estimates
are
based
on
lamp
replacement
at
four
lamps
per
hour
and
three
hours
per
cleaning.

Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
19
June
2003
K.
7
Summary
of
Alternatives
and
their
Advantages
and
Disadvantages
A
comparison
of
the
three
alternatives
is
presented
in
this
section.
The
alternative
comparisons
are
based
on
cost,
feasibility
of
construction,
and
ease
of
maintenance.
The
advantages
and
disadvantages
of
each
alternative
are
summarized
in
Table
K.
10.

Table
K.
10
Advantages
and
Disadvantages
of
Each
Alternative
Alternative
1
 
Filter
Gallery
 
New
building
not
necessary
 
Damp
during
periods
of
the
year
 
Below
existing
hydraulic
grade
line
(
HGL)
 
Needs
protective
cabinet
for
each
UV
reactor
 
Relatively
simple
construction
 
Tight
quarters
for
construction
and
maintenance
 
Likely
no
plant
down­
time
for
construction
 
Less
manufacturer
flexibility
 
Lowest
capital
cost
 
Must
take
a
filter
off­
line
for
UV
reactor
maintenance
 
Does
not
accommodate
expansion
easily
 
No
redundancy
at
maximum
flow
Alternative
2
 
Chemical
Building
 
New
building
not
necessary
 
UV
reactors
above
the
plant
HGL
 
Main
floor
access
 
Difficult
construction
constraints
 
Uncertainty
associated
with
structural
upgrades
 
Very
limited
space
for
relocation
of
existing
 
Chemical
equipment
to
other
parts
of
the
plant
 
Low
lift
pump
upgrades
necessary
Alternative
3
 
New
Building
 
Ample
space
for
UV
reactors
and
controls
 
Highest
capital
and
total
project
cost
 
UV
reactors
placed
below
the
plant
HGL
 
Necessitates
a
new
building
 
Room
for
future
expansion
 
Longer
construction
schedule
 
Flexibility
in
UV
installation
design
options
 
Custom
designed
space
for
UV
reactors
K.
8
Conclusions
and
Recommendations
Alternative
1,
which
involves
the
installation
of
the
UV
reactors
in
the
filter
gallery,
is
the
least
expensive
option.
However,
there
are
concerns
about
the
moisture
associated
with
the
location
that
may
adversely
affect
the
performance
of
the
UV
reactors
and
cause
maintenance
problems.
Servicing
and
maintaining
the
UV
reactors
in
the
filter
gallery
and
installing
the
necessary
control
panels
might
be
problematic
based
on
space
constraints.
In
addition,
to
expand
the
capacity
of
the
UV
installation
to
accommodate
the
ultimate
flow
(
40
mgd),
the
size
of
the
pipes
connecting
the
UV
reactors
to
the
filter
effluent
pipes
would
have
to
be
increased
and
an
additional
UV
reactor
would
have
to
be
installed.
Furthermore,
Alternative
1
does
not
provide
adequate
redundancy
at
the
ultimate
flow.

Alternative
2,
retrofitting
the
existing
chemical
feed
room
has
the
advantage
of
not
requiring
a
new
building.
However
the
disadvantages
of
this
option
include
the
space
limitations,
associated
pump
upgrade
needs
and
the
need
to
find
a
new
location
for
the
chemical
feed
equipment
currently
housed
in
that
room.

Proposal
Draft
Appendix
K.
Preliminary
Engineering
Report
UV
Disinfection
Guidance
Manual
K­
20
June
2003
Alternative
3,
constructing
a
building
addition
to
the
WTP,
is
the
most
expensive
alternative,
but
this
option
has
some
important
advantages
over
Alternatives
1
and
2.
The
new
building
would
offer
flexibility
in
design
options.
The
design
would
not
be
limited
to
using
the
MP
or
LPHO
reactors;
any
UV
reactors
could
be
accommodated.
There
would
be
room
for
future
expansion
of
the
UV
installation,
if
necessary,
and
ample
space
would
be
provided
for
mechanical
and
instrument
layouts.
The
UV
reactors
would
be
installed
below
the
existing
hydraulic
grade
line
of
the
plant
to
ensure
submergence
of
the
reactors.

Although
Alternative
3
has
some
distinct
advantages
over
Alternatives
1
and
2,
the
capital
cost
is
significantly
higher,
due
to
the
cost
of
the
new
building
and
appurtenances.
Alternative
1
is
the
most
economical
alternative,
but
does
not
accommodate
expansion
easily
and
provides
no
redundancy
at
maximum
flow.
The
disadvantages
of
Alternative
2,
including
lack
of
space
for
the
existing
chemical
equipment,
make
this
alternative
the
least
desirable.

Given
the
above
discussion,
and
based
on
both
economical
and
non­
economical
criteria
for
comparison
in
this
example,
including
anticipated
future
expansion
needs,
the
ranking
of
the
alternatives
from
most
desirable
to
least
desirable
is
as
follows:

 
Most
desirable
 
Alternative
3,
New
Building
 
Next
best
option
 
Alternative
1,
Filter
Gallery
 
Least
desirable
 
Alternative
2,
Existing
Chemical
Room
K.
9
References
Carollo
Engineers.
2001.
Weber
basin
water
treatment
plant
No.
3
expansion.
Layton,
Utah.

Mackey,
E.
D.,
R.
S.
Cushing,
and
G.
F.
Crozes.
2001.
Practical
Aspects
of
UV
Disinfection.
Denver,
Colo.:
AWWA
Research
Foundation.

Proposal
Draft
Appendix
L.
Regulatory
Timeline
The
purpose
of
this
appendix
is
to
provide
utilities
with
a
timeline
(
Figure
L.
1)
to
assist
in
planning
and
implementation
of
tasks
to
achieve
compliance
with
the
Long­
Term
2
Enhanced
Surface
Water
Treatment
Rule
(
LT2ESWTR).
The
timelines
present
the
important
tasks
that
utilities
are
likely
to
complete;
however,
the
tasks
and
their
duration
will
change
based
on
utilityspecific
priorities
and
constraints.

Tasks
have
been
divided
into
two
general
categories:
regulatory
and
engineering.
Compliance
dates
and
resulting
planning
activities
are
based
on
utility
size
(
i.
e.,
systems
serving
fewer
than
10,000
persons
or
systems
serving
10,000
or
more
persons).

L.
1
Regulatory
Tasks
Regulatory
tasks
and
milestones
include
key
dates
in
the
regulatory
schedule
such
as
monitoring
requirements
and
compliance
dates.

L.
1.1
Cryptosporidium
Monitoring
One
of
the
key
provisions
of
the
LT2ESWTR
is
the
requirement
to
conduct
monitoring
to
determine
Cryptosporidium
removal/
inactivation
requirements
(
40
CFR
141.702).
Monitoring
results
will
be
used
to
determine
a
"
bin
classification,"
which
prescribes
the
Cryptosporidium
inactivation/
removal
required.
More
information
regarding
the
monitoring
requirements
is
available
in
the
Source
Water
Monitoring
Guidance
Manual
for
Public
Water
Systems
for
the
LT2ESWTR.

L.
1.2
Compliance
Deadlines
for
Cryptosporidium
Treatment
For
utilities
required
to
provide
additional
treatment
for
Cryptosporidium,
the
compliance
deadline
is
the
date
when
a
utility
must
have
implemented
the
selected
treatment
techniques
(
40
CFR
141.701).
Table
L.
1
summarizes
the
Cryptosporidium
treatment
compliance
deadlines
for
the
LT2ESWTR.

Table
L.
1
LT2ESWTR
Compliance
Schedule
Summary1
System
Size
Compliance
Deadline
for
Systems
Making
No
Capital
Improvements
for
Compliance2
Serving
10,000
or
more
people
6
years
after
LT2ESWTR
promulgation
Serving
fewer
than
10,000
people
8
½
years
after
LT2ESWTR
promulgation
1
(
40
CFR
141.701)
2
State
may
grant
an
additional
two
years
for
systems
making
capital
improvements.

UV
Disinfection
Guidance
Manual
L­
1
June
2003
Proposal
Draft
Appendix
L.
Regulatory
Timeline
UV
Disinfection
Guidance
Manual
L­
2
June
2003
L.
2
Engineering
Tasks
and
Milestones
Engineering
tasks
and
milestones
include
tasks
that
should
be
completed
by
a
utility
to
develop
and
implement
an
LT2ESWTR
compliance
strategy.

L.
2.1
Process
Evaluation
and
Planning
Compliance
with
the
LT2ESWTR
Cryptosporidium
treatment
requirements
will
necessitate
varied
levels
of
process
evaluation
and
planning.
After
compliance
strategy
options
have
been
reviewed
(
see
section
3.1.5)
and
a
decision
has
been
made
to
implement
UV
disinfection,
planning
may
include
one
or
more
of
the
following
activities:

 
Engaging
the
State
during
planning
to
ensure
the
installation
of
UV
disinfection
is
approved
 
Conducting
disinfection
benchmarking
and
profiling
if
distribution
system
total
trihalomethane
(
TTHM)
and
five
haloacetic
acids
(
HAA5)
concentrations
are
at
least
80
percent
of
the
Stage
1
DBPR
maximum
contaminant
levels
for
TTHM
and
HAA5
(
40
CFR
141.711­
713)

 
Developing
a
capital
improvement
program
that
includes
the
necessary
modifications
for
LT2ESWTR
compliance
(
i.
e.,
UV
disinfection)

 
Evaluating
and
implementing
funding
alternatives
Utilities
are
encouraged
to
seek
approval
of
their
LT2ESWTR
compliance
plan
from
the
State
before
implementation
of
a
compliance
strategy.
This
may
take
several
months
and
can
have
a
significant
impact
on
the
implementation
schedule,
particularly
when
the
State
requires
modifications.
Because
UV
disinfection
is
a
relatively
new
technology,
the
State
may
take
longer
to
approve
UV
disinfection
or
require
more
significant
involvement
in
the
compliance
strategy
development.

L.
2.2
UV
Installation
Design
The
duration
of
the
facility
design
phase
will
be
contingent
on
a
number
of
utility­
specific
factors,
including
scope
of
design
(
i.
e.,
new
facility
or
retrofit),
scale
of
design
(
size
of
facility),
available
in­
house
resources,
procurement
methods,
and
validation
testing
requirements
(
discussed
in
detail
in
chapters
3
and
4).
The
design
will
likely
include
one
or
more
of
the
following
tasks:

 
Evaluation
of
equipment
and
contractor
procurement
methods
 
UV
reactor
procurement
 
UV
installation
design
Proposal
Draft
Appendix
L.
Regulatory
Timeline
UV
Disinfection
Guidance
Manual
L­
3
June
2003
 
UV
reactor
validation
strategy
determination
Many
States
require
final
approval
of
process
improvements.
As
such,
utilities
should
review
the
UV
installation
design
and
validation
strategy
with
the
State.
If
the
State
is
not
consulted
during
these
phases,
additional
time
may
be
necessary
to
receive
final
approval.

L.
2.3
Construction
and
Startup
The
timeline
in
Figure
L.
1
reflects
a
construction
period
of
two
years
for
both
large
and
small
utilities.
However,
the
actual
duration
of
construction
and
startup
can
vary
significantly,
depending
on
the
scope
of
the
project,
the
significance
of
the
changes
to
the
existing
treatment
plant,
and
other
utility
specific
factors.
Utilities
should
consider
these
factors
during
planning
phases
and
adjust
accordingly
to
ensure
regulatory
milestones
are
achieved
by
the
necessary
dates.

L.
3
Example
Timeline
Figure
L.
1
presents
example
timelines
that
encompass
the
regulatory
and
engineering
tasks
discussed
in
the
previous
sections.
Utilities
may
have
site­
specific
constraints
that
may
shorten
or
extend
the
duration
of
the
engineering
tasks
listed;
however,
regulatory
milestones
are
not
flexible.

