►
From YouTube: NTSB Public Meeting Day 1 Disc 2 of 2 March 1, 2011
Description
NTSB Public Meeting Day 1 March 1, 2011
Disc 2 of 2
Natural Gas Pipeline Explosion and Fire
San Bruno, CA September 9, 2010
A
A
A
Mr
ruiz,
I'm
sorry
before
you
begin.
If
I
could
just
remind
remind
all
the
parties
there
are
some
lights
up
on
the
on
the
table
and
you'll
see
their
green,
yellow
and
red,
and
we're
going
for
five
minute
rounds
for
all
the
parties,
as
well
as
the
board.
Members
and
you'll
get
a
yellow
about
a
minute
when
you
have
a
minute
left
in
a
red
when
you've
hit
your
time,
and
so
I'd
prefer
for
you
to
self
police.
But
if
not,
I
will
jump
in
and
also
I'd
like
to
remind
the
parties.
A
C
C
First
question
I
have
relates
to
procedures,
particularly
those
that
could
affect
scada
systems,
and
I
was
just
curious
what
your
procedures
call
for
when
you're
doing
work
that
could
affect
the
segata
system,
whether
it
was
testing
new
software
or
doing
electrical
work.
That
could,
you
know,
affect
the
scada
system
in
some
way.
C
I
I
just
point
out
that
fimsa
had
done
a
advisory
a
number
of
years
ago,
that
that
relates
to
scada
systems
and
work
that
could
influence
scada
systems
not
immediately
and
directly
relevant,
but
clearly
any
work
that
you
do
that
affects
your
ability
to
monitor
your
system
in
any
way
is
something
you
want
to
take
seriously.
As
I'm
sure
you
do
next
question
really
is
just
the
five-year
pressure
increase
that
you
talked
about
as
a
practice.
I
was
just
curious
if
pg
e
continues
that
practice
at
this
time.
E
C
F
F
There's
there's
two
nuances
to
the
question.
I
guess
the
first
easy
answer
is:
we
are
going
to
be
deploying
remote
and
automated
control
valves
in
our
system,
we're
evaluating
where
that's
going
to
have
the
greatest
benefit
or
effect
as
it
relates
to
our
high
consequence
area,
we're
in
the
process
of
working
with
films.
Other
regulators
to
help
inform
the
the
new
design
criteria
that
we
intend
to
propose.
F
Having
said
that,
we're
going
to
pilot
aggressively
this
summer,
the
placement
of
some
of
some
valves
just
to
make
sure
that
we're
clear
about
process
as
it
relates
to
the
specific
root
cause
of
the
failed
pipe
in
question.
The
we're
going
to
wait
until
we
have
all
the
facts
in
the
conclusion
for
probable
cause
from
the
ntsb
before
we
commence
any
mitigative
action
as
it
relates
to
that.
F
I
In
answer
to
your
question,
we
have
already
added
additional
monitoring
sites
as
a
result
of
some
of
the
reconfiguration
efforts
we
had
to
take
on
to
continue
to
serve
the
the
customer
base
during
the
winter
months,
and
we
are
looking
at
improvements
in
the
area
in
our
study
of
adding
remote
control
valves
to
our
system.
That
would
allow
us
to
do
a
more
effective
isolation.
Sooner.
H
E
So
what
we're
doing
during
that
pressure
increase
is
exactly
what
you
say
is
we
are
monitoring
the
pressure
to
a
very
specific
value,
and
this
is
on
these
lines
that
may
have
not
reached
this
five-year
maximum
operating
pressure
during
their
normal
course
of
operations,
because
they
could
have
reached
their
on
their
own
just
by
meeting
customer
demand.
So
that's
so
we
are
looking
exactly
at
pressure.
H
Maybe
you
could
describe
to
me
how
what
is
the
indication
of
the
ability
of
the
pipeline
to
maintain
the
pressure?
Then
I
guess
would
be
my
question:
are
you?
Are
you
looking
for
a
leak
or
how?
How
might
the
pipe
behave
in
that
situation?
That
would
either
satisfy
your
interest
or
give
you
rise
to
additional
concerns,
so.
E
The
pressures
that
we
are
raising
the
pipeline
to
are
the
maximum
allowable
operator
press
operating
pressure
or
something
less,
and
I
should
say
that
the
maximum
allowable
operating
pressure
is
not
it's.
It's
it's
a
pressure
with
significant
safety
factors
built
in
already.
You
know
the
pipeline
is
tested
to
a
much
greater
pressure
when
it's
manufactured
at
the
mill
there's
also
reductions
that
are
taken.
You
know,
restrictions
based
on
what
we
call
class
location
so
raising
that
pressure
up
to
the
pressures
through
these
planned
increases
is
not
a
cause
of
of
concern.
F
H
A
J
Okay,
madam
chair
panel,
I'm
paul
klan
and
I'll
be
representing
the
public
utilities.
Commission,
just
a
quick
follow-up
on
that.
You've
now
been
asked
questions
I
think
by
everybody
about
the
five-year
policy
and
we've
heard
about
the
criteria
that
you
use
to
select
the
lines
that
you
apply
it
to.
We've
heard
what
you're
looking
for
when
you're
doing
that
I'll
call
it
a
test,
and
we've
heard
that
you're
not
doing
it
anymore.
J
E
So
through
the
five-year
maximum
operating
pressure
increases
if
the
pipeline
has
not
seen
that
five-year
maximum
operating
pressure,
I
think
it's
probably
maybe
best
to
use
an
example
to
illustrate
the
point
if
you
have
a
maximum
allowable
operating
pressure
of
it
of
say,
400
pounds
in
the
pipeline,
but
you
operate
it
over
the
course
of
the
next
five
years
at
300
pounds.
If,
if
it
has
not
raised
through
the
course
of
operations
to
400
pounds,
the
pressure
is
now
300
pounds.
E
K
J
E
J
J
Paragraph
and
one
of
the
things
that
you're
saying
here
in
this
paragraph
is
that
there
is
an
alternative
to
the
ratcheting
down
effect
that
you
that
you
mentioned,
and
that
is
doing
extra
testing
of
a
pipeline.
That
goes
above
a
certain
limit.
So
this
was
a
choice
by
pg
e,
to
avoid
two
things:
a
ratcheting
down
of
the
maop
and
or
the
extra
hydro
testing
that
the
section
would
call
for
is
that
right.
E
J
And
I
didn't
mean
to
use
a
word
that
would
put
you
on
the
spot
like
a
void,
but
I
still
want
to
understand
that
the
purpose
of
the
policy
was
was
in
pg
e's
view
that
it
was
necessary
to
keep
the
the
maximum
pressure
at
the
former
level
that
you
had
to
do
that
or
you
would
have
had
to
do.
The
hydra
testing
is
that
right.
J
J
I
want
to
to
know
what
information
you
had
before
september,
9th
just
in
your
day-to-day
operations
in
administering
the
integrity
management
program.
What
were
your
sources
of
information
about?
What's
under
the
ground?
How
did
you
verify
those
sources
of
information,
and
did
you
have
any
program
for
benchmarking?
That
information,
for
example,
as
as
there
are
issues
that
arise
in
other
states,
whether
you're,
aware
of
that
and
and
you're
able
to
factor
that
in.
E
J
And
what
about
benchmarking
with
other
with
other
states
with
other
pipeline
operators?
Are
you
in
a
dialogue
or
is
pg
e
in
a
dialogue
with
other
folks
who
are
operating
systems
like
this,
especially
now
that
the
infrastructure
is
aging,
it's
aging
for
everybody
at
more
or
less
the
same
pace
now,
and
I
know
that
the
industry
itself
is
is
doing
a
lot
of
thinking
about
this.