Proposal
Draft
Appendix
L.
Regulatory
Timeline
Figure
L.
1
Example
LT2ESWTR
Compliance
Timeline
2003
2004
2
005
2006
2007
2008
2009
2010
2011
2012
2013
2014
Task
Description
LT2ESWTR
Proposal
LT2ESWTR
Promulgation
Systems
serving
10,000
or
more
persons
Cryptosporidium
monitoring
Initial
Bin
Classification
Process
Evaluation
and
Planning
Facility
Design
Construction
and
Startup*

Compliance
Deadline
Systems
serving
less
than
10,000
persons
Cryptosporidium
monitoring
Initial
bin
classification
Process
Evaluation
and
Planning
Facility
Design
Construction
and
Startup*

Compliance
Deadline
*
State
may
grant
an
additional
2
years
for
system
making
capital
improvements
Regulatory
Tasks/
Milestones
Engineering
Tasks/
Milestones
UV
Disinfection
Guidance
Manual
L­
4
June
2003
Proposal
Draft
Appendix
M.
Compliance
Forms
This
appendix
is
intended
to
supplement
the
monitoring
information
provided
in
section
5.4
with
examples
of
monthly
compliance
report
and
monitoring
log
forms
that
utilities
might
use
for
reporting
to
the
State.
(
Note,
these
are
only
examples;
the
States
may
develop
their
own
compliance
forms
and
require
additional
monitoring.)
The
specific
monitoring
and
reporting
requirements
for
each
utility
should
be
confirmed
with
the
State,
and
the
forms
should
be
modified
accordingly.
For
those
utilities
with
advanced
control
systems
(
e.
g.,
Supervisory
Control
and
Data
Acquisition
(
SCADA)),
it
may
be
possible
to
automatically
generate
these
reports
and
compliance
forms.

To
receive
disinfection
credit,
the
Long
Term
2
Enhanced
Surface
Water
Treatment
Rule
(
LT2ESWTR)
requires
validation
testing
of
UV
reactors
to
demonstrate
a
set
of
operating
conditions
where
the
UV
reactor
will
deliver
the
required
dose
(
40
CFR
141.729
(
d)).
These
operating
conditions
must
include
flowrate,
UV
intensity,
and
UV
lamp
status,
and
the
utility
must
monitor
these
parameters
during
routine
operation
to
ensure
dose
delivery.
(
40
CFR
141.729
(
d)).
States
may
specify
additional
monitoring
or
reporting
requirements.
The
example
forms
presented
in
this
appendix
list
both
required
and
recommended
monitoring
parameters
(
required
parameters
are
identified
with
the
applicable
rule
citation).

Table
M.
1
summarizes
the
recommended
minimum
level
of
monitoring
and
record
keeping
for
utilities
utilizing
UV
disinfection.
For
many
of
the
UV
reactor
components,
the
required
or
recommended
performance
level
is
based
on
the
measurement
uncertainty
of
the
specific
equipment
that
was
used
when
the
UV
reactor
was
validated.
This
uncertainty
is
used
to
determine
the
validation
safety
factor
and
recommended
reduction
equivalent
dose
(
i.
e.,
operational
UV
dose),
as
described
in
section
4.2.

UV
Disinfection
Guidance
Manual
M­
1
June
2003
Proposal
Draft
Appendix
M.
Compliance
Forms
UV
Disinfection
Guidance
Manual
M­
2
June
2003
Table
M.
1.
Summary
of
Compliance
Monitoring
and
Reporting
Activities1,
2
Item
Description
Measured
Parameter
Recommended
Monitoring
Frequency
Reporting
Frequency
Offspecification
Validated
Parameters
for
UV
Dose
Monitor
reactor
to
ensure
operation
within
conditions
validated
for
required
UV
dose
(
40
CFR
141,
Subpart
W,
Appendix
D).
Flowrate,
UV
intensity,
lamp
status,
and
other
parameters
(
e.
g.,
UVT)
used
to
monitor
dose.
Continuously.
Record
at
least
once
every
four
hours
(
daily
for
very
small
systems).
Required
monthly
report
of
offspecification
operation
as
a
percent
of
distributed
flow
or
operating
time
(
40
CFR
141.730).
3
Calibration
of
UV
Intensity
Sensors
Calibration
checks
compare
the
duty
sensor
to
the
reference
sensor
and
are
recommended
at
the
power
setting
utilized
during
normal
operation.
Percent
difference
between
duty
and
reference
sensors
relative
to
the
level
of
uncertainty
used
in
determining
the
RED.
(
see
section
C.
4.7)
Monthly.
If
a
sensor
fails
for
three
consecutive
months,
then
the
sensor
should
be
checked
weekly
and
the
manufacturer
contacted.
Requirements
in
a
State
approved
protocol.

Calibration
of
UV
Transmittance
(
UVT)
Monitor
It
is
recommended
that
grab
samples
be
collected
to
confirm
performance.
Percent
difference
relative
to
the
manufacturer's
guaranteed
uncertainty.
Weekly
initially.
Reduced
frequency
following
oneyear
of
supporting
data.
NR.

1
Section
5.4.2
presents
all
recommended
monitoring
activities,
including
the
compliance
monitoring
shown
in
this
table.
2
Unless
noted
in
the
table
with
an
LT2ESWTR
citation,
the
monitoring
is
recommended
and
not
required.
3
The
reported
off­
specification
value
is
the
percentage
of
water
entering
the
distribution
that
was
not
treated
with
UV
reactors
operating
within
validated
conditions.
This
is
required
by
the
LT2ESWTR
(
40
CFR
141.730).
NR
 
No
requirement
The
LT2ESWTR
requires
utilities
to
submit
monthly
reports
to
the
State
(
40
CFR
141.730).
At
a
minimum,
the
reports
must
detail
operating
performance
during
the
reporting
period
and,
specifically,
the
percent
of
total
distributed
volume
treated
during
periods
when
the
UV
reactor(
s)
was
off­
specification.
An
example
monthly
monitoring
form
is
shown
in
Table
M.
2.
Tables
M.
3
through
M.
6
present
a
format
that
the
utility
can
use
to
log
operating
data
for
development
of
the
monthly
reports.
With
minor
modification,
the
example
forms
are
applicable
for
any
of
the
three
control
strategies
discussed
in
section
4.3.2.2:
UV
intensity
setpoint,
UV
intensity
and
UV
transmittance
(
UVT)
setpoint,
and
calculated
dose.

For
those
utilities
utilizing
multiple
reactors,
the
operation
of
each
reactor
must
be
monitored,
recorded,
and
reported.
Requirements
for
compliance
monitoring
beyond
those
established
by
the
LT2ESWTR
and
the
specific
content
of
the
monthly
report
will
be
established
by
the
State
and
coordinated
with
all
other
reporting
requirements.
Additional
information
on
UV
reactor
monitoring
and
maintenance
is
provided
in
Chapter
5.

Proposal
Draft
Appendix
M.
Compliance
Forms
UV
Disinfection
Guidance
Manual
M­
3
June
2003
Greater
detail
on
each
of
the
example
forms
is
provided
below:

 
Form
M.
2
is
an
example
of
a
summary
report
that
would
be
completed
by
the
utility
and
submitted
to
the
State
on
a
monthly
basis.

 
Forms
M.
3A,
M.
3B,
and
M.
3C
are
example
reference
forms
for
each
of
the
three
control
strategies
discussed
in
section
4.3.2.2.
These
forms
would
be
completed
by
the
utility
based
on
validation
results
and
then
referenced
throughout
the
operation
of
the
UV
installation
to
confirm
compliance.

 
Form
M.
4
is
an
example
operating
log
that
would
be
completed
on
a
daily
basis.
The
form
would
be
used
to
record
the
operating
status
of
the
UV
installation
and
to
estimate
the
volume
of
water
that
was
discharged
during
off­
specification
operation.

 
Form
M.
5
is
an
example
sensor
calibration
log.
This
log
would
be
completed
whenever
sensor
calibration
checks
are
performed.
The
log
would
be
used
to
record
the
results
of
the
calibration
testing
as
well
to
track
any
sensor
recalibration
or
repair
work
that
was
completed.

 
Form
M.
6
is
an
example
on­
line
UVT
monitor
calibration
log.
This
log
would
only
be
completed
by
those
utilities
that
have
included
on­
line
UVT
monitors
as
part
of
their
design.
The
log
would
be
completed
whenever
UVT
monitor
calibration
checks
are
performed.
The
log
would
be
used
to
record
the
results
of
the
calibration
testing
as
well
to
track
any
recalibration
or
repair
work
that
was
completed.

Proposal
Draft
Appendix
M.
Compliance
Forms
Table
M.
2
Example
Monthly
Report
to
State
Reporting
Period:

System/
Treatment
Plant:

PWSID:

Monthly
Reactor
Operating
Report
Operating
Data
Off­
Specification1
Unit
No.
Operating
Time
(
Hours)
Volume
Treated
(
Gallons)
No.
of
Events
Volume
(
Gallons)
Time
(
Hours)
Percent
of
Volume
Treated
%
%
%
%
%
%
%
%
%

TOTALS:

Compliance
Certification:

By
volume
the
total
percent
of
off­
specification
operation
during
the
reporting
period
=
%

Of
the
_____
sensors
within
those
reactors
that
operated
during
this
reporting
period,
_____
have
been
checked
for
calibration
and
were
within
the
acceptable
range
of
tolerance.

Signature
of
Principal
Executive
Officer
or
Authorized
Agent:
Date:

1
From
Table
M.
4
UV
Disinfection
Guidance
Manual
M­
4
June
2003
Proposal
Draft
Appendix
M.
Compliance
Forms
Table
M.
3A
UV
Intensity
Setpoint
Control
Strategy1
Validated
Operating
Requirements
to
Achieve
Disinfection
Credit
for
(
Target
Pathogen)

(
Reference
Document
­
See
Instructions
Below)

Validated
Operating
Conditions
Flow
Rate
UV
Intensity
Log
Inactivation2
1
This
form
is
for
use
with
a
UV
Intensity
setpoint
operating
strategy.
If
a
different
operating
strategy
is
employed
by
the
utility,
then
one
of
the
other
sample
reference
forms
should
be
used
or
a
form
summarizing
the
specific
validation
criteria
necessary
to
confirm
compliance
using
the
selected
operating
strategy
should
be
developed.

2
For
those
utilities
that
do
not
employ
variable
power
settings
or
that
have
a
single
target
log
inactivation,
this
column
may
not
be
necessary.

Instructions:

This
form
should
be
completed
based
on
the
validation
testing
results
and
used
as
a
reference
document.

This
form
should
be
referenced
to
determine
if
the
UV
system
is
operating
within
its
validated
conditions
and
meeting
the
performance
requirements
for
inactivation
credit
for
the
target
pathogen.

UV
Disinfection
Guidance
Manual
M­
5
June
2003
Proposal
Draft
Appendix
M.
Compliance
Forms
Table
M.
3B
UV
Intensity
and
UVT
Setpoint
Control
Strategy1
Validated
Operating
Requirements
to
Achieve
Disinfection
Credit
for
(
Target
Pathogen)

(
Reference
Document
­
See
Instructions
Below)

Validated
Operating
Conditions
Flow
Rate
UV
Intensity
UVT
Log
Inactivation2
1
This
form
is
for
use
with
a
UV
Intensity/
UVT
setpoint
operating
strategy.
If
a
different
operating
strategy
is
employed
by
the
utility,
then
one
of
the
other
sample
reference
forms
should
be
used
or
a
form
summarizing
the
specific
validation
criteria
necessary
to
confirm
compliance
using
the
selected
operating
strategy
should
be
developed.

2
For
those
utilities
that
do
not
employ
variable
power
settings
or
that
have
a
single
target
log
inactivation,
this
column
may
not
be
necessary.

Instructions:

This
form
should
be
completed
based
on
the
validation
testing
results
and
used
as
a
reference
document.

This
form
should
be
referenced
to
determine
if
the
UV
system
is
operating
within
its
validated
conditions
and
meeting
the
performance
requirements
for
inactivation
credit
for
the
target
pathogen.

UV
Disinfection
Guidance
Manual
M­
6
June
2003
Proposal
Draft
Appendix
M.
Compliance
Forms
Table
M.
3C
Calculated
Dose
Control
Strategy1
Validated
Operating
Requirements
to
Achieve
Disinfection
Credit
for
(
Target
Pathogen)

(
Reference
Document
­
See
Instructions
Below)

Validated
Operating
Conditions
Flow
Rate
UV
Intensity
UVT
Calculated
Dose
Log
Inactivation2
1
This
form
is
for
use
with
a
calculated
dose
operating
strategy.
If
a
different
operating
strategy
is
employed
by
the
utility,
then
one
of
the
other
sample
reference
forms
should
be
used
or
a
form
summarizing
the
specific
validation
criteria
necessary
to
confirm
compliance
using
the
selected
operating
strategy
should
be
developed.

2
For
those
utilities
that
do
not
employ
variable
power
settings
or
that
have
a
single
target
log
inactivation,
this
column
may
not
be
necessary.

Instructions:

This
form
should
be
completed
based
on
the
validation
testing
results
and
used
as
a
reference
document.

This
form
should
be
referenced
to
determine
if
the
UV
system
is
operating
within
its
validated
conditions
and
meeting
the
performance
requirements
for
inactivation
credit
for
the
target
pathogen.

UV
Disinfection
Guidance
Manual
M­
7
June
2003
Proposal
Draft
Appendix
M.
Compliance
Forms
Table
M.
4
Daily
UV
Intensity
Sensor
and
UVT
Monitoring
and
Compliance
Log1
UVT
Monitor
b
Reading
Reading
Date:
_______________________________________________
Unit
Did
Not
Operate
During
This
Monitoring
Interval:

System/
Treatment
Plant:
_______________________________

PWSID:
_____________________________________________
Operator
Signature:

Unit
Number:
_______________________________

UV
Transmittance
Monitor:
Grab
Sample
Collected:
Y/
N
If
yes,
complete
Calibration
Log
Sheet.

Calibration
Check
Performed:
Y/
N
UV
Intensity
Sensors:
Calibration
Check
Performed:
Y/
N
If
yes,
complete
Calibration
Log
Sheet.

Cumulative
UV
Intensity
Sensors
Flow
Rate
Volume
No.
Time
(
gpd)
(
gals)
Reading
Reading
Reading
Reading
Reading
Reading
Reading
Reading
123456
Time
Flow
Rate
UVT
Acceptable
(
Y/
N)
UV
Intensity
Acceptable
(
Y/
N)
2
Calculated
Dose
2
Volume
3
Operating
Time
Off­
Spec
(
Y/
N)
Volume
Off­

Spec
4
Reading
No.
1
Reading
No.
2
Reading
No.
3
Reading
No.
4
Reading
No.
5
Reading
No.
6
Daily
Total:
A
B
Percent
Off­
Specification
by
Volume:
%
>
5%
(
Y/
N)
5
(
B/
A
x
100)

Notes:

1
As
presented,
this
form
is
most
applicable
to
a
calculated
dose
control
strategy,
but
can
be
modified
to
suit
all
control
strategies.

2
Only
if
applicable
to
selected
control
strategy.

3
Volume
treated
since
last
reading.

4
For
systems
that
rely
solely
on
manual
readings,
the
volume
off­
specification
shall
be
calculated
as
the
total
volume
discharged
since
the
last
recording
interval
during
which
operation
within
validated
conditions
was
observed.
If
continuous
monitoring
is
provided,
then
the
volume
off­
specification
shall
be
only
that
portion
of
the
volume
distributed
during
the
recording
interval
that
was
outside
of
the
validated
conditions.

5
LT2ESWTR
requires
that
off­
specification
be
less
than
5
percent
by
volume
on
a
monthly
basis
for
unfiltered
systems
(
§
141.721).
For
filtered
systems,
it
is
that
off­
specification
operation
be
less
than
5
percent
by
volume
on
a
monthly
basis.
The
specific
requirements
for
allowable
off­
specification
for
filtered
systems
will
be
established
by
the
state.

Instructions:

This
form
should
be
completed
daily
for
each
operating
unit.
If
a
unit
did
not
operate
during
the
24­
hour
monitoring
interval,
please
note
as
such
in
the
box
at
the
top
of
the
log.

This
form
is
intended
for
periodic
manual
monitoring
of
system
operating
conditions
at
the
recommended
interval
of
once
every
four
hours.
If
automated
or
more
frequent
monitoring
is
performed,

then
this
form
should
be
modified
by
the
utility.

UV
Disinfection
Guidance
Manual
M­
8
June
2003
Proposal
Draft
Appendix
M.
Compliance
Forms
Table
M.
5
Daily
UV
Intensity
Sensor
and
UVT
Monitoring
and
Compliance
Log1
where,
is
measured
intensity
is
measurement
uncertainty
Duty
Sensor
Reading
(
A)
Reference
Sensor
Reading
(
B)
Error
[(
A­
B)/
Bx100]
Acceptable
(
Y/
N)

alibration.

ange
of
tolerance.

ented
below.
Total
Uncertainty
Criteria
=
(
)
2
1
2
Duty
2
Ref
Ref
uty
D
100
*

1
I
I
 
+

 
 

 
  
 

 
  
 
 

 
I
Reporting
Period:

System/
Treatment
Plant:

PWSID:
Duty
UV
Intensity
Sensor
Uncertainty
(%):
where,

Reference
Intensity
Sensor
Uncertainty
(%):
is
measured
intensity
Total
Uncertainty
Criteria
(%):
is
measurement
uncertainty
UVT
Intensity
Sensor
Calibration
Report
(
Make
Additional
Copies
of
Form
as
Necessary)

Reactor/
Duty
Sensor
No.
Reference
Sensor
ID
Duty
Sensor
ID
Date
Time
Duty
Sensor
Reading
(
A)
Reference
Sensor
Reading
(
B)
Error
[(
A­
B)/
Bx100]

Certification:

Of
the
______
sensors
within
those
reactors
that
operated
during
this
reporting
period,
______
have
been
checked
for
calibration.

One
or
more
of
the
sensors
within
those
reactors
that
operated
during
this
reporting
period
were
outside
the
acceptable
range
of
tolerance.

UV
Intensity
Sensor(
s)
sent
to
manufacturer
to
be
recalibrated
as
documented
below.

UV
intensity
sensors
sent
to
manufacturer
for
calibration:

Sensor
ID
Unit
No.
Date
Sent
Date
Received
Operator
Signature:
Date:
Total
Uncertainty
Criteria
=
(
2
Ref
Ref
uty
D
100
*

1
I
I
 
 

 
  
 

 
  
 
 

 
I
UV
Disinfection
Guidance
Manual
M­
9
June
2003
Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
Lamps
used
in
UV
reactors
typically
contain
mercury
or
an
amalgam
composed
of
mercury
and
another
element,
such
as
indium
or
gallium.
Other
elements,
such
as
xenon,
cadmium,
zinc,
and
magnesium,
are
also
capable
of
generating
UV
light;
however,
the
temperatures
required
to
volatilize
these
elements
are
much
higher
than
to
volatilize
mercury.
In
contrast,
mercury
has
a
sufficient
vapor
pressure
at
ambient
temperatures
to
provide
the
optimum
pressure
for
efficient
production
of
resonance
radiation.
Moreover,
mercury
has
a
low
ionization
energy
to
facilitate
starting
a
lamp
(
Phillips
1983).
In
order
to
provide
a
cost­
efficient
lamp
while
addressing
perceived
risks
and
disposal
issues
associated
with
mercury,
lamp
manufacturers
are
continuing
to
develop
ways
to
reduce
the
mercury
content
of
lamps
without
impacting
their
efficiency
(
USEPA
1997b;
Walitsky
2001).

The
mercury
contained
within
a
UV
lamp
is
isolated
from
exposure
to
water
by
a
lamp
envelope
and
surrounding
lamp
sleeve.
In
order
for
mercury
to
be
released
into
the
water,
both
the
lamp
and
lamp
sleeve
must
break.
For
the
purposes
of
this
appendix,
lamp
breakage
is
defined
as
fracture
of
the
lamp
sleeve
and
the
lamp
envelope.
This
is
further
divided
into
off­
line
and
on­
line
breaks.
Off­
line
breaks
occur
during
handling
or
maintenance
functions
when
the
lamps
are
not
installed
in
the
reactor.
On­
line
lamp
breaks
occur
while
UV
reactors
are
in
operation.

Due
to
the
general
public
health
concern
with
mercury,
this
appendix
discusses
the
issues
associated
with
UV
lamps
used
for
drinking
water
disinfection
by
addressing
potential
causes
of
lamp
breakage,
preventive
measures,
disposal
issues,
the
fate
of
mercury
after
release,
and
regulatory
issues.

N.
1
Off­
Line
Lamp
Breaks
Off­
line
breaks
occur
when
a
lamp
breaks
during
shipping,
handling,
or
storage.
These
releases
do
not
pose
a
hazard
to
the
water
consumer
but
are
a
concern
for
operators
or
employees
in
the
vicinity
of
the
break.

N.
1.1
Potential
Causes
of
Off­
Line
Lamp
Breaks
and
Corresponding
Prevention
Measures
Mercury
is
sealed
in
a
UV
lamp
within
the
lamp
envelope;
therefore,
there
is
no
risk
of
mercury
exposure
from
handling
an
unbroken
UV
lamp.
The
UV
manufacturer
should
train
operators
in
proper
handling
and
maintenance
of
UV
lamps
to
avoid
mishandling
and
potential
off­
line
breaks.
In
addition,
proper
storage
procedures
will
also
reduce
the
potential
for
lamp
breakage.
Lamps
should
be
stored
horizontally
in
individual
packaging.
Lamps
should
not
be
stacked
unpackaged
on
one
another
or
vertically
propped
in
corners
(
Dinkloh
2001a).

UV
Disinfection
Guidance
Manual
N­
1
June
2003
Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
2
June
2003
N.
1.2
Off­
Line
Mercury
Release
Cleanup
Procedures
Off­
line
lamp
breaks
resulting
in
a
release
of
mercury
can
occur;
therefore,
Standard
Operating
Procedures
(
SOPs)
should
be
developed
that
describe
the
procedures
for
containing
and
cleaning
the
off­
line
spills.
The
local
poison
control
center,
fire
department,
or
public
health
board
can
assist
in
the
development
of
SOPs.

Small
spills,
defined
as
less
than
about
0.6
to
2.25
grams
(
USEPA
1992)
or
the
amount
in
a
broken
thermometer
(
USEPA
1997a),
can
be
contained
and
collected
with
commercially
available
mercury
spill
kits.
Mercury
and
materials
used
during
the
cleanup
procedure
are
regulated
as
hazardous
wastes
and
should
be
disposed
of
properly
as
described
in
section
N.
3.3.
The
USEPA
Office
of
Emergency
and
Remedial
Response
recommends
that
"[
i]
n
the
event
of
a
large
mercury
spill
(
more
than
a
broken
thermometer's
worth),
immediately
evacuate
everyone
from
the
area,
seal
off
the
area
as
well
as
possible,
and
call
your
local
authorities
for
assistance"
(
USEPA
1997a).
Local
authorities
can
help
determine
the
appropriate
response
for
various
spill
sizes
to
be
included
in
SOPs.
Given
that
the
mercury
content
in
a
single
UV
lamp
typically
ranges
from
0.005
to
0.4
grams
(
as
discussed
in
section
N.
4.3),
large
mercury
spill
actions
would
not
be
warranted
for
a
single
lamp
break
or
multiple
lamp
breaks
that
result
in
release
of
less
than
roughly
two
grams.

Superfund
Amendments
and
Reauthorization
Act
(
SARA)
Title
III
regulations
address
emergency
release,
inventory,
and
release
reporting
requirements
for
hazardous
materials.
The
reportable
quantity
for
mercury
spills
is
one
pound
(
454
grams)
as
mercury.
Based
on
typical
mercury
levels
in
UV
lamps
(
discussed
in
section
N.
4.3),
this
would
necessitate
the
breakage
of
approximately
1,100
medium
pressure
(
MP)
lamps
and
up
to
90,000
low
pressure
(
LP)
lamps;
as
such,
spilling
more
than
one
pound
of
mercury
is
highly
unlikely.

N.
2
On­
Line
Lamp
Breaks
A
recent
survey
of
domestic
water
and
wastewater
municipalities,
UV
lamp
manufacturers,
and
UV
reactor
manufacturers
identified
relatively
few
instances
of
on­
line
lamp
breaks
and
mercury
release
(
Malley
2001).
This
section
discusses
potential
causes
of
lamp
breakage
and
corresponding
prevention
measures,
followed
by
a
summary
of
documented
incidents
of
on­
line
lamp
breaks.

N.
2.1
Potential
Causes
of
On­
Line
Lamp
Breaks
and
Corresponding
Prevention
Measures
Lamp
breaks
can
potentially
be
caused
by
debris
in
the
water,
temperature
variations,
exceeding
positive
or
negative
pressure
limits
(
water
hammer),
electrical
surges,
or
improper
maintenance.
Lamps
may
also
break
as
a
result
of
inherent
mechanical
or
physical
limitations
of
the
lamp
and
improper
material
selection.

Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
3
June
2003
N.
2.1.1
Debris
Debris
in
the
water
can
potentially
break
the
lamp
sleeves
and
lamps.
Although
the
majority
of
UV
reactors
will
be
installed
after
the
filters
in
the
treatment
train,
it
is
possible
that
equipment
failure
upstream
may
release
parts
or
fragments,
such
as
nuts
or
bolts.
In
addition,
if
UV
disinfection
is
applied
prior
to
the
filters
the
probability
of
having
debris
in
the
water
might
be
higher
compared
to
post­
filter
UV
installation.
Ground
water
systems
have
reported
stones
or
gravel
from
wells
entering
UV
reactors
and
breaking
lamps
(
Malley
2001;
Roberts
2000).