How
do
you
benchmark
your
activities,
and
I
see
that
my
light
is
red
so
I'll?
F
So
we
have
a
fairly
structured
benchmarking
program
that
we've
instituted
as
it
relates
to
both
gas
and
electric.
We
use
aga
and
other
fora
to
to
gather
data.
Having
said
that,
getting
asset
level
data
is
is
a
little
bit
more
challenging
and
that's
more
ad
hoc.
We
will
get
system
performance
data
and
we
probably
get
more
robust
data
at
the
distribution
layer
of
our
system
than
we
do
at
the
transmission
layer
of
the
system.
So
that's
an
opportunity
for
improvement.
G
Thank
you
very
much.
I've
have
questions
from
mr
lee
earlier
this
morning.
During
the
technical
discussion
you
referenced
industry
references
when
talking
about
automatic
and
remote
control,
shutoff
valves
and
you
specifically
stated
that
those
studies
directly
impacted
and
discussed
injury
and
property
damage
and
concluded
that
the
damage
was
done
in
the
first
30
seconds.
Is
that
correct.
G
L
G
L
G
Thank
you,
mr
slipsager.
Earlier
this
morning
you
also
mentioned
in
your
conversation,
references
to
the
ability
to
impact
flow
at
milpita
station
and
impact
flow
at
martin
station.
Had
you
utilized
that
those
opportunities
to
reduce
flow
at
both
stations?
Do
you
have
an
idea
how
long
it
would
have
taken
before
the
gas
flow
would
have
been
down
to
zero
at
the
rupture
site.
I
G
E
Sure
what
eyes?
What
I'd
started
to
say
was
that
there
were
two
two
incidents
on
on
pipelines
and
other
operators,
companies
that
they
had
both
been
hydro
tested,
as
well
as
inline
inspected,
using
a
variety
of
tools
and
using
including
a
crack
tool
which
is
capable
of
of
assessing
for
seam
defects,
and
they
both
failed
along
the
long
seam.
And
so
while
these
were
liquid
events
and
I'm
not
trying
to
draw
the
correlation
between
liquid
operations
and
natural
gas
because
they
are
uniquely
different.
E
G
G
B
Thank
you.
As
I
look
across
the
table,
we
have
a
senior
vice
president
and
a
vice
president.
We
have
a
couple
of
managers.
We
have
a
supervisor
so
who
of
those
would
be
the
the
most
junior
in
the
management
chain.
Would
that
be
the
mr
kashmir
kazmi,
who
I've
practiced
that
name
many
times,
and
I
still
messed
it
up.
My
apologies.
B
D
B
B
Thank
you.
I
was
just
trying
to
get
a
feel
for
how
rigorous
the
the
training
and
the
standardization
program
was.
So
thank
you.
This
can
be
directed
towards
either
of
you,
gentlemen.
Does
pg
e
have
a
confidential
instant
reporting
system
whereby
employees
can
report
safety
issues
that
they
may
be
concerned
of.
I
F
B
B
D
D
F
Expand
on
that,
we
we
operate
in
an
environment
with
with
a
bargaining
unit,
so
we
have
two
bargaining
units
and
the
rules
with
regard
to
discipline
are
defined
in
the
context
of
those
those
agreements,
and
so
we
have
processes
that
we
would
have
to
comply
with,
which
are
positive
discipline
in
nature
and
there's
a
a
process
for
that,
and
we
have
also
a
similar
process,
although
not
negotiated
for
our
management
or
or
non-marketing
unit.
Employees.
F
M
M
G
G
Done
the
analysis
of
the
30
inch
pre-1962
pipe
done
that
validation
effort.
We
are
undergoing
a
complete
validation
of
the
gis
system,
for
if
I
may,
class
locations
three
and
four
as
well
as
class
locations,
one
and
two
in
high
consequence
areas,
and
then
we
are
committed
to
continuing
that
effort
throughout
the
entire
transmission
network.
G
Several
different
ways,
one
of
which
is
just
going
through
the
process.
If
you
don't
mind,
we
are
validating
all
of
the
information
within
gis,
with
the
engineering
specifications
in
the
project
folders
and
then
we
are
also
using
validation,
digs
radiography
in
certain
areas,
and
we
are
doing
inline
camera
inspections
on
several
areas
to
confirm
that
data
as
well.
M
Let
me
change
subjects
here
for
a
moment.
A
lot
of
discussion
this
morning
about
how
long
it
took
to
get
somebody
out
to
close
the
manually,
operated
valves
and
given
that
you've
got
a
system
where
you
can't
simply
close
a
valve
without
having
some
repercussions
elsewhere
in
the
system,
assuming
that
the
two
valves
that
were
closed
manually.
M
I
Believe
I
believe
it
would
have
been
at
least
15
minutes,
because
when
you
look
at
the
event
timeline,
you
know
the
the
rupture
occurred.
I
totally
told
it
at
6
11
pm
the
alarm
went
off
in
the
control
room
near
6
15
and
by
there
were
actually
actions
taken
in
in
between
time
where
employees
were
already
moving,
but
gas
control
made
its
callouts
just
around
6
30
2
p.m.
So
it
would
have
taken
that
long.
If
the
operators
had
done
their
analysis,
they
could
have
closed
valves
remotely
at
that
point.
Okay,.
M
I
A
A
I
M
G
If
you
could
discuss
the
ups
lost
power,
the
pressure
change
and
if
you
would
describe
the
troubleshooting
process
that
went
on
the
decision
making
timing
response
outcome,
are
there
standard
troubleshooting
protocols
in
the
control
room
and
give
us
a
sense
if
there
are
scenarios
that
are
practiced
so
give
us
a
feel
for
when
these
things
happen?
What
was
going
on
there
within
all
those
different.
D
In
this
case,
it
was
not
the
ups
that
failed.
The
ups
was
out
of
service
for
several
months.
Prior
to
that
downstream.
From
the
ups
we
had
24
volt
power
supplies
that
fed
instrumentation
and
while
we
still
had
120
volt
120
volt
power
out
of
the
temporary
ups's,
the
24
volt
output
from
the
power
supply
is
what's
been
lost.
That
night.
G
And
and
actually
I'm
less
interested
in
the
specifics
of
this
one
and
understanding
your
generalized
procedures
when
you
have
an
outage,
a
pressure
change,
an
outage
that
happens,
that's
going
to
ripple
through
what
are
your
procedures
to
respond
to
that?
So
is
there
a
standard
protocol
of
troubleshooting?
Is
there
a
timeline
for
that?
Do
you
practice
those
scenarios?
Do
you
actually
sort
of
look
back
and
determine
we
need
to
do
that
faster
differently,
etc.
D
We
don't
practice
specific
scenarios
because
I
don't
think
there's
any
two
identical
scenarios
when
something
fails
in
the
field.
What
we
do
do
is
we
go
through
the
training
for
our
maintenance
personnel
and
the
training
would
normally
involve
typical
troubleshooting
steps
in
case
of
failures
in
the
system,
be
it
power,
loss
or
loss
of
a
specific
transmitter
or
any
component
of
the
control
system.
F
I
think
we
may
be
talking
a
little
bit
past
each
other.
I
think
mark
is
referring
to
the
work
activity
that
was
on
site
by
the
construction
crew.
They
have
a
method
of
procedure
and
they
have
a
clearance
document
that
identifies
step
by
step.
What
they
should
do
to
execute
the
work
that
they've
got
to
do
to
the
extent
that
work
in
some
shape
way
or
form
fails.
They
have
to
go
into
a
local,
troubleshooting
sectionalizing
process
to
identify
kind
of
where,
where
the
issues
are.