Placement
of
screens,
baffles,
or
low
velocity
collection
areas
upstream
of
UV
reactors
or
vertical
installation
of
UV
reactors
(
when
applicable)
may
reduce
the
risk
of
debris
in
the
water
from
entering
the
reactor
(
Cairns
2000;
Malley
2001,
McClean
2001b).
The
extent
of
containment
provided
by
these
safety
measures
is
unknown.
Utilities
and
designers
should
determine
the
applicability
of
these
isolation
techniques
on
a
site­
specific
basis.

N.
2.1.2
Loss
of
Water
Flow
and
Temperature
Considerations
UV
lamps
are
designed
to
operate
within
a
specific
temperature
range
to
maximize
the
UV
light
output
of
the
lamp.
Without
flowing
water
to
cool
the
lamp,
the
lamp
temperature
can
rise
to
dangerous
levels
and
may
break
(
Dinkloh
2001a;
Malley
2001;
Srikanth
2001a;
Srikanth
2001b).
This
overheating
is
more
likely
to
occur
with
MP
than
LP
lamps
(
due
to
lamp
operating
temperatures)
and
occurs
much
faster
in
air
than
stagnant
water.
Even
if
upper
temperature
levels
are
not
exceeded,
after
restoration
of
water
flow,
the
lower
temperature
water
entering
the
reactor
may
cause
the
lamp
sleeve
and
the
lamp
to
break
due
to
temperature
differentials
(
Dinkloh
2001a;
Malley
2001).
In
order
to
prevent
lamp
breaks,
operating
procedures
should
ensure
that
the
following
conditions
are
met:

 
Water
is
flowing
through
the
UV
reactor
if
the
UV
lamps
are
energized.

 
The
lamps
are
not
energized
while
the
reactor
is
not
flowing
full
(
i.
e.,
no
air
in
the
reactor).

Temperature
sensors
should
be,
and
typically
are,
incorporated
into
the
reactor
design
and
will
shut
down
the
reactor
before
critical
temperatures
are
exceeded
(
Cairns
2000;
Dinkloh
2001a;
Malley
2001;
Srikanth
2001b).
Proper
hydraulic
design
is
also
necessary
to
ensure
that
lamps
are
submerged
at
all
times
during
reactor
operation.
Reactor
designs
should
incorporate
low
flow
alarms,
air
relief
valves,
or
other
devices
to
ensure
that
lamps
are
operating
only
when
the
reactor
is
completely
flooded
and
water
is
flowing.
These
sensors
should
be
linked
to
an
alarm
and
automatic
shutoff
system
(
Cairns
2000;
Dinkloh
2001a;
Srikanth
2001b).
Lamp
overheating
and
temperature
differentials
could
break
all
the
lamps
within
the
affected
reactor.

N.
2.1.3
Pressure­
Related
Issues
Hydraulic
pressures
within
the
reactor
that
are
not
within
UV
installation
operating
limits
may
also
break
the
lamp
sleeve.
Although
breaking
the
lamp
sleeve
does
not
automatically
break
the
lamp
envelope,
the
lamp
is
more
vulnerable
when
its
lamp
sleeve
has
been
Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
4
June
2003
compromised,
potentially
allowing
the
lamp
envelope
to
come
into
direct
contact
with
the
surrounding
water.

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
2001c;
Srikanth
2001a;
Srikanth
2001b).
However,
negative
gauge
pressures
below
­
1.5
psig
have
been
shown
to
adversely
affect
lamp
sleeve
integrity
(
Dinkloh
2001c).
The
tolerance
level
of
the
lamp
sleeve
depends
on
the
quality
of
the
quartz
and
the
thickness
and
length
of
the
lamp
sleeve;
therefore,
pressure
thresholds
vary
between
lamp
sleeves.
Positive
and
negative
pressures,
such
as
those
associated
with
water
hammer,
that
exceed
these
levels
may
compromise
the
integrity
of
the
lamp
sleeve.
Manufacturers
should
provide
lamp
sleeves
with
the
appropriate
material,
thickness,
geometry,
and
seals
for
the
specified
pressure
and
flow
ranges
of
a
given
UV
installation.
Water
hammer
can
affect
all
UV
reactors
and
break
all
lamps;
therefore,
utilities
should
perform
a
surge
analysis
to
determine
if
water
hammer
is
a
potential
problem.

N.
2.1.4
Procedural
Errors
Operation
and
maintenance
training
can
help
prevent
lamp
breaks
during
on­
line
operations
because
a
lamp
damaged
by
off­
line
handling
or
improper
maintenance
operations
may
potentially
break
under
on­
line
pressure
or
temperature
stresses.
For
example,
a
common
procedural
error
that
can
occur
during
lamp
replacement
is
over­
tightening
compression
nuts
when
securing
the
lamp
sleeve
(
Aquafine
2001;
Dinkloh
2001a;
Srikanth
2001a;
Srikanth
2001b;
Swaim
2002).
Over­
tightening
can
cause
a
fracture
of
the
lamp
sleeve
or
a
leak
around
the
sleeve
or
compression
nut
cavity
that
may
not
become
apparent
until
after
start­
up
and
operation
of
the
UV
reactor.

N.
2.1.5
UV
Reactor
Design
The
UV
reactor
manufacturer
should
design
the
UV
reactor
to
reduce
the
possibility
of
lamp
sleeve
and
lamp
breaks.
This
subsection
describes
design
problems
that
may
cause
lamp
sleeve
and/
or
lamp
breakage
if
not
properly
addressed.

Electrical
Considerations
If
the
UV
installation
electrical
support
system
is
improperly
designed
(
e.
g.,
inadequate
circuit
breakers
and
ground
fault
indicator
circuits),
electrical
surges
can
cause
short­
circuiting
and
lamp
socket
damage
(
Srikanth
2001a;
Srikanth
2001b).
In
addition,
system
electronics
that
can
provide
voltages
that
exceed
lamp
ratings
(
overdriving
lamps)
may
also
result
in
breaking
the
lamp
(
Malley
2001).

Cleaning
Mechanism
Considerations
The
cleaning
mechanism
may
break
the
lamp
sleeve
and
lamp
envelope
if
it
is
not
aligned
properly.
Although
the
cleaning
mechanism
closely
surrounds
the
lamp
sleeve
for
cleaning,
manufacturers
should
ensure
that
the
mechanism
is
flexible
and
able
to
adjust
to
minor
misalignment
of
the
lamp
sleeves.

Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
5
June
2003
At
high
lamp
temperatures,
the
cleaning
mechanism
in
some
UV
reactors
may
fuse
to
the
lamp
sleeve
when
not
in
use.
As
a
result,
during
the
next
cleaning
event,
the
lamp
sleeve
may
crack
when
the
cleaning
mechanism
is
activated
or
when
the
cleaning
mechanism
passes
back
over
the
residual
left
on
the
lamp
sleeve
(
Dinkloh
2001a).
Routine
inspection
according
to
manufacturers'
recommendations
will
help
detect
problems
with
the
cleaning
mechanism
before
damage
occurs.
In
some
UV
reactors,
wipers
rest
away
from
the
lamp
sleeve
when
not
in
use
and
an
alarm
sounds
when
the
wiper
stops
along
the
lamp
sleeve.

Thermal
Expansion
and
Contraction
Other
potential
causes
of
lamp
breaks
include
improper
matching
of
lamp
materials
with
respect
to
thermal
expansion
characteristics.
Compatible
materials
within
the
lamp
should
be
used
by
the
manufacturer
to
avoid
stress
and
damage
that
can
be
caused
by
thermal
expansion
and
contraction
differences
between
materials
under
various
operating,
shipping,
or
handling
conditions
(
Cairns
2000).
In
addition,
improper
seal
design
or
lamp
envelope
swelling
may
cause
water
leaks
around
the
seals
that
may
result
in
electrical
shorts
and
cracking
of
lamps
(
Cairns
2000).

N.
2.1.6
Summary
of
Potential
Causes
and
Methods
of
Prevention
of
On­
Line
UV
Lamp
Breaks
Table
N.
1
summarizes
the
potential
causes
of
on­
line
lamp
breaks
and
provides
a
brief
description
of
the
preventive
measures
that
UV
installation
designers
and
operators
can
implement
to
reduce
each
risk.
There
are
few
documented
cases
where
lamps
have
been
broken
during
on­
line
operations,
which
are
discussed
in
section
N.
2.2.

Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
6
June
2003
Table
N.
1
Summary
of
Potential
Causes
and
Methods
of
Prevention
of
On­
Line
UV
Lamp
Breaks
Potential
Cause
Description
Preventive
Measure
Debris
 
Physical
impact
of
debris
on
lamp
sleeves
may
cause
lamp
breaks.
 
Installation
of
screens,
baffles,
or
low
velocity
collection
areas
upstream
of
UV
reactors
or
vertical
installation
of
UV
reactors
will
help
prevent
debris
from
entering
the
reactor.
Loss
of
Water
Flow
and
Temperature
Considerations
 
Lamps
may
overheat
and
break.
 
The
temperature
differential
between
stagnant
water
or
air
and
flowing
water
may
cause
lamp
breaks.
 
Reactors
should
always
be
completely
flooded.
Temperature
and
flow
sensors
that
are
linked
to
an
alarm
and
automatic
shutoff
system
can
be
used
to
indicate
irregular
temperature
or
flow
conditions.
Pressure­
Related
Considerations
 
Excessive
positive
or
negative
pressures
may
exceed
lamp
sleeve
tolerances
and
break
the
lamp
sleeve.
 
A
surge
analysis
should
be
completed
to
determine
the
occurrence
of
water
hammer.
 
Pressure
relief
valves
or
other
measures
can
be
used
to
reduce
pressure
surges.
 
Applicable
pressure
ranges
should
be
specified
for
lamp
sleeves.
Procedural
Errors
 
Improper
handling
or
maintenance
may
compromise
the
integrity
of
the
lamp
sleeve
and/
or
lamp.
 
Operators
and
maintenance
staff
should
be
trained
by
the
manufacturer.

UV
Reactor
Design
 
Electrical
surges
can
cause
short­
circuiting
and
lamp
socket
damage.
 
Applying
power
that
exceeds
design
rating
of
lamps
can
cause
lamps
to
burst
from
within.
 
Adequate
circuit
breakers/
ground
fault
indicators
should
be
specified
to
prevent
damage
to
the
reactor.
 
Replacement
lamps
should
be
electrically
compatible
with
reactor
design.

 
Misaligned
or
heat­
fused
cleaning
mechanism
may
break
or
damage
the
lamp
sleeve
and
lamp
envelope.
 
Operators
and
maintenance
staff
should
perform
routine
inspection
and
maintenance
according
to
manufacturers'
recommendations.

 
Thermally
incompatible
materials
do
not
allow
for
expansion
and
contraction
of
lamp
components
under
required
temperature
range.
 
Designers
should
specify
temperature
ranges
likely
to
be
encountered
during
shipping,
storage,
and
operation
of
lamps
to
aid
the
manufacturer
in
the
selection
of
thermally
compatible
materials.

N.
2.2
Frequency
of
On­
Line
Lamp
Breaks
There
have
been
relatively
few
documented
incidents
of
on­
line
lamp
breaks.
As
part
of
a
survey
of
domestic
water
and
wastewater
municipalities,
UV
lamp
manufacturers,
and
UV
reactor
manufacturers,
Malley
(
2001)
identified
nine
cases
of
on­
line
lamp
breaks.
Both
the
lamp
sleeve
and
lamp
envelope
were
damaged
in
all
nine
cases,
resulting
in
mercury
release
(
Table
N.
2).
No
cases
of
on­
line
failures
using
LP
or
low
pressure
high
output
(
LPHO)
lamps
were
identified.
However,
LPHO
lamps
are
relatively
new
to
the
UV
disinfection
market
and
all
LPHO
lamp
installations
have
been
operating
for
5
years
or
fewer
(
Malley
2001).
All
nine
cases
involved
MP
lamps.
Four
of
the
nine
lamp
breaks
were
caused
by
impacts
from
stones
on
lamps
Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
7
June
2003
oriented
perpendicular
to
flow.
In
one
of
the
nine
lamp
breaks,
the
applied
power
exceeded
design
rating
of
lamp
(
30kW)
causing
the
lamp
to
burst
from
within.
Differential
sleeve
heating
resulted
in
two
of
the
nine
documented
lamp
breaks.
The
lamps
were
mounted
vertically
in
the
UV
reactor
allowing
heat
to
accumulate
at
the
top
of
the
lamp,
eventually
cracking
the
sleeve.
In
two
of
the
nine
instances,
operating
lamps
reached
extremely
high
temperatures
(>
600
°
C)
in
air
because
the
reactors
lost
water
flow.
When
water
flow
resumed,
the
cooler
water
(
20
°
C)
broke
the
lamps.
Most
of
these
documented
cases
of
lamp
failure
were
the
result
of
design
issues
that
have
been
addressed
in
modern
reactor
designs.
As
mentioned
previously,
temperature
and
flow
alarms
should
shut
the
UV
reactor
down
when
the
potential
for
overheating
or
differential
heating
exist.

Another
documented
instance
of
MP
lamp
breakage
occurred
in
a
UV­
peroxide
reactor
designed
for
well­
head
treatment
of
tetrachloroethene­
contaminated
ground
water
(
Moss
2002a).
The
UV
reactor
was
positioned
between
the
ground
water
extraction
pump
and
distribution
system
booster
pumps.
The
7­
foot
long
MP
lamp
sleeve
sagged
and
came
into
contact
with
the
lamp
envelope.
The
lamp
envelope
and
lamp
sleeve
broke,
releasing
mercury
to
the
water
in
the
reactor.
The
lamp
failure
triggered
an
alarm,
shutting
down
both
the
ground
water
extraction
and
distribution
system
booster
pumps.
Mercury
liquid
was
found
settled
in
the
bottom
of
the
reactor.
Water
sampling
at
a
nearby
fire
hydrant
detected
mercury
concentrations
below
the
maximum
contaminant
level
(
MCL)
of
2
micrograms
per
liter
(
µ
g/
L)
(
Moss
2002a;
Moss
2002b).

European
drinking
water
utilities
have
an
extensive
history
with
UV
technologies.
Unfortunately,
no
written
documentation
of
lamp
failures
was
identified;
however,
two
instances
of
lamp
breakage
during
UV
disinfection
of
drinking
water
were
noted
by
European
manufacturers
(
Roberts
2000;
Table
N.
2).
In
one
instance,
a
ground
water
well
pump
discharged
gravel
or
stones
into
the
reactor,
resulting
in
a
lamp
break.
A
strainer
was
placed
in­
line
prior
to
the
reactor
to
prevent
any
future
instances.
The
other
documented
case
of
a
lamp
breaking
was
due
to
operator
error.
A
forklift
was
driven
into
an
operating
reactor
and
physically
damaged
the
UV
reactor.
The
event
activated
an
alarm
and
pneumatic
valve
closure,
which
contained
the
contamination
(
Roberts
2000).
In
addition,
there
was
an
incident
in
which
equipment
debris
(
a
bolt
from
the
filter
underdrain)
impacted
a
lamp
sleeve.
Although
the
lamp
sleeve
was
broken,
the
lamp
envelope
remained
intact
and
mercury
was
not
released
because
of
the
immediate
UV
installation
shutdown
and
prompt
operator
response
(
McClean
2001a).
Table
N.
2
summarizes
the
documented
lamp
breakages
discussed
in
this
section.

Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
8
June
2003
Table
N.
2
Mercury
Release
Incidents
Involving
UV
Lamp
Breaks
Identified
Cause
Number
of
Incidents
Description
of
Incident
Debris
5
(
4)
1
Stones
entered
the
reactors
and
impacted
and
broke
the
lamps.

(
1)
2
Gravel
entered
reactor
through
the
booster
pump
and
impacted
and
broke
the
lamp.

Loss
of
Water
Flow
and
Temperature
Considerations
2
(
2)
1
Lamps
were
left
on
and
allowed
to
reach
high
temperatures
(
600
oC)
in
empty
non­
operating
reactors.
Restoration
of
flow
resulted
in
cooler
water
(
20
oC)
breaking
the
lamps.

Operator
Error
1
(
1)
3
Forklift
collided
with
on­
line
reactor
resulting
in
lamp
breakage.

Manufacturer
Design
4
(
1)
1
Applied
power
exceeded
design
rating
of
lamp
(
30kW)
causing
the
lamp
to
burst
from
within.
(
2)
1
Vertical
orientation
of
lamps
resulted
in
differential
heating
and
eventual
cracking
of
lamp
sleeve
as
surrounding
water
cooled
the
submerged
portion
of
lamp
and
the
exposed
portion
of
the
lamp
accumulated
heat.
(
1)
4
High
operating
temperatures
resulted
in
deformation
of
the
lamp
sleeve.
The
lamp
sleeve
sagged
and
on
contact
with
the
lamp
envelope,
both
envelope
and
lamp
sleeve
broke.
1
Survey
of
domestic
water,
wastewater,
and
hazardous
waste
treatment
utilities
(
Malley
2001)
2
European
drinking
water
facilities
(
Roberts
2000)
3
European
brewery
(
Roberts
2000)
4
UV­
peroxide
ground
water
remediation
reactor
(
Moss
2002a)

N.
2.3
On­
Line
Mercury
Release
Response
Plan
On­
line
lamp
breaks
are
rare
occurrences
that
are
preventable
with
appropriate
design
and
operation
of
UV
reactors.
However,
utilities
may
consider
developing
a
mercury
release
response
plan
for
an
on­
line
UV
lamp
break.
The
plan
may
include
the
following
components:

 
Site­
specific
containment
measures
 
Mercury
sampling
and
compliance
monitoring
guidelines
 
Clean­
up
procedures
 
Reporting
requirements
In
the
event
of
an
on­
line
lamp
failure
alarm,
the
UV
reactor
should
be
immediately
shut
down
and
operators
should
attempt
to
determine
the
cause
of
the
alarm.
Unfortunately,
lamp
failure
alarms
or
sensors
cannot
typically
determine
the
cause
of
the
alarm,
whether
it
is
partial
or
complete
breakage
of
the
lamp
sleeve
or
lamp
envelope
(
Kolch
2001)
or
another
problem
unrelated
to
the
lamps.
Thus,
it
is
recommended
that
the
reactor
be
taken
off­
line
when
investigating
the
cause
of
a
lamp
failure
alarm
(
Kolch
2001).

Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
9
June
2003
In
the
event
of
an
on­
line
lamp
break
and
mercury
release,
operators
should
attempt
to
isolate
the
mercury
in
the
reactor
or
downstream.
Utilities
may
install
spring­
return
actuated
valves
with
a
short
closure
time
on
the
reactor
inlet
and
outlet
piping
(
McClean
2001b)
to
isolate
the
mercury.
Given
the
short
residence
time
of
many
MP
reactors,
the
outlet­
side
valve
may
need
to
be
located
quite
a
distance
downstream
so
that
the
valve
has
time
to
close
and
isolate
the
mercury
upstream.
UV
installation
designers
should
evaluate
valve
closure
times
with
respect
to
creating
water
hammer.

Condensed
mercury
may
collect
in
areas
of
low
water
velocity
such
as
the
bottom
of
a
shutdown
reactor,
sump
areas,
or
a
clearwell.
In
addition,
a
strainer
positioned
on
the
reactor
outlet
piping
may
prevent
lamp
fragments
from
entering
the
water
supply
system
(
McClean
2001b;
Srikanth
2001a;
Srikanth
2001b).
The
headloss
associated
with
such
measures
should
be
considered
in
the
hydraulic
profile.
Designers
may
also
consider
installation
of
drains,
vacuum
relief
valves,
and
piping
to
allow
disposal
of
potentially
contaminated
water
in
the
reactor
to
a
waste
container
or
truck.

The
extent
of
containment
provided
by
these
safety
measures
is
unknown.
Utilities
and
designers
should
determine
the
applicability
of
these
isolation
techniques
on
a
site­
specific
basis.

Utilities
should
coordinate
with
their
State
primacy
agency
when
developing
the
following
action
items:

 
Mercury
sampling
plan
 
Sampling
procedures
may
outline
sample
locations,
sampling
frequencies,
and
analysis
methods.
Sample
locations
should
be
chosen
with
consideration
of
where
mercury
may
settle
and
to
assess
the
mercury
concentrations
potentially
reaching
the
consumer.
Sampling
frequencies
should
consider
flowrate,
detention
time,
and
travel
time
to
the
first
potential
consumer.

 
Site­
specific
cleanup
procedures
 
Site­
specific
cleanup
procedures
should
be
incorporated
into
a
utility
process
hazard
analysis
(
PHA).
Issues
to
consider
are
detection
and
disposal
of
isolated
or
condensed
mercury,
potential
disposal
or
treatment
of
contaminated
water,
and
cleanup
responsibilities
(
by
utility
staff
or
contracted
hazardous
materials
team).

 
Reporting
to
State
 
Reporting
may
include
a
description
of
the
release,
estimated
quantity
of
release,
shutdown
or
containment
procedures,
cleanup
or
disposal
methods,
sampling
procedures
(
including
sampling
locations,
frequencies,
and
results).

 
Public
notification
requirements,
if
applicable
 
Revised
public
notification
requirements
(
40
CFR
141.203)
outline
three
tiers
of
public
notification,
depending
on
the
severity
of
the
violation
or
situation.
Exceeding
the
mercury
MCL
of
2
µ
g/
L
is
classified
as
a
Tier
2
notice,
where
public
notification
is
required
within
30
days,
unless
extended
to
90
days
by
the
State
primacy
agency.
Public
notification
requirements
do
not
specifically
address
mercury
releases
due
to
UV
lamp
breakage
where
the
MCL
is
not
exceeded.

Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
10
June
2003
N.
3
Regulatory
Review
This
section
presents
a
review
of
regulations
that
may
apply
to
the
use
or
breakage
of
UV
lamps
containing
mercury
in
water
treatment
plants
(
WTPs).

N.
3.1
Safe
Drinking
Water
Act
The
Safe
Drinking
Water
Act
(
SDWA)
established
a
primary
MCL
of
2
µ
g/
L
for
inorganic
mercury
(
40
CFR
141.62(
b)).
The
required
monitoring
frequency
depends
on
the
water
source
and
the
frequency
of
detections.
Utilities
using
ground
water
sources
are
required
to
sample
once
every
3
years.
WTPs
using
surface
water
sources
are
required
to
sample
annually.
If
mercury
is
detected
above
the
MCL
in
any
ground
water
or
surface
water
utility,
the
utility
must
sample
quarterly.

These
regulations
are
independent
of
the
use
of
UV
disinfection
at
a
facility.
As
discussed
in
section
N.
2.3,
utilities
should
consult
with
their
primacy
agencies
when
developing
a
sampling
plan
for
responding
to
an
on­
line
UV
lamp
break.

N.
3.2
Operator
Health
and
Safety
­
Exposure
Limits
The
Mercury
Study
Report
to
Congress
(
USEPA
1997c)
provides
detailed
information
on
health
effects
associated
with
exposure
to
elemental
mercury
and
mercury
compounds.
Mercury
exposure
to
employees
in
WTPs
falls
under
the
regulatory
authority
of
the
Occupational
Safety
and
Health
Administration
(
OSHA).
The
exposure
limits
set
by
OSHA
focus
on
exposure
by
inhalation.

OSHA
regulations
have
established
permissible
exposure
limits
(
PELs)
for
mercury
compounds
and
organo
alkyls
containing
mercury.
A
PEL
is
a
time
weighted
average
concentration
for
an
8­
hour
workday
during
a
40­
hour
work
week
that
is
not
to
be
exceeded.
When
a
PEL
is
designated
as
a
ceiling
level
(
cPEL),
the
concentration
cannot
be
exceeded
during
any
part
of
the
workday.
PELs
and
cPELs
are
enforceable
standards.
The
National
Institute
for
Occupational
Safety
and
Health
(
NIOSH)
also
publishes
Immediately
Dangerous
to
Life
or
Health
(
IDLH)
concentrations
for
a
variety
of
compounds.
IDLH
concentrations
represent
the
maximum
concentrations
that
one
could
escape
within
30
minutes
without
symptoms
of
impairment
or
irreversible
health
effects.
These
values
are
not
enforceable,
but
can
be
used
as
guidance
for
safety
procedures.
Table
N.
3
outlines
the
PELs,
cPELs,
and
IDLHs
for
mercury
compounds
and
organo
alkyls
containing
mercury.

Table
N.
3
Health
and
Safety
Standards
for
Mercury
Compounds
in
Air
Compound
PEL
(
mg­
Hg/
m3)
cPEL
(
mg­
Hg/
m3)
IDLH
(
mg­
Hg/
m3)

Mercury
compounds
NR
0.1
10
Organo
alkyls
containing
mercury
0.01
0.04
2
NR
­
not
reported.

Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
11
June
2003
In
the
event
of
a
spill,
the
volatilization
and
the
resultant
concentration
of
mercury
in
air
depends
on
the
vapor
pressure
(
0.002
mm
Hg;
Table
N.
4),
air
currents,
temperature,
surface
area/
dispersion
of
mercury
droplets,
and
time.
Calculations
using
the
ideal
gas
law
(
PV=
nRT)
indicate
that
these
levels
may
be
exceeded
if
cleanup
of
the
mercury
spills
does
not
occur;
however,
prompt
response
and
proper
cleanup
procedures
should
prevent
exposure
levels
over
these
standards.

N.
3.3
UV
Lamp
Disposal
Regulations
Lamp
manufacturers
are
required
to
determine
whether
their
products
exhibit
the
toxicity
characteristic
for
mercury
using
a
test
called
the
Toxic
Characteristic
Leaching
Procedure
(
TCLP,
40
CFR
261).
If
the
TCLP
level
of
a
lamp
is
above
the
regulatory
limit
of
0.2
mg/
L,
the
lamp
is
regulated
as
a
universal
hazardous
waste
(
Universal
Waste
Rule,
40
CFR
273)
under
Subtitle
C
of
the
Resource
Conservation
and
Recovery
Act
(
RCRA).
As
such,
these
lamps
should
be
sent
to
a
mercury
recycling
facility
where
the
mercury
is
recovered
and
lamp
components
are
recycled.
Although
some
mercury
lamps
do
not
exceed
the
TCLP
regulatory
level,
utilities
are
encouraged
to
recycle
these
lamps
to
reduce
mercury
loading
to
the
environment.
Some
UV
reactor
and
lamp
manufacturers
will
accept
spent
or
broken
lamps
for
recycling
or
proper
disposal
(
Dinkloh
2001a;
Lienberger
2002;
Gump
2002).
Alternatively,
utilities
should
contact
their
primacy
agency
for
a
list
of
local
recycling
facilities.

N.
4
Additional
Factors
Affecting
Risk
This
section
provides
further
information
that
may
be
helpful
in
evaluating
risk
associated
with
on­
line
lamp
breakage.
The
ultimate
fate
of
mercury
after
a
lamp
is
broken
is
currently
unknown
but
is
expected
to
depend
on
the
following
conditions:

 
Physical
and
chemical
properties
of
mercury
species
in
air
and
water
 
Mercury
behavior
in
operating
UV
lamps
 
Quantity
of
mercury
released
(
type,
age,
and
number
of
broken
lamps)

 
Potential
mercury
reactions
in
water
treatment
plants
and
the
distribution
system
N.
4.1
Physical
and
Chemical
Properties
of
Mercury
Mercury
can
exist
in
three
oxidation
states:
elemental
(
Hg0),
mercurous
(
Hg+
1),
and
mercuric
(
Hg+
2).
Mercury
cycles
between
oxidation
states
as
a
function
of
the
redox
conditions
of
the
surrounding
environment
and
the
availability
of
other
reactive
compounds.

Elemental
mercury
is
a
liquid
at
ambient
temperature
and
pressure;
however,
given
its
high
vapor
pressure
(
Table
N.
4),
elemental
mercury
is
easily
vaporized
at
ambient
temperatures.
Other
physical
and
chemical
properties
of
elemental
mercury
that
affect
its
fate
and
transport
are
outlined
in
Table
N.
4.

Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
12
June
2003
Table
N.
4
Physical
and
Chemical
Properties
of
Elemental
Mercury
(
Merck
&
Co.,
Inc.
1983)

Property
Value
Melting
point
(
oC)
­
38.87
Boiling
point
(
oC)
356.72
Density
(
g/
mL
at
25
oC)
13.534
Solubility
(
g/
L
at
25
oC)
0.061
Vapor
pressure
(
mm
Hg
at
25
oC)
0.002
1
Further
information
regarding
mercury
solubility
in
water
can
be
found
in
Glew
et
al.
1971
N.
4.2
Mercury
Behavior
in
UV
Lamps
It
is
important
to
characterize
the
quantity
and
form
of
mercury
in
an
operating
lamp
because
they
represent
the
starting
point
for
mercury
dispersion,
speciation,
and
reaction
chemistry
in
the
water
system
following
a
lamp
break.
However,
the
quantity
and
form
of
mercury
placed
in
UV
lamps
typically
is
considered
proprietary
information
by
manufacturers
because
these
parameters
affect
the
efficiency,
operation,
and
life
of
the
lamp.
In
general,
the
form
of
mercury
contained
in
a
UV
lamp
is
elemental
mercury
(
LP
and
MP)
or
a
mercury
amalgam
(
LPHO).
An
amalgam
is
an
alloy
of
elemental
mercury
with
another
metal
(
typically
indium
in
lamp
applications)
that
can
be
either
solid
or
liquid
at
room
temperature,
depending
on
the
relative
proportions
of
the
two
metals.
In
operating
lamps,
elemental
mercury
(
from
pure
or
amalgamated
mercury)
is
vaporized
in
the
presence
of
an
inert
gas.
Vapor
phase
mercury
is
excited
and
then
ionized
by
the
energy
transfer
from
the
excited
inert
gas
and
the
supply
of
electrons
generated
from
the
applied
voltage
(
Phillips
1983).
It
is
the
transition
of
mercury
electrons
from
excited
state
back
to
ground
state
that
releases
energy
in
the
wavelength
range
of
the
UV
spectrum.

The
concentration
of
mercury
in
the
vapor
phase
in
LP
and
LPHO
lamps
is
controlled
predominantly
by
temperature.
Manufacturers
of
these
lamps
use
different
methods
to
control
or
maintain
the
temperature
of
the
liquid
mercury
or
mercury
amalgam
to
establish
the
desired
vapor
phase
mercury
concentrations.
Methods
of
controlling
the
temperature
of
mercury
and,
consequently,
the
vapor
pressure
in
LP
and
LPHO
include
using
either
a
mercury
amalgam
attached
to
the
lamp
envelope
(
LPHO
only),
a
cold
spot
on
the
lamp
wall,
or
a
mercury
condensation
chamber
located
behind
each
electrode.
At
typical
LP
and
LPHO
lamp
operating
temperatures,
mercury
remains
predominantly
in
the
liquid
or
solid
amalgam
phase
with
a
small
proportion
in
the
vapor
phase.

MP
lamps
are
dosed
with
elemental
mercury
liquid.
In
operating
MP
lamps,
mercury
is
primarily
present
in
the
vapor
phase
due
to
high
operating
temperatures
(
600
to
900
oC;
Table
2.1)
that
cause
all
liquid
elemental
mercury
to
volatilize
(
Phillips
1983).
In
order
to
control
the
concentration
of
vapor
phase
mercury,
manufacturers
strictly
control
the
amount
of
mercury
dosed
or
added
to
the
lamp
during
production.
This
is
different
from
the
LP
and
LPHO
lamps
where
an
excess
of
mercury
is
placed
in
the
lamp
and
only
a
portion
of
the
elemental
mercury
enters
the
vapor
phase.
Once
elemental
mercury
enters
the
vapor
phase,
mercury
ionization
in
a
MP
lamp
occurs
the
same
way
as
in
LP
or
LPHO
lamps.

Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
13
June
2003
The
relative
proportion
of
mercury
in
the
liquid/
amalgam
phase
and
the
vapor
phase
becomes
important
when
an
operating
lamp
breaks
in
water.
Vapor
phase
elemental
or
ionized
mercury
may
be
released
as
very
fine
particles.
These
particles
may
more
readily
dissolve
in
water
as
opposed
to
condensed
liquid
or
amalgamated
mercury
that
settles
in
low
velocity
areas.

In
addition
to
these
functional
mercury
interactions,
Altena
(
2001)
reported
reactions
of
vapor
phase
mercury
with
fluorescent
lamp
components,
such
as
the
glass
bulb,
glass
stems,
coatings,
and
the
emission
material
(
electrodes).
This
process
results
in
the
embedding
of
mercury
in
lamp
components
and
the
accumulation
of
mercury­
containing
deposits,
such
as
mercury
oxide,
on
the
internal
lamp
envelope
surface.
Altena
(
2001)
theorized
that
mercury
reactions
with
UV
lamp
components
would
be
comparable
to
fluorescent
lamps.
These
deposits
represent
approximately
2
to
15
percent
of
the
total
mercury
present
in
a
lamp
as
calculated
from
Altena
(
2001).
After
breakage,
these
deposits
are
available
to
dissolve
in
water;
however,
mercury
oxide
has
a
low
solubility
in
water
(
Merck
&
Co.
1983).

Figure
N.
1
outlines
the
expected
forms
of
mercury
in
an
operating
lamp.
Note
that
all
liquid
elemental
mercury
will
volatilize
in
an
operating
MP
lamp,
leaving
no
mercury
in
the
liquid
phase.
Also,
amalgams
are
only
used
in
LPHO
lamps.

Figure
N.
1
Mercury
Speciation
In
Operating
UV
Lamps
N.
4.3
Quantity
of
Mercury
in
Lamps
The
amount
of
mercury
in
a
UV
reactor
is
a
function
of
the
type
of
lamp,
the
number
of
lamps
in
a
reactor,
and
the
number
of
reactors.
Mercury
content
within
lamps
depends
on
type
(
LP,
LPHO,
or
MP),
length,
and
power
rating.
Although
mercury
content
data
are
specific
to
manufacturer
and
lamp,
lamps
with
higher
pressures,
power
ratings,
and
lengths
typically
contain
more
mercury.
Table
N.
5
summarizes
the
quantities
of
elemental
mercury
dosed
into
lamps
during
manufacturing
according
to
a
confidential
manufacturer
survey
and
published
literature
values.

Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
14
June
2003
Table
N.
5
Elemental
Mercury
Content
in
UV
Lamps
Mercury
Content
(
mg
per
lamp)
Lamp
Type
Electrical
Power
Rating
(
W)
Phillips
(
1983)
Clear
et
al.
(
1994)
Manufacturer
Survey
LP
15­
70
"
a
single
drop"
1
202
5­
50
120­
260
NR
263,
364
150
LPHO
400
NR
75.5
NR
1000
NR
250
NR
MP
1­
25
kW
1.4
­
14.5
mg/
cm5
NR
200­
400,
0.3
­
7
mg/
cm
length,
7.9
mg/
cm
length
1
Phillips
1983
2
75
W
mercury
vapor
lamp
3
175
W
mercury
vapor
lamp
4
250
W
mercury
vapor
lamp
5
mg
per
cm
of
lamp
length,
reported
lamp
lengths
are
6­
300
cm
(
Primarc
Limited
2001)
NR
­
Not
Reported
N.
4.4
Quantification
of
Mercury
in
Example
UV
Installations
This
section
illustrates
example
calculations
of
the
amount
of
mercury
contained
in
hypothetical
UV
installations.
Two
UV
reactor
manufacturers
established
design
parameters
for
three
treatment
flowrates
(
0.18,
3.5,
and
210
million
gallons
per
day
(
mgd))
with
a
specified
water
quality
and
design
dose
(
Table
N.
6).
Design
parameters
included
the
number
of
lamps
needed
to
obtain
a
dose
of
40
mJ/
cm2
and
the
total
number
of
reactors
for
each
of
the
three
design
flows.
Calculations
assume
50,
150,
and
400
mg
of
mercury
per
LP,
LPHO,
and
MP
lamp,
respectively.
Utilities
should
use
site­
specific
UV
installation
information
to
determine
quantities
because
mercury
content
varies
with
lamp
type
and
manufacturer.

Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
15
June
2003
Table
N.
6
Mercury
Quantity
in
Example
UV
Installations1,2
Design
Flow
(
mgd)
Average
Flow
(
mgd)
Lamp
Type
Average
Number
of
Reactors
Average
Number
of
Lamps
(
per
reactor)
Total
Hg
in
UV
Installation3
(
g)

LP
1
2
0.1
LPHO
1
1
0.2
0.18
0.054
MP
1
1
0.4
LPHO
1
30
4.5
3.5
1.4
MP
1
4
1.6
LPHO
6
72
64.8
210
120
MP
6
7
16.8
1
UV
Dose
=
40
mJ/
cm2
2
Water
quality
criteria:
UVT
=
89%
(
A254
=
0.05
cm­
1),
Turbidity
=
0.1
NTU,
Alkalinity
=
60
mg/
L
as
CaCO3,
Hardness
=
100
mg/
L
as
CaCO3
3
Values
given
represent
the
amount
of
elemental
mercury
dosed
in
lamps
during
manufacturing.