F
Having
said
that
at
a
broader
level-
and
I
think
keith
could
speak
to
that,
the
control
room
has
a
different
set
of
protocols
as
they
look
at
situational
awareness
in
what's
going
on
with
regard
to
the
system
at
large
and
how
they
would
react
to
the
extent
they
have
lost,
control
or
or
visibility
to
a
given
terminal
or
station.
And-
and
maybe
you
can
talk
a
little
bit
about
how
you
would
troubleshoot.
I
I
N
Thank
you.
I
have
a
question
that
I'm
not
sure
is
appropriate
for
you
and
I'm
not
sure
who
else
it
might
be
appropriate
for
so
I'm
going
to
tell
you
first.
This
is
about
the
pups,
I'm
just
curious
about
pups.
Do
they
are
they
made
by
taking
a
full
length
pipe
and
cutting
it,
or
they
made
some
other
way
and-
and
my
ultimate
question
is
how
what
do
you
know
about
how
how
the
pups
were
made
in
this
particular
piece
of
pipe.
F
F
They
had
the
authority
to
take
sections
that
were
good,
that
passed
and
and
joined
them
and
and
then
used
them
as
as
filling
part
of
the
order.
There
were
limits
to
how
much
of
that
could
be
a
part
of
any
given
order,
and
and
so
that
that's
the
way
you
could
get
joined
sections
of
pipe
into
the
system.
A
A
I
A
I
A
Okay,
so
you
know,
unfortunately,
in
our
business
we
look
at
accidents
and
and
seconds
sometimes
are
very
precious
and
in
a
situation
where
you're
describing
it
takes
15
minutes
to
actually
determine
what
next
steps
need
to
be
taken.
When
there's
a
catastrophic
rupture
like
this,
I'm
trying
to
understand
what
are
the
you
know,
minimum
expectations
I
mean:
are
we
going
into
red
alert
when
they
see
this
this
happening,
or
is
everyone
all
hands
on
deck
available
supervisors?
I
Ma'am,
my
gas
control
operators
take
these
situations
very
seriously.
We
have
a
priority
system
in
our
in
our
skate
alarms
and,
in
this
case,
a
low
low
alarm
or
in
the
case
of
the
high
high
alarms
that
it
occurred
earlier,
they're
both
the
highest
priority
and
require
the
operator
action
to
move
forward
as
quickly
as
possible,
and
I
have
five
five
operators
in
my
control
room
that
are
watching
these
systems.
A
I
A
A
A
Just
for
the
for
the
general
public,
I
think
kind
of
understanding,
usually
when
we
think
about
something
failing
if
something
goes
wrong,
if
something
loses
power
like
a
grade
crossing,
the
fail-safe
position
would
be
for
it
to
fail
down
or
fail
closed
in
this
situation.
We're
hearing
that
when
there's
a
loss
of
power
or
interruption
of
a
signal,
the
position,
the
fail-safe
position
for
pipelines
is
for
it
to
fail
open.
Can
you
explain
to
me
why
you
wouldn't
have
a
fail,
closed
situation
with
valves
when
they
lose
signal
or
power.
D
D
P
F
Certainly
from
a
local
distribution
company,
a
low
pressure
system,
uncontrolled
low
pressure
is
as
dangerous
to
the
public
safety
as
an
overpressure
situation.
So
what
we're
trying
to
do
is
maintain
pressure
within
a
tolerance
of
safety
and
and
what
we
have
are
these
two
forms
of
control
and
regulation.
Managing
that
so
an
uncontrolled
shutdown
can
be
as
hazardous
to
public
safety
in
many
cases
as
as
a
rupture.
A
F
It
could
be
if
it
didn't
have
precisely
the
exact
logic
controlling
it
so
that
it
was
operating
with
all
of
the
right
inputs
in
making
the
right
decision,
so
it
will
be
insufficient
just
to
have
at
least
in
a
network
system,
a
valve
that
that
works
off
of
simplistic
or
incomplete
inside
information.
You
really
need
to
have
full
situational
awareness
so
that
you
know
you're
not
creating
any
unintended
consequences
by
shutting
down
a
line.
A
K
K
E
E
So,
in
the
first
steps
of
the
process
for
say,
inline
inspection,
all
job
files
are
are
pulled
for
the
creation
of
what
we
call
our
pipeline
features
list
to
see
what
what
upgrades
may
be
needed
on
the
pipeline
for
the
purposes
of
external
corrosion
direct
assessment.
We
are
specifically
focused
on
those
values
where
we,
those
assumed
values
where
we
have
conservatively
assumed
a
value
in
the
fields
and
we
pulled
job
files.
According
with
those
values.
K
E
We
are
constantly
trying
to
improve
our
our
systems
and
our
programs,
and
we
we
do
this
continuously
through
the
integrity
management
program
for
external
corrosion
direct
assessment.
We
are
not
pulling
a
comprehensive
set
of
job
packages
and,
as
mr
dauben
spoke
to
earlier,
we
are
in
the
process
of
of
comprehensively
reviewing
our
entire
gis
system
and
populating
those
with
source
job
documents.
K
I
I
don't
recollect
that
I
said
that
I
think
if,
if
you're
speaking
to
my
my
what
I
have
said
this
morning,
I
said
that
it
would
be
proper
for
it
and
it
is
required
for
the
operators
to
do
the
analysis
of
the
system
and
then
do
start
to
perform
call
outs
as
quickly
as
possible.
If
that's
warranted,
I
don't
pg
e.
My
gas
control
operators
do
not
call
9-1-1.
K
There
are
three
techniques
you
mentioned:
one
is
pressure,
testing
other
is
in-line
inspection
and
the
third
is
external
corrosion,
which
has
been,
I
guess,
applied
to
line
132..
Can
you
please
clarify?
Will
the
inline
inspection
identify
all
defects
if
there
are
any
in
the
pipeline
segment
that
you
use
the
tool.
K
E
So
the
assessment
methods-
and
this
may
be
a
better
topic
to
explore
in
the
in
the
integrity
management
panel,
but
the
inspection
methods
already
are
designed
to
address
the
threats
identified
in
the
pipeline.
So
they
don't
look
at
all
threats.
They
look
at
the
threats
identified
on
the
pipeline
and
I
think
this
issue
is
better
explored
in
the
integrity
management
panel.
K
E
So
when
we
perform
external
corrosion
direct
assessment,
we
dig
on
the
most
severe
signals.
If
you
will
so
we
are
digging
where
we
have
any
concern.
So
the
the
logic
in
terms
of
external
corrosion
direct
assessment
is
that
anything
left
in
the
pipeline
is
less
than
what
you've
directly
examined.
The
pipe
condition
for
thank.
K
E
C
Thank
you
very
much.
Just
have
a
couple
of
quick
questions
if
you'll
allow
the
first
one,
because
I
think
several
people
were
asking
questions
on
this.
If
you'll
allow
me
touch
on
emergency
response
plans,
some
some
people
were
asking.
What
do
your
policies
procedures
require
you
to
do
in
regard
to
contacting
local
communities
emergency
responders?
C
F
Correct,
having
said
that,
another
form
of
another
area
of
improvement
that
we're
undergoing
right
now
is
to
better
integrate
all
of
our
emergency
response
programs
to
improve
the
training
and
exercising
with
first
responders
and
we're
trying
to
establish
a
standard
in
conjunction
with
our
sister
utility
in
the
south
of
california,
so
that
we
have
one
way
of
of
managing
training,
communications
and
and
development.
With
that
with
all
first
responders.
C
Great
thank
you.