N.
4.5
Fate
of
Mercury
After
Release
The
previous
sections
define
the
quantity
and
form
of
mercury
in
an
operating
lamp
and
thus
define
the
starting
point
for
the
investigation
of
the
fate
of
mercury
in
the
water
system.
Unfortunately,
little
documentation
exists
on
the
fate
of
mercury
in
WTPs
or
distribution
systems.
The
few
case
studies
that
do
exist
are
mainly
in
the
wastewater
industry
and
focus
primarily
on
removal
of
influent
mercury
by
the
following
wastewater
treatment
processes:

 
Primary
sedimentation
(
Lester
1983;
Firk
1986;
Balogh
and
Liang
1995;
Goldstone
et
al.
1990;
Oliver
and
Cosgrove
1974)

 
Activated
sludge
(
Gilmour
and
Bloom
1995;
Lester
1983;
Chen
et
al.
1974;
and
Wu
and
Hilger
1985)

 
Conventional
treatment
process
(
Mugan
1996;
Balogh
and
Liang
1995;
Bodaly
et
al.
1998)

Much
of
the
knowledge
about
mercury
and
its
potential
fate
in
water
systems
is
derived
from
studies
performed
in
natural
environments.
It
is
expected
that
this
knowledge
of
mercury
cycling
within
the
natural
environment
can
be
applied
to
mercury
dynamics
within
a
WTP
and
distribution
system
where
environmental
conditions
are
largely
controlled
and
remain
fairly
stable.
However,
WTPs
employ
a
number
of
chemicals
that
are
not
typically
found
in
natural
environments.
No
studies
were
identified
on
influence
and
reaction
of
mercury
with
coagulants,
polymers,
corrosion
inhibitors,
ammonia,
strong
oxidants,
and
other
disinfectants
(
e.
g.,
chlorine
and
ozone).
This
subsection
projects
mercury
reactions
within
a
WTP
and
distribution
system
based
on
documented
mercury
reactions
in
the
natural
environment.

Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
16
June
2003
N.
4.5.1
Potential
Mercury
Reactions
in
Water
Treatment
Plants
and
Distribution
System
Liquid
phase
elemental
mercury
is
considerably
denser
than
water
(
specific
gravity
=
13.6;
Table
N.
4)
and
does
not
readily
dissolve
in
water.
Therefore,
liquid
elemental
mercury
and
mercury
amalgams
may
settle
in
areas
of
low
water
velocity,
thereby
providing
an
option
for
early
containment
and
removal.
For
example,
in
cases
where
mercury
was
released
from
other
water
treatment
equipment
(
such
as
manometers,
flow
instrumentation,
or
pump
seals),
mercury
was
found
to
have
condensed
and
settled
in
the
clearwell.
However,
the
amount
of
mercury
recovered
relative
to
the
amount
of
mercury
released
is
unknown
(
Cotton
2002).
Kolch
(
2001)
monitored
the
mercury
concentrations
in
a
50­
liter
batch
reactor
following
the
destruction
of
one
LPHO
lamp
(
approximately
150
mg­
Hg).
Mercury
concentrations
reached
approximately
2.5
µ
g/
L
in
water.
Amalgamated
mercury
was
found
settled
on
the
bottom
at
the
reactor
(
Dinkloh
2001b).
However,
it
was
not
reported
whether
all
the
150
mg
of
mercury
present
in
the
operating
lamp
was
recovered
with
the
amalgam
or
accounted
for
in
the
aqueous
phase.

Also,
Kolch
(
2001)
did
not
determine
whether
the
source
of
aqueous
phase
mercury
was
dissolved
mercury
from
the
amalgam
or
vapor
phase
mercury
present
in
the
lamp
prior
to
when
it
was
broken.
This
issue
may
become
important
when
considering
the
aqueous
behavior
of
mercury
following
a
MP
lamp
break.
Mercury
in
a
MP
lamp
is
predominantly
in
the
vapor
phase
during
operation.
It
is
unknown
how
the
vapor
phase
mercury
will
react
with
the
water.
Vapor
phase
elemental
and
ionized
mercury
may
become
very
fine
particles
when
contacting
the
water
as
opposed
to
liquid
or
amalgamated
mercury
that
settles
in
low
velocity
areas.

Depending
on
the
age
of
the
lamp,
mercury­
containing
deposits,
such
as
mercury
oxide,
may
accumulate
on
the
inner
surface
of
the
lamp
envelope
in
all
lamp
types.
Dissolution
of
the
deposits
would
result
in
additional
ionized
mercury
entering
the
water.
Prompt
response
to
a
lamp
break
would
include
removal
of
lamp
fragments;
therefore,
the
compounds
on
the
lamp
fragments
should
not
have
the
opportunity
to
enter
the
WTP.

Once
in
the
water,
aqueous
(
dissolved)
elemental
and
ionized
mercury
are
expected
to
cycle
between
phases
and
oxidation
states
as
determined
by
temperature,
pH,
organic
carbon
concentration,
minerals
and
inorganic
species,
and
dissolved
oxygen
level.
The
mercurous
ion
(
Hg2
+
2)
is
formed
by
the
oxidation
of
elemental
mercury
or
the
reduction
of
the
mercuric
ion
(
Hg+
2).
The
mercurous
ion
is
capable
of
bonding
with
inorganic
constituents;
however,
it
does
not
bind
with
organic
compounds.
Hg2
+
2
is
rarely
stable
under
typical
environmental
conditions
and
is
readily
reduced
to
Hg0
or
oxidized
to
Hg+
2.

Inorganic
reactions
involving
the
mercuric
ion
(
Hg+
2)
include
binding
with
inorganic
ligands
such
as
chloride,
hydroxide,
and
sulfide.
Considering
sulfide
is
present
in
anoxic
environments,
mercury
sulfide
(
HgS)
is
not
expected
to
form
in
a
water
treatment
environment.
Reactions
of
Hg+
2
with
chloride
and
hydroxide
resulting
in
mercuric
chloride
and
mercuric
hydroxide
compounds
depend
on
the
pH
and
chloride
concentrations
(
Beckvar
et
al.
1996).

Additional
discrepancies
arise
in
the
comparison
of
mercury
reactions
and
fate
in
a
drinking
water
environment
versus
the
natural
environment
when
organic
carbon
concentrations
and
existing
microbial
populations
are
considered.
The
mercuric
ion
(
Hg+
2)
is
the
only
oxidation
state
of
mercury
capable
of
association
with
organic
compounds
such
as
phenyl
and
methyl
Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
17
June
2003
groups.
The
resultant
organic
compounds,
commonly
known
as
methylated
mercury,
have
different
chemical,
physical,
and
toxicological
properties
than
inorganic
mercury
and
offer
more
cause
for
concern
due
to
toxicity
and
bioaccumulation
properties
(
Beckvar
et
al.
1996).
Methylation
of
mercury
to
form
methyl
and
dimethyl
mercury
is
primarily
a
biological
process
involving
sulfate­
reducing
bacteria
although
it
can
also
occur
abiotically.
The
extent
to
which
methylation
occurs
depends
on
the
availability
of
Hg+
2
and
the
presence
of
an
appropriate
microbial
population.
Methylation
rates
are
higher
under
anoxic
conditions,
low
pH,
elevated
temperatures,
and
high
organic
matter
concentrations
(
USEPA
1997c).
Therefore,
even
though
Hg+
2
may
be
present
in
the
water
column,
all
of
the
above
factors
oppose
the
occurrence
of
methylation
in
a
drinking
water
environment.

Another
divergence
of
a
water
treatment
environment
from
the
natural
environment
is
the
presence
of
treatment
chemicals
such
as
coagulants,
strong
oxidants,
polymers,
corrosion
inhibitors,
and
ammonia.
Seigneur
(
1994)
researched
the
reaction
chemistry
of
mercury
with
inorganic
species
and
strong
oxidants
such
as
chlorine
and
ozone
in
the
aqueous
and
gas
phases
that
are
present
in
the
atmosphere.
Ionized
mercury
can
form
inorganic
compounds
with
chloride
and
hydroxide
ions.
Depending
on
reactant
concentrations,
these
compounds
may
be
present
in
the
aqueous
phase
and
as
solid
precipitate.
Also,
based
on
reduction
oxidation
potentials,
it
is
possible
that
the
oxidation
of
elemental
mercury
would
occur
in
the
presence
of
chlorine
and
ozone,
forming
mercury
ions
and
thereby
increasing
the
solubility
of
elemental
mercury.

Further
research
and
investigation
is
necessary
to
determine
the
mechanisms
of
any
potential
mercury
reactions.
Figure
N.
2
outlines
this
preliminary
assessment
of
mercury
speciation
and
reaction
in
a
drinking
water
environment.

Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
18
June
2003
Figure
N.
2
Potential
Reactions
of
Mercury
in
a
Drinking
Water
Environment
(
Compounds
released
from
a
broken
lamp
are
in
boldface
type.)

Mercury
(
methylated
and
ionic)
sorption
to
dissolved
and
particulate
organic
matter
is
commonly
found
in
natural
environments
(
Beckvar
et
al.
1996;
USEPA
1997c).
This
observation
was
also
made
in
water
and
wastewater
treatment
plant
studies,
where
mercury
became
incorporated
into
coagulant
flocs
and
activated
sludge
waste,
respectively
(
Logsdon
1973;
Gilmour
and
Bloom
1995;
Lester
1983;
Chen
et
al.
1974;
Wu
and
Hilger
1985).

If
UV
disinfection
of
raw
water
is
used
and
a
UV
lamp
breaks,
the
mercury
could
potentially
be
removed
within
the
WTP.
Logsdon
(
1973)
investigated
the
efficiency
of
mercury
removal
in
conventional
WTPs.
Bench­
scale
laboratory
tests
indicated
that
inorganic
mercury
was
removed
via
coagulation,
softening,
adsorption
on
turbidity,
powdered
activated
carbon
(
PAC)
adsorption,
and
granular
activated
carbon
(
GAC)
column
adsorption.

N.
5
Summary
and
Conclusions
The
risk
associated
with
a
mercury
release
to
the
water
system
depends
on
many
factors.
More
research
is
needed
to
close
the
knowledge
gap
that
exists
regarding
the
fate
of
mercury
in
a
drinking
water
environment
following
a
UV
lamp
break.
The
influence
of
treatment
chemicals
such
as
oxidants,
disinfectants,
and
coagulants
is
largely
unstudied.
Likewise,
dispersion
and
transport
of
mercury
through
a
WTP
and
distribution
system
has
yet
to
be
evaluated.
Although
Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
19
June
2003
these
issues
are,
at
present
time,
largely
unknown,
there
are
procedures
and
actions
that
can
be
taken
to
reduce
or
mitigate
mercury
release
caused
by
UV
lamp
breakage.

For
the
purposes
of
this
appendix,
lamp
breakage
was
defined
as
fracture
of
the
lamp
sleeve
and
the
lamp
envelope.
This
was
further
divided
into
off­
line
and
on­
line
breaks.
Off­
line
lamp
breaks
typically
occur
during
storage
or
handling
and
cause
small
spills
(<
2g).
Small
spills
should
be
contained,
cleaned
up,
and
disposed
of
properly.

On­
line
lamp
breaks
occur
while
the
UV
reactor
is
in
operation.
There
have
been
reported
incidents
of
on­
line
UV
lamp
breaks
associated
with
impact
from
debris,
loss
of
water
flow,
temperature
differentials,
faulty
electrical
equipment
and
design,
and
procedural
errors.
However,
on­
line
lamp
breaks
are
a
rare
occurrence
and
are
largely
preventable
with
appropriate
design,
operation,
maintenance,
and
operator
care.
The
following
engineering
and
administrative
methods
have
been
proposed
that
may
help
prevent
UV
lamp
breakage:

 
Screens,
baffles,
or
low
velocity
collection
areas
prior
to
the
reactor
influent
to
prevent
entrance
of
debris
 
Temperature
and
flow
sensors
and
alarms
to
detect
critical
conditions
and
shut
the
UV
reactors
down
 
Surge
analysis
to
determine
if
water
hammer
may
be
a
potential
problem
 
Comprehensive
training
and
maintenance
program
In
the
event
of
a
mercury
release,
the
following
engineering
controls
are
additional
precautions
that
may
aid
in
the
containment
and
collection
of
mercury:

 
Strainers
and
low
velocity
collection
areas
downstream
of
the
reactor
 
Isolation
valves
activated
by
an
alarm
to
attempt
to
isolate
potentially
contaminated
water
The
extent
of
containment
and
prevention
provided
by
these
measures
is
unknown.
Utilities
and
designers
should
consider
the
applicability
of
these
isolation
techniques
on
a
sitespecific
basis.
Utilities
should
consult
with
their
State
primacy
agency
in
the
development
of
standard
operating
procedures,
clean­
up
procedures,
and
reporting
requirements
in
preparation
for
a
potential
UV
lamp
break
and
mercury
release.
It
is
recommended
that
a
utility
prepare
a
mercury
release
response
plan
to
address
these
issues.
This
plan
should
address
compliance
with
the
SDWA,
OSHA
health
and
safety
standards,
and
RCRA
universal
waste
rules.
Utilities
are
encouraged
to
recycle
or
return
all
mercury­
containing
lamps
to
mercury
re­
generating
facilities
or
the
lamp
manufacturer.

Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
20
June
2003
N.
6
References
Aquafine
Corporation.
2001.
CSL
Series:
Installation,
Maintenance,
and
Operation
Manual
Part
No.
115­
1.

Altena,
F.
W.,
J.
B.
J.
van
Overveld,
and
H.
Giller.
2001.
Technological
advances
in
disinfection
lamps
leading
to
more
compact
UV
sources.
Presented
at
the
First
International
Congress
on
Ultraviolet
Technologies,
Washington,
DC,
June
14­
16.

Balogh,
S.
and
L.
Liang.
1995.
Mercury
pathways
in
municipal
wastewater
treatment
plants.
Water,
Air,
and
Soil
Pollution
80:
1181­
1190.

Beckvar,
N.,
J.
Field,
S.
Salazar,
and
R.
Hoff.
1996.
Contaminants
in
aquatic
habitats
at
hazardous
waste
site:
mercury.
Hazardous
Materials
Response
and
Assessment
Division
of
the
National
Oceanic
and
Atmospheric
Administration
Technical
Memorandum
NOS
ORCA
100.
Seattle,
WA.

Bodaly,
R.
A.,
J.
W.
M.
Rudd,
and
R.
J.
Flett.
1998.
Effect
of
urban
sewage
treatment
on
total
and
methyl
mercury
concentrations
in
effluents.
Biogeochemistry
40:
279­
291.

Cairns,
B.
2000.
Comments
on
USEPA
questions
about
the
risk
of
using
UV
disinfection
technologies
in
the
potable
water
industry.
Trojan
Technologies
Inc.,
London,
Ontario,
Canada,
May
24.

Chen,
K.
Y.,
C.
S.
Young
and
N.
Rohatgi.
1974.
Trace
metals
in
wastewater
effluents.
J.
Water
Poll.
Control
Fed.
46:
2663.

Clear,
R.,
and
S.
Berman.
1994.
Environmental
and
health
aspects
of
lighting:
mercury.
Journal
of
the
Illuminating
Engineering
Society
138­
156.

Cotton,
C.
2002.
Confidential
communication
with
major
water
municipality.
June.

Dinkloh,
L.
2001a.
Wedeco
Ideal
Horizons.
Personal
communication
with
Jennifer
Hafer,
Malcolm
Pirnie,
regarding
UV
reactors.
Tucson
AZ,
May
10.

Dinkloh,
L.
2001b.
Wedeco­
Ideal
Horizons.
Personal
communication
with
Jennifer
Hafer,
Malcolm
Pirnie,
regarding
UV
reactors.
Tucson
AZ,
August
10.

Dinkloh,
L.
2001c.
Wedeco­
Ideal
Horizons.
Email
to
Christine
Cotton
regarding
UV
negative
pressure.
Tucson
AZ,
August
15.

Firk,
W.
1986.
Heavy
metals
in
sewage
and
occurring
sludges
 
the
balance
of
three
sewage
treatment
plants,
edited
by
J.
N.
Lester,
R.
Perry
and
R.
M
Sterritt.
Lisbon:
International
Conference
Chemicals
in
the
Environment.

Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
21
June
2003
Gilmour,
C.
C.
and
N.
S.
Bloom.
1995.
A
case
study
of
mercury
and
methylmercury
dynamics
in
a
Hg­
contaminated
municipal
wastewater
treatment
plant.
Water
Air
and
Soil
Pollution
80:
799­
803.

Glew,
D.
N.
and
D.
A.
Hames.
1971.
Aqueous
nonelectrolyte
solution:
mercury
solubility
in
water.
Canadian
Journal
of
Chemistry
49:
3114­
3118.

Goldstone,
M.
E.,
C.
Atkinson,
P.
W.
W.
Kirk,
and
J.
N.
Lester.
1990.
The
behavior
of
heavy
metals
during
wastewater
treatment
III:
mercury
and
arsenic.
The
Science
of
the
Total
Environment
95:
271­
294.

Gump,
D.
2002.
Severn
Trent.
Personal
communication
with
Jennifer
Hafer,
Malcolm
Pirnie,
Inc.,
regarding
UV
reactors.
Tucson
AZ,
November
6.

Kolch,
A.
2001.
Assessing
the
risk
of
mercury
release
from
ultraviolet
discharge
lamps
in
drinking
water.
Wedeco
AG
Water
Technology.

Lester,
J.
N.
1983.
Significance
and
behaviour
of
heavy
metals
in
waste
water
treatment
processes.
I.
sewage
treatment
and
effluent
discharge.
Science
of
the
Total
Environment
30:
1­
44.

Lienberger,
J.
2002.
Trojan
Technologies.
Personal
communication
with
Jennifer
Hafer,
Malcolm
Pirnie,
Inc,
regarding
UV
reactors.
November
5.

Logsdon,
G.
and
J.
Symons.
1973.
Mercury
removal
by
conventional
water­
treatment
techniques.
Journal
American
Water
Works
Association
554­
562.

Malley,
J.
2001.
Addressing
concerns
over
mercury
releases
from
UV
lamps.
IUVA
News
3,
no
3:
28­
29.

McClean,
J.
2001a.
Hanovia
Corporation.
Personal
communication
with
Doug
Owen
and
Christine
Cotton,
Malcolm
Pirnie,
Inc.,
regarding
UV
reactors.
Slough,
England,
April
9.

McClean,
J.
2001b.
Hanovia
Corporation.
Response
to
memo
by
Christine
Cotton,
Malcolm
Pirnie,
Inc.,
regarding
UV
reactors.
Tucson
AZ,
May
4.

Merck
&
Co.,
Inc.
1983.
The
Merck
Index
10th
Edition,
edited
by
M.
Windholz.
Rahway,
NJ.

Moss,
L.
2002a.
Personal
communication
with
Jennifer
Hafer,
Malcolm
Pirnie,
Inc.,
regarding
UV
reactors.
Tucson
AZ,
April
29.

Moss,
L.
2002b.
Personal
communication
with
Jennifer
Hafer,
Malcolm
Pirnie,
Inc.,
regarding
UV
reactors.
Tucson
AZ,
November
18.

Mugan,
T.
J.
1996.
Quantification
of
total
mercury
discharges
from
publicly
owned
treatment
works
to
Wisconsin
surface
waters.
Water
Environment
Research
68:
2,
229­
234.

Proposal
Draft
Appendix
N.
UV
Lamp
Breakage
Issues
UV
Disinfection
Guidance
Manual
N­
22
June
2003
Oliver,
B.
G.,
and
E.
G.
Cosgrove.
1974.
Efficiency
of
heavy­
metal
removal
by
a
conventional
activated­
sludge
treatment
plant.
Water
Research
8,
no
11:
869­
874.

Phillips,
R.
1983.
Sources
and
Applications
of
Ultraviolet
Radiation.
Academic
Press
Inc.,
New
York,
New
York.

Primarc
Limited.
2001.
September
5,.
www.
primarc.
com.

Roberts,
A.
2000.
Hanovia
Corporation.
Letter
to
Dan
Schmelling,
EPA.
April
10.

Seigneur,
C.,
J.
Wrobel,
and
E.
Constantinou.
1994.
A
chemical
kinetic
mechanism
for
atmospheric
inorganic
mercury.
Environmental
Science
and
Technology
28:
1589­
1597.

Srikanth,
B.
2001a.
Aquafine
Corporation.
Letter
to
James
Malley,
UNH,
regarding
accidental
UV
lamp
breakage
and
potential
scenarios
and
prevention
strategies.
June
13.

Srikanth,
B.
2001b.
Aquafine
Corporation.
Personal
communication
with
Jennifer
Hafer,
Malcolm
Pirnie,
Inc.,
regarding
UV
reactors.
Tucson,
AZ,
June
21.

Swaim,
P.
D.,
M.
A.
Morine,
R.
G.
Brauer,
M.
A.
Neher,
J.
L.
Gebhart,
and
W.
D.
Bellamy.
2002.
Operating
data
from
Henderson's
full­
scale
UV
disinfection
facility.
Proceedings
from
the
Water
Quality
Technology
Conference,
November
10­
14,
Seattle,
WA.

U.
S.
Environmental
Protection
Agency.
1992.
Characterization
of
Products
Containing
Mercury
in
Municipal
Solid
Waste
in
the
United
States,
1970­
2000.
Office
of
Solid
Waste,
Washington,
D.
C.

U.
S.
Environmental
Protection
Agency.
1997a.
Mercury
­
Emergency
Spill
&
Release
Facts.
EPA
540­
K­
97­
004,
OSWER
9378.0­
10FS,
PB97­
963405.
Office
of
Emergency
and
Remedial
Response,
Washington,
D.
C.

U.
S.
Environmental
Protection
Agency.
1997b.
Mercury
emissions
from
the
disposal
of
fluorescent
lamps.
Office
of
Solid
Waste,
Washington,
D.
C.
June
30.

U.
S.
Environmental
Protection
Agency.
1997c.
Mercury
Study
Report
to
Congress.
EPA­
425/
R­
97­
004:
2­
14.
Office
of
Air
Quality
Planning
and
Standards
and
Office
of
Research
and
Development,
Washington,
D.
C.

Walitsky,
P.
Philips
Lighting.
2001.
Personal
communication
with
Jennifer
Hafer,
Malcolm
Pirnie,
Inc.,
regarding
UV
reactors.
Tucson
AZ,
May
14.

Wu,
J.
S.
and
H.
Hilger.
1985.
Chemodynamic
behavior
of
mercury
in
activated
sludge
process.
Am.
Inst.
Chem.
Eng.
81:
109.

Proposal
Draft
Appendix
O.
Case
Studies
This
appendix
will
be
included
in
the
final
draft
when
more
information
is
available.

UV
Disinfection
Guidance
Manual
O­
1
June
2003
Proposal
Draft
Appendix
P.
Validation
Protocol
Calculator
Tool
The
validation
protocol
described
in
Chapter
4
and
Appendix
C
of
this
guidance
manual
involves
several
calculations
to
determine
the
log
inactivation
credit
achieved
during
validation
of
a
UV
reactor.
For
this
protocol,
a
safety
factor
calculated
from
uncertainties
and
known
variability
associated
with
UV
reactors,
monitoring,
and
validation
methods
is
used
to
relate
the
reduction
equivalent
dose
(
RED)
demonstrated
during
validation
to
the
UV
dose
required
to
achieve
a
specified
log
inactivation
credit
(
Table
1.4).
EPA
developed
a
spreadsheet
that
enables
a
user
to
input
information
associated
with
the
uncertainty
of
validation
and
monitoring
and
calculate
the
safety
factor
and
resulting
target
RED.
The
calculator
was
used
to
develop
Tier
1
RED
targets
and
can
be
used
to
apply
the
Tier
2
approach.
(
See
section
C.
4.10.2
for
a
description
of
the
Tier
2
approach,
including
the
safety
factor
calculation.)

The
Microsoft
Excel
®
spreadsheet
contains
the
following
five
worksheets:

 
Instructions
 
provides
step­
by­
step
instructions
for
entering
data
into
the
"
RED
Bias",
"
Polychromatic
Bias",
and
"
Safety
Factor"
worksheets
and
executing
macros
to
calculate
safety
factor
and
resulting
target
RED.

 
RED
Bias
 
calculates
RED
bias
from
input
of
Chapter
1
UV
dose
requirements
and
inactivation
and
RED
measured
during
validation.

 
Polychromatic
Bias
 
calculates
the
polychromatic
bias
for
medium­
pressure
UV
systems
using
spectral
data
on
the
lamp
output,
sleeve
UV
transmittance,
water
UV
transmittance,
and
sensor
response.

 
Safety
Factor
 
calculates
safety
factor
from
RED
bias,
polychromatic
bias,
and
expanded
uncertainty
associated
with
reactor
validation
and
monitoring.

 
Default
Data
 
contains
assumed
data
for
calculating
the
polychromatic
bias;
alternatively,
the
user
can
provide
validation
testing
data
as
specified
in
the
instruction
worksheet.

UV
Disinfection
Guidance
Manual
P­
1
June
2003
Proposal
Draft