The
last
couple
of
quick
questions
relate
to
remote,
controlled
valves,
automatic,
shut
off
valves,
some
background
with
integrity
management
myself,
and
so,
while
mr
hee
did
this
address
the
notion
about
a
policy
and
the
company's
policy
as
it
relates
to
these
valves.
The
rule
also
requires
you
to
document
your
through
an
engineering
assessment,
your
decisions
about
what
preventive
and
mitigative
actions
to
implement.
Do
you
know
whether
you're
doing
that,
as
it
relates
to
the
valves
prior
to
the
failure.
C
F
H
I
have
to
admit
to
a
little
bit
of
confusion
about
the
existing
communication
plan
in
place
relative
to
a
significant
emergency
such
as
we
experienced
in
san
bruno.
My
question
is
whether
you
are
considering
or
whether
you
can
consider
designate
an
employee,
such
as
a
district
supervisor,
for
emergency
responders
to
contact
in
the
event
of
a
significant
emergency.
F
We're
looking
at
the
entire
process,
we
we
subscribe
to
the
incident
command
system
with
regard
to
management
of
any
significant
event.
Having
said
that,
we
know
that
we
we
can
improve
the
manner
in
which
we
we
make
sure
that
the
incident
command
on
site
knows
who
to
go
to
who
to
work
with
from
a
utility
standpoint,
and
that's
part
of
the
improvement
effort
that
we've
that
we've
employed.
N
I
Low
pressure
in
in
a
system
may
allow
us
to
go
into
an
uncontrolled
shutdown,
and
if
we
have
low
pressure
that
allows
the
pilot
lights
in
a
house
to
extinguish
extinguish,
then
you
have
the
potential
for.
If
gas
were
to
re-enter
that
house,
through
migration
of
this
low
pressure
around
the
system
before
it's
all
evacuated
out
of
the
pipe
it
could
it
could,
the
gas
could
intrude
back
into
the
house
and
create
an
explosive
situation.
F
Thank
you,
member.
I
think
it.
I
think
a
lot
happens.
We
do
share
information
and
I
think
it
is
from
where
I
sit
very
it's
a
little
bit
ad
hoc.
We
come
together
with
regard
to
industry
associations.
We
share
data
and
information
when
we
have
very
significant
large
events
such
as
this
horrible
tragedy.
I
would
say
we
could
make
improvements.
F
We
we're
not
quite
at
what
I
would
call
the
impo
level
from
a
nuclear
industry
standpoint
where
we
are
sharing
detailed
failure
or
near
misinformation
across
the
board,
so
that
we
treat
you
know
a
failure
in
in
iowa
as
seriously
as
as
a
failure
in
california
that
we
all
learn
from
from
every
incident.
So
I
think
that's
that's
an
area
that
we
could
collectively
improve
in.
B
Thank
you,
and
I
know
on
day
three
we'll
be
talking
about
industry-wide
technology,
and
it's
my
hope
that
we
can
also
ask
questions
along
the
same
types
of
questions
to
the
industry
organizations
not
necessarily
related
to
technology,
but
with
respect
to
sharing
of
best
practices.
Madam
chairman,
no
further
questions.
F
A
D
I
We
earlier
showed
a
slide
that
showed
the
the
large
amount
of
data
that
exists
from
a
scada
perspective
downstream
of
milpita
station
towards
the
san
francisco
peninsula,
and
there
is
a
site:
that's
within
a
very
close
distance.
Several
miles
I
mean
less
than
I
believe
less
than
five
miles,
and
it
didn't
register
that
sort
of
pressure.
A
F
Madam
chairwoman,
we
haven't
concluded
all
of
the
protocols
and
the
decision,
trees
and
they're
they're,
I
would
say,
they're
fairly
mature,
but
they're
not
complete
yet
and
we're
looking
to
to
vet
them
with
industry
experts
kind
of
nationwide.
Having
said
that,
we
know
we
want
to
focus
on
hca
geography.
F
We
know
we
need
to
do
significant
upgrade
with
regard
to
control
logic,
so
that
the
people
in
the
control
room
floors
can
manage
many
many
more
streams
of
data
intelligently.
We
can't
have
them
just
look
at
more
alarms.
We
need
to
have
tools
that
help
parse
and
interpret
those
alarms
so
that
good
action
is
taken.
So
we
have,
we
have
work
in
a
number
of
areas,
but
the
real
focus
is
going
to
be
in
high
consequence
areas
where
we've
got
population
and
and
to
give
us
more
visibility
inside
and
control.
With
regard
to
our
system,.
F
A
F
G
L
Certainly,
my
name
is
frank:
dolby,
I'm
a
supervising
engineer
in
the
integrity
management
group
at
pg
e
I've
been
employed
by
pg
e
for
27
years
in
a
variety
of
gas
engineering
positions
for
the
last
seven
years.
I've
been
supervisor
for
the
ili
group
and
for
the
last
year
also
including
external
corrosion
direct
assessment,
and
I
hold
a
bachelor's
degree
of
civil
engineering
from
georgia,
tech.
O
I've
had
a
variety
of
engineering
responsibilities
from
the
engineering
level
as
well
as
supervision
level.
I've
also
had
a
number
of
construction
level
responsibilities
as
well
in
general.
My
responsibilities
as
integrity
management
director
is
responsibilities
for
the
integrity
management
programs,
as
well
as
the
maintenance
processes
associated
with
pipelines.
G
A
L
N
L
L
I
don't
have
an
exact
mileage
figure
for
that.
I
can
tell
you
in
general,
pg
e's
gas
transmission
system
was
not
built
with
pig
ability
in
mind
prior
to
the
mid
90s,
when
rules
went
into
effect
requiring
that
all
new
pipelines
built
be
made
pickable
and
thus
anytime,
we
implement
an
in-line
inspection
project.
L
Yes,
we
are:
we've
had
plans
in
place
for
some
time
to
continue
to
upgrade
various
sections
of
our
pipeline
system
to
make
them
capable
of
of
of
running
in-line
inspection
tools
that
definitely
predates
the
the
incident
in
in
san
bruno,
and
we
will
continue
to
upgrade
pipelines
such
that
we
have
an
option
of
running
either
inline
inspection
tools
or
potentially
doing
ecda
or
hydrated
tests,
but
at
least,
if
they're
upgraded.
We
have
the
option
of
running
those
kind
of
tools.
L
L
L
N
O
There
was
no
triggering
of
the
items
that
would
make
it
considered
to
be
unstable,
manufacturing
threat
for
some
of
the
segments
that
did
not
include
segment.
180
was
not
one
of
those
segments.
There
were
no
triggering
of
that.
By
that
I
mean
there's
no
record
of
a
leak
on
a
longstand
or
that
the
maximum
operating
pressure
had
been
exceeded.
N
O
O
They
are
tools
that
are
compatible
with
each
other,
but
they
do
different
things
and
they
tell
you
different
things.
A
hydra
test
requires
putting
water
into
your
pipeline
and
on
vintage
pipelines.
That's
a
concern,
because
if
the
pipe
isn't
figurable
we're
concerned
that
getting
the
water
out
would
be
difficult
and
even
with
modern
day
dryers
we're
not
100,
it's
not
100
dry,
so
you
would
leave
water
behind
and
the
reason
you
leave
water
buying
is
the
changing
diameters.
O
The
the
partially
blocking
excuse
me
valves
that
mr
dhoby
alluded
to,
as
that
poly
pig
goes
through
it,
and
the
pig
is
basically
a
very
stiff.
Rubber-Coated
sponge
shaped
like
a
cylinder
as
it
tries
to
negotiate
its
way
around
those
pinch
points.
You
expect
water
to
be
left
behind
it.
We
don't
like
leaving
water
behind.
That's
a
concern
for
internal
corrosion.
It's
also
a
serious
impact
to
our
customers
to
take
them
out.
So
unless
we
have
to
assess
that
way,
hydrotest
is
not
not
the
tool
used.
Ili
again,
as
mr
doby
stated,
you
have
many.
O
These
vintage
pipelines
have
issues
associated
with
them.
Ili
would
be
a
step
up
from
hydrotest,
for
example,
because
the
hydrotest,
when
you
pass
the
test,
all
it
tells
you
is
you
pass
the
test.
It
doesn't
tell
you
that
you
may
have
some
corrosion
cells,
for
example,
that
were
small
enough
to
not
be
failed
by
the
hydro
test,
but
still
in
your
pipeline,
an
ili
would
tell
you
that
you
still
have
flaws.
You
need
to
go
look
at
and
where
to
look
at
them
and
da
also.
N
O
O
So
the
integrity
management
program
is
based
on
high
consequence
areas
and
high
consequence
areas
calculated
as
as
discussed
earlier
in
a
couple
different
ways,
but
it
is
relative
to
population.
Essentially,
a
high
consequence
area
is
an
area
of
of
high
density
around
the
pipeline
or
an
area
around
the
pipeline.
O
N
N
O
You
can
be
brief,
so
risk
as
we
understand
it
in
the
industry.
We
all
use
a
relative
risk
ranking,
which
is
the
likelihood
that
a
threat
should
have
should
cause
a
problem
times.
The
consequence,
that's
what
risk
is
and
when
we're
told
what
an
hca
is,
what
a
high
consequence
area
is.
That
means
high
consequence
or
consequence
has
already
been
defined,
and
really
the
only
thing
we're
addressing
is
the
likelihood
portion
of
that
equation.
F
If
I
could,
if
I
could
have
just
maybe
have
to
expand
a
little
bit
the
role
that
population
consequence
is
really
driven
by
three
three
considerations:
impact
on
population
impact
on
the
environment
and
impact
on
reliability.
By
far
and
away
the
heaviest
weighted
component
is
the
impact
and
population
of
that
consequence
of
failure.
Calculation,
so
population
bears
heavily
on
the
on
the
value.
N
N
O
O
There
is
an
analog
equation
that
was
spoken
of
earlier
this
morning
that
uses
different
categories
such
as
third-party
damage,
corrosion
outside
force,
those
types
of
things
there
are
committees
of
subject
matter,
experts
that
meet
to
put
the
waiting
on
them
to
establish
that
equation
and
also
to
incorporate
things
that
have
come
up
through
attendance
and
conferences
or
papers
from
dot.
That
kind
of
thing.
N
O
There's
validation
digs
both
on
the
inline
inspection
side,
as
well
as
the
direct
assessment
side
that
requires
us
to.
Essentially
you
come
up
initially
with
your
hypothesis
or
your
evaluation
of
where
you
think
your
concerns.
Are
you
go
excavate
them
you
validate
that,
and
then
you
go
excavate
some
other
areas
that
you
don't
think
you
have
concerns
with,
and
you
make
sure
that
what
you
see
in
the
ground
is
what
you
expected
to
see.
Based
on
your
analysis,.
O
N
O
O
There
are
various
ways
of
assessing
individual
procedures
and
ways
of
assessing
what
we
call
risk
management
procedure
number
six,
which
is
the
over
arching
procedure
for
how
we
will
manage
the
program
that
that
particular
one
has
been
audited
by
third
party
experts.
It's
also
in
the
process.
All
of
our
procedures
following
san
bruno
have
been
now
in
the
process
of
third
party
consultants,
reviewing
the
entire
all
the
procedures
evaluating
relative
decode,
evaluating
relative
to
industry,
consensus
and
we'll
also
be
making
recommendations
that
go
beyond
either
code
or
industry.
Consensus.
O
O
In
review
the
procedures,
the
technical
teams
they
get
together,
they
review
not
just
the
waiting,
but
they
also
review
the
procedure.
If
there's
no
changes
that
need
to
be
made,
they
won't
necessarily
meet,
but
that's
done
by
subject
matter
experts.
It's
done,
there's
a
monthly
meeting
of
the
team
where
issues
are
brought
up
where
schedules
are
brought
up,
concerns
that
are
brought
up.
O
O
So
rmp6,
because
it's
the
overarching
document
would
be
the
maximum
that
would
go
to
the
vice
president
of
gas,
transmission
and
distribution.
The
risk
management
instructions
which
are
more
detailed,
specific
levels
associated
with
tasks
within
the
integrity
management
team.
Those
could
be
approved
at
the
supervisor
level
and
the
risk
management
procedures
outside
of
rmp6
could
be
approved
at
the
manager's
level.
L
P
You
mentioned
the
types
of
inspections
and
tests
available
to
detect
a
flaw
of
the
type
that
caused
the
fracture
of
line
132
on
september
9th,
you
mentioned
the
hydro
test,
inline
inspection
and
with
respect
to
inline
inspection
mention
a
crack
tool
between
you
and
mr
fassett.
You've
also
described
the
challenges
of
doing
a
hydro
test
on
line
132.
P
L
Well,
one
of
the
options
that
we
have
already
in
process
of
pursuing
and
have
pursued
to
validate
that
there's
an
inside
weld
on
the
pipe
is
to
actually
open
up
the
pipe
and
then
do
a
camera
survey
to
validate
that
they're
they're.
Actually,
the
pipe
is
welded
on
the
inside.
If
it
was,
it
was
purchased
to
be
double
submerged
arc,
welded
pipe
in
regards
to
the
different
types
of
crack
tools.
L
Those
do
exist
and
and
could
be
employed
in
certain
situations
around
the
transmission
system.
One
of
the
big
limitations
on
those
with
the
present
technology
is
that
they
can
only
inspect
one
one
diameter
at
a
time
and
thus,
depending
on
how
much
footage
you
have
at
a
single
diameter
that
limits
the
length
of
inspection
you
can
do
in
one
in
one
inspection.
P
With
respect
to
running
a
camera
through
the
line
again,
what
limitations
would
that
present
to
you?
I
mean
you
mentioned
it?
It
would
detect
whether
there
was
an
inside
weld,
but
beyond
that
are
you
able
to
does?
Would
it
have
the
resolution
to
pick
up
other
types
of
defects
like
flaw,
cracks
or
perhaps
dents
or.
L
No
sir,
the
the
camera
inspection
is
basically
just
a
a
visual.
You
know
giving
the
opportunity
to
visually
inspect
the
inside
to
confirm
that
there
actually
isn't
and
weld
on
the
inside
of
the
pipeline,
but
outside
of
any
very
significant
defects,
it
would
not
provide
detailed
information
regarding
cracks.
B
O
O
The
technology
that's
was
just
recently
supported
or
released,
was
a
product
called
explorer
ii,
and
that
was
supported
also
through
the
department
of
transportation,
fimsa
organization
supported
that
development.
These
are
instruments
that
the
camera
alone,
besides
being
able
to
identify
whether
you
have
a
long
seam,
that's
missing
a
weld.
What
we
found
when
we
did
the
investigation
as
part
of
this
ntsb.
P
O
The
tethered
type
is
available
and
widely
in
use,
currently
the
swimming
type
or
the
self-propelled
type.
We
believe
we'll
have
one
by
may
that
could
be
used
on
this
project.
The
other.
The
one
thing
I
wanted
to
add
quickly.
I'm
sorry
is
what
we
found
by
using
the
cameras
we
found
labeling
on
the
inside
of
the
pipe
which
we
understand
has
come
from
the
mill.
They
would
label
it
in
a
specific
sequence
associated
with
the
american
petroleum
institute
standard
for
labeling
right
and
by
collecting
that
we
learn
more
and
more.
P
L
F
O
P
O
Of
the
three
major
categories
that
those
22
threats
I
mentioned
earlier,
fall
under
there's
time
dependent,
which
is
categorized
as
internal
external
or
stress,
corrosion,
cracking
and
there's
stable,
understable
there's
several
one
of
them
is
manufacturing
defects.
If
the
defect
is
considered
to
be
stable,
it
is
not
expected
to
grow
over
the
expected
life
of
the
pipeline
to
failure.
This
particular
flaw
is
something
that
I've
been
in
this
business
for
a
couple
of
decades.
O
I
never
would
have
expected
to
see
a
diesel
long
seam,
not
just
missing
a
weld
on
a
four
foot
can
but
the
two
four
foot
cans.
Next
to
it.
Since
the
metallurgical
report
has
come
out,
I've
had
colleagues
from
all
across
the
industry
who
have
my
experience
and
in
some
cases
twice
as
much
time
that
were
equally
surprised
to
see
this.
P
P
O
O
The
manufacturing
threats
is
categorized
under
typically
understable,
unless
you
know
something
that
would
make
them
unstable,
at
which
point
direct
assessment
would
not
be
the
methodology
we
would
use.
We
have
experience
with
a
pipeline
up
north,
where
this
was
the
situation,
and
we've
used
a
crack
tool
to
inspect
that
pipeline.
P
O
Unstable,
so
the
way
that's
defined
is
is
if
we've
had
a
leak
on
a
long
seam,
specifically
the
guidelines
or
the
guidance
on
that
is,
if
there's
a
leak
on
a
long
seam
and
it's
select
seam
corrosion,
we
would
call
that
something
that
needs
to
be
inspected
for
the
manufacturing
threat.
If
we
have
a
leak
on
the
long
sim,
regardless
of
what
caused
it,
we
would
say
that
that
has
a
manufacturing
threat.
Therefore,
we
would
not
use
direct
assessment
on
it.
That
was
the
case
of
the
example
I
used
up
north.
O
But
if
I
may
add,
oh
sure,
quickly,
I'm
sorry,
this
is
an
area
that,
especially
from
the
surprise
of
what
we
saw
with
what
failed
this
pipeline.
What
we
believe
may
have
failed
this
pipeline,
that
the
concept
of
manufacturing
of
vintage
pipes,
what
we
know
about
them,
what
were
their
processes
1948
was
when
we
believe
this
pipe
was
created.
That
was
when
the
technology
was
invented.
Also,
we
think
we
need
to
look
further
into
those
vintage
processes.
O
There
was
one
I
believe
in
2007
and
then
in
2009
we,
it
wasn't
an
audit
necessarily.
We
had
asked
a
risk
management
expert
to
come
in
and
look
at
what
we
would
need
to
do
to
move
from
the
current
relative
risk
evaluation
equations
that
we
use
as
many
in
industry
do
to
a
probabilistic
methodology
that
mr
salas
referred
to
earlier.
O
F
Mr
trainer,
what
I
would
do
would
would
be
to
reiterate
what
what
mr
fassett
had
said
earlier,
but
at
a
broader
level
where
we
would
be,
I
would
be
interested
in,
and
I
think
the
leadership
organization
in
terms
of
are
we
are
we
continue.
Are
we
meeting
the
program
objectives
as
it
relates
to
execution,
so
we
have
as
defined
by
by
the
objectives
with
regard
to
integrity
management,
the
need
to
to
complete
the
first
lap,
as
it
were,
of
assessment
by
december
17
2012..
F
F
We
are
relying
on
input
from
our
regulator
that
comes
and
audits
us
or
looks
at
the
quality
of
our
program.
Third
parties.
Looking
at
our
program
are
we
acting
on
the
findings
that
they
identify
as
we
improve
the
program,
we
recognize
that
this
was
something
that
was
new
and
introduced
to
the
industry
and
in
2004
and
and
so
we
we
expect
the
process
to
be
one
of
continuous
improvement.
So
we
look
at
it
from
a
variety
of
vectors
to
assess
the
effectiveness
of
it.
F
Obviously,
we're
we're
looking
to
manage
a
risk
profile
over
time
down
and-
and
I
think
one
of
one
of
the
issues
that
that
we
see
with
a
relativistic
model
said,
is
that
you
know
we're
self-assessing,
so
you
know
we
need.
We
want
a
more
objective
view
as
as
we
as
we
look
at
ourselves.
We
think
the
probabilistic
transition
will
offer
us
that.
F
I
rely
on
the
management
team
to
to
be
looking
at
performance.
We
have
monthly
performance
meetings,
as
relates
to
pipeline
performance
to
the
the
manner
in
which
we're
scheduling,
work
that
need
for
for
corrective
or
safety
related
programs
connected
with
the
integrity
management
program.
So
we're
we're
watching
to
to
make
sure
that
we're
we're
hitting
the
the
targets
that
we've
got
with
regard
to
corrective
actions.
F
F
Well,
given,
quite
frankly,
the
tragedy
and
and
the
huge
loss,
I
think
we
we
need
to
stand
back
and
question
ourselves.
You
know
from
from
you
know
from
the
beginning,
and
I
think
that's
that's
our
intent
with
regard
to
integrity
management,
as
well
as
the
risk
management
protocols
is
to
to
challenge
ernie
and
everything
that
we've
done
to
see
what
we
could
do
differently.
F
What
we
could
do
better
what
we
could
discover
as
a
result,
a
result
of
the
root
cause
findings
of
this
of
this
particular
tragedy
and
incorporate
those
findings
into
the
into
the
ongoing
program.
So
we
can
absolutely
positively
prevent
similar
occurrence
into
the
future,
and
I
think
that's
why
the
the
reference
to
dsaw
is
so
so
critical
to
us
to
the
extent.
F
O
K
O
None
of
the
three
assessment
tools
allowed
to
us
through
the
program
would
tell
grade
of
the
pipe.
There
are
recently
efforts
through
gti
and
others,
a
gas
technical
institute
and
others
to
develop
in
the
field
or
in
the
ditch
technologies
for
looking
at
chemistries,
and
it's
not
quite
approved
by
asme.
Yet
but
there's
a
there's,
an
ability
to
evaluate
the
yield
strength
of
the
pipe
without
having
to
do
destructive
testing.
O
As
it
relates
to
the
integrity
management
program,
as
ms
peralta
mentioned
earlier
this
morning,
as
each
project
is
performed,
records
are
researched
and
evaluated,
and
those
records
in
the
case
of
direct
assessment
that
don't
have
have
what
we
call
assumed
values,
they're,
conservatively,
assumed
values
and
accepted
through
the
industry
based
on
support
those
values
that
are
assumed.
We
do
a
record
search
to
value,
to
evaluate
them.
That's
one
method.
O
The
other
method
is
we
have
a
rather
extensive
inspection
form
that
when
we
do
excavate
the
pipelines,
we
compare
what
we're
seeing
to
what
we
believe
we're
seeing
and
that's
an
area
that
we
continue
to
to
work
on.
Prior
to
september
9th
we
had
sent,
I
believe,
all
of
our
direct
assessment
data
from
all
phases.
K
O
O
That
failed
shows
a
weld
penetration
of
approximately
two-thirds
of
the
weld
seam
of
the
thickness.
Rather,
it
did
not
line
up
on
the
directly
on
the
scene,
so
only
about
50
of
that
penetration
was
actually
binding
them
together.
But
there
were
segments,
as
I
believe
you
mentioned,
or
showed
in
your
presentation
where
it
did
get
both
the
outside
and
the
inside.
O
O
O
O
O
H
H
O
So
the
2007
audit-
that
was
that
I
referred
to
where
outside
third
party
expert
came
in
there
were
people
who
are
very
familiar
with
what's
used
on
the
outside.
We
also
are
members
of
various
committees.
I
mentioned
a
few
of
those
that
I'm
part
of
and
the
information
that
we
learned
there
is
brought
back
to
the
company
there's
one
area,
for
example,
where
we
believe
that
further
further
direction
needed
to
be
provided
on
when
to
use
da
and
and
ili
and
hydro
test
in
a
complementary
fashion.
O
So
we
have,
we
have
pipelines
that
we've
dealed
and
also
follow
up
with
ili,
and
vice
versa.
We
went
to
the
national
association
of
corrosion
engineers.
They
agreed
to
open
a
committee.
The
committee
is
technical
group
401,
which
is
how
to
how
to
best
use
those
tools
and
that's
been
underway.
I
think
that
group
is
about
a
year
old
now.
O
I
don't
have
the
specific
locations.
I
can
provide
that.
I
believe
what
I
was
referring
to
is.
I
didn't
explain
the
whole
issue
when
we
saw
what
happened
with
this
line.
We
took
an
evaluation
where
we
looked
at
all
30
inch
pipe
that
was
installed
prior
to
1962
and
if
it
didn't
have
a
strength
test
on
it,
we
reduced
the
operating
pressure
by
20,
which
is
an
industry
agreed
way
of
doing
supported
by
fimsa.
O
B
O
O
H
A
H
F
We
are
looking
at
the
entire
process
again
from
beginning
to
end,
not
as
a
function
of
having
any
anything.
That
concludes
that
it
was.
It
was
broken
or
wrong,
but
just
the
fact
that
this
this
event
occurred
causes
us
to
step
back
and
question
everything
that
that
we're
doing,
I
think
when
we
get
the
root
cause
analysis,
we
understand
the
full
probable
root
cause.
We
can
then
step
back
and
then
and
then
assess
whether
there
was
something
more
or
different
that
we
we
could
have
or
should
have
done
right
now,
that's
not
apparent.
A
A
C
C
I
wanted
to
start
these.
I've
got
a
series
and,
if
I
run
out
of
time
I'll
be
glad
to
if
we
get
an
opportunity
to
come
back
to
finish
it
up,
but
first
of
all
I
wanted
to
complement
the
ntsb,
the
sort
of
the
emergency
notice
on
risk
assessment.
We
think,
was
very
much
on
the
spot
and
we
responded
quickly.
C
What
sort
of
level
what
sort
of
safety
factor
do
you
think
is
appropriate
to
apply
to
that
particular
outcome?
Does
that
make
sense?
To
you
I
mean
when
you
have
100
confidence
in
all
your
data
streams,
going
into
your
integrative
analysis
as
part
of
the
risk
assessment.
You
know,
I
think
you,
you
feel
a
lot
more
comfortable
going
forward.
So
just
invite
that
for
your
comment
so.
F
We
have
a
lot
of
stand-alone
legacy
systems
and
tools
that
we
have
for
use
within
gas
transmission,
and
so
we
have
a
plan
to
consolidate
all
of
those
and
to
migrate
towards
directional
systems.
So
we
we
intend
to
cure
the
system
problem.
You
know
aggressively
by
by
moving
towards
consolidating
pulling
in,
simplifying
and
and
part
of
that
process
will
be
validation
so
that
we
we're
not
just
pulling
records
together,
making
sure
that
records
agree
but
actually
doing
field
tests
so
that
we
can.
C
C
Well
I'll
skip
the
question
on
risk
assessment,
because
I
you
know
I'm
aware
of
the
fact
that
you've
done
kind
of
a
major
you
know
look
at
your
risk
assessment
methodology
and
I
think
our
state
partners
and
ourselves
are
very
interested
in
that
and
hope
to
work
with
you
a
little
more
detail
on
that
one.
I
would
ask
for
any
comments
you
have.
You
know.
You've
you've
learned
a
lot
of
lessons
going
through
this
and
you've
looked
at
some
of
your
assumptions
and
regarding
the
stability
of
certain
defects,
particularly
manufacturing
and
construction.
C
O
So
I
think
the
issue
well
the
issue
we're
addressing,
and
it's
in,
in
cooperation
with
our
cpuc
for
the
pipeline
that
was
established
using
the
historical
pressure.
There
has
been
no
specific
direction
that
if,
if
the
maop
of
the
line
was
established
using
the
historical
perspective,
what
that
means
relative
to
a
stable
or
unstable
flaw.
I
think
that
would
be
an
area
that
we
would
like.
We
are
looking
into
further
and
I
think
it
would
be
a
good
area
for
the
industry
looking
to.
C
J
O
The
and
ms
peralta
spoke
to
it
earlier
so
I'll
just
emphasize
I've
I've
been
at
this,
not
as
long
as
some,
but
in
my
20
years
the
the
d
saw
long.
Seam
has
always
been
considered
as
if
you're
going
to
join
a
pipe
on
a
seam,
that's
the
most
effective
way
of
doing
it
so
to
have
one
missing,
as
I
said,
not
only
just
approximately
four
feet
of
weld
on
the
inside
but
closer
to
12
feet,
because
there
was
three
cans.
O
That's
just
a
huge
surprise.
We've
been
doing
a
lot
of
research
on
this
since
the
at
some
point.
After
this
vintage,
the
standard
in
the
industry
changed
to
where
the
weld
on
the
inside
was
applied
first
and
the
weld
on
the
outside
was
applied.
Second,
during
this
vintage
with
this
manufacturer,
they
applied
the
weld
on
the
outside
first,
so
you
could
see
how,
through
a
quality
control
program
that
may
confuse
some
people,
but
that's
what
surprised
me
about
it.
J
O
I
believe
it's
unusual
generally,
the
only
time
I
would
see
and
I've
been
a
construction
supervisor.
I've
had
welders
and
pipe
gangs
under
my
responsibility.
Generally,
you
wouldn't
see
short
pieces
like
that.
Unless
it
was
around
a
fitting
there
was,
I
believe,
a
picture
that
was
in
one
of
the
exhibits
showing
the
the
project
at
the
tie-in
day
and
there's
three
or
four
full-length
pieces
of
pipe
on
the
job,
so
why
they
would
take
short
pieces
of
pipe
and
weld
them
up
to
fit.
O
A
O
I'm
sorry
there's
there's
a
picture
in
one
of
the
exhibits.
I
believe
it's
anyway
there's
a
picture
that
was
taken
at
the
night.
We
believe
it's
the
1956
project
on
pretty
close
to
tie-in
day,
so
they
when
they
were
going
to
tie
the
new
pipe
into
the
old
pipe,
and
in
the
background
of
that
picture,
you
see
manufactured
segments,
so
form
90,
degree
elbows
and
you
see
full
length
pieces
of
pipe.
O
We
know
from
the
investigation
that
these
short
cans,
some
of
them
were,
were
completely
square
and
some
of
them
had
a
miter
on
them,
so
the
miter
joint,
the
angle
was
probably
made
in
the
field,
but
the
cans
that
didn't
of
that
were
of
square
equal
length
it.
I
just
had
a
hard
time,
believing
that
a
welder
would
have
taken
six
pieces
of
pipe
and
spent
all
that
time,
welding
them
together
when
there
was
full
lengths
of
pipe
left
over
on
the
project.
O
In
my
mind,
if
they
were
making
up
an
angle
and
what
we
know
from
from
the
field
reports
is
that
the
each
end
of
what's
left
out?
There
is
approximately
the
same
elevation,
but
is
off
in
alignment
by
a
degree
in
half
one
and
a
half
degrees.
There
appears
to
have
been
about
a
three
degree
change
on
one
end,
when
you
look
at
the
manufacturing
report,
it
shows
the
cord
lengths
the
lengths
of
the
pipe
at
the
three
o'clock
nine
o'clock
and
12
o'clock
position
by
looking
at
those
lengths.
A
J
Let
me
direct
this
one
to
mr
salas.
Mr
traynor
walked
you
through
some
questions
about
the
integrity
management
plan
and
how
it
does,
or
should
deal
with
manufacturers,
defects
that
might
worsen
over
time
and
how
you
might
take
that
into
account.
The
kinds
of
defects
that
we've
that
we've
talked
about
so
far
mr
facets
talked
about
them,
are
the
particular
kinds
of
design
like
erw
and
and
and
so
on.
J
F
My
understanding
is
there
is:
there
is
a
a
method
for
for
dealing
with
lack
of
of
information
and
again
it's
a
conservative
value.
That's
placed
in.
I
think
what
what
then
needs
to
be
done
and
again
the
experts
are
sitting
next
to
me
with
regard
to
the
the
program
is
that
we
would
need
to
have
based
on
the
criticality
of
the
attribute
or
aspect
or
data
field
that
was
missing
some
set
of
steps
taken
to
determine
physically
in
the
field.
J
And
finally,
I
think,
I'm
probably
out
of
time
now,
but
I
wanted
to
just
raise
an
issue
to
see
whether
anyone
else
wants
to
pick
it
up,
and
that
is
whether
there
are
potential
threats
that
are
specifically
associated
with
aging
pipelines,
with
pipelines
of
this
vintage,
50s
and
60s.
I
think
that
about
half
of
your
your
existing
today
pipeline
system
was
installed
during
the
50s
and
60s,
and
mr
trainer
began
this
line
of
questioning.
But
I
I'm
curious
to
know
whether
you
think
that
there
are
particular
issues
just
associated
with
the
aging
of
the
infrastructure.
F
I
have
an
opinion-
it's
probably
not
going
to
be
terribly
popular
from
an
industry
standpoint.
I
think
our
our
historical
thinking
as
it
relates
to
asset
age
has
been
really
it's
really
not
much
about
age
as
much
about
age
as
it
is
about
the
condition
of
the
asset
and
and
it's
really
understanding
the
life
of
that
asset
and
having
that
be
the
basis
for
how
you
treat
it.
F
I
think,
having
given
our
experience,
we
need
to
step
back
and
challenge
that
that
precept
and
I
think
that
that
there
may
be
a
place
to
consider
age
or
other
attributes,
and
I
think
that's
part
of
why
we
are
suggesting
again
an
a
fairly
aggressive
modernization
program.
So,
while
we
talk
a
lot
about
integrity,
management
and
inline
inspection
and
hydro
testing
pressure,
testing
ecda
all
these
different
techniques
to
determine
the
character
of
of
pipe,
I
think
what
we
haven't
talked
about
is
at
what
point
do
we
simply
replace
pipe?
B
L
B
O
The
specifically
what
I
referenced
was
the
information
that
we've
provided
about
the
three
purchase
orders
of
pipe
one
of
those
purchase
orders
had
with
it
or
we
sent
in
with
it
a
report
from
moody
engineering,
which
was
then
a
third
party
quality
assurance
specializing
in
mill
inspection.
They
are
today
also
the
in
that
report.
It
called
out
the
design
of
that
pipe
to
be
3.
O
O
So
we
have
reason
to
believe
that
the
process
used
then
would
have
been
the
process
that
was
used
on
the
earlier
pipe.
In
that
discussion
of
the
long
seam
it
not
only
mentioned
that
it
was
back.
Then
they
called
them
electric
welded,
their
electric
fusion
welded.
Today
we
call
that
submerged
dark
welded,
but
it
specifically
described
that
the
design
requirements
for
the
long
seam
weld
was
to
have
an
outside
weld
whereby
the
fusion
would
penetrate
two-thirds
of
the
wall
thickness
of
the
pipe
and
an
inside
weld.
O
O
B
And
it's
double
submerged
because
it's
it's
weld
it's
welded
from
the
from
the
outside
and
from
the
inside.
That's
what
the
double
part
of
that
means
correct.
Yes,
sir!
So
so,
if
that's
what
the
design
specs
call
for,
why
was
there
confusion
in
the
pgene
records
concerning
this
being
unseen,
pipe
or
seamless,
pipe.
O
As
mr
dobbin
described
earlier
this
morning,
there
was
an
accounting
voucher
was
used,
the
slide
that
was
put
up
there
and
I'm
sorry,
I
don't
remember
which
one
it
was
otherwise.
I'd
asked
for
it
to
come
up
showed
that
there
was
an
accounting
code,
the
account
or
the
material
code,
which
called
out
essentially
what
I
just
said.
X52
dsaw,
but
the
description
that
was
typed
manually
at
the
time
was
one
of
seamless.
H
B
B
If
somebody
had
to
inspect
the
the
weld
from
the
outside,
you
could
see
that
the
well
was
there
or
not
there.
But
if
you
went
inside
the
pipe
it'd,
be
it's
easier
to
miss
an
inside,
well
being
miss
being
not
there
than
it
is
from
the
outside,
but
I
think
I'm
looking
at
the
in
an
interview
from
a
retired
employee,
in
fact
a
pge
employee,
in
fact
he's
the
one
who
took
the
photograph
that
that
you
referenced
earlier
in
1956.
B
O
So
his
purpose
was
to
crawl
through
and
and
look
for
those
things
typically
on
the
bottom
of
the
pipe.
I
believe
there's
also
discussion
in
that
transcript.
Where
they
say
did
you
look
for
the
long
seam
and
he
said
I
was
in
a
pipe.
My
head
was
down.
All
I
could
look
at
was
that
the
feet
that's
significant,
because
when
we
weld
up
pipes
we
it's
been
for
years.
The
standard
to
put
the
long
seam
weld
at
the
10
and
two
position
is
typical
for
a
pipeline.
A
M
M
O
O
We
took
the
30-inch
pipe
pre-1962
and
I
know
I've
said
that
before,
but
I
want
to
explain
a
little
bit.
Pre-62
is
for
a
couple
of
reasons.
We
looked
at
all
the
pipe
purchased,
all
the
pipe
installed
from
really
1948
and
continued
to
run
it
out
and
there's
a
spreadsheet.
That
shows
the
detail.
The
short
of
it
is
by
about
19
1959.
O
We
were
bringing
in
new
pipes,
so
any
pipe
that
was
left
over
from
these
purchase.
Orders
would
have
been
installed,
and
the
other
significance
about
1962
is
that
by
the
end
of
1961,
our
state
of
california
had
enacted
general
order,
112,
which
is
the
maintenance
and
operation
safety
standard
for
the
state.
At
the
time,
one
of
the
things
they
required
was
a
strength
test
for
any
pipe
intended
to
operate
above
20
percent
of
its
theoretical
strength.
M
O
Both
of
those
pipelines
have
had
the
pressure
reduced
by
20
and
we're
we've
had
getting
permits
and
we're
doing
efforts
to
get
inside
of
them
by
reference
to
technology
that
we're
working
with
a
couple
of
folks
on
to
improve
our
ability
to
go
in
hot,
which,
which
means
that
we
can
go
through
a
lot
more
pipe,
a
lot
easier
than
a
tethered
camera,
which
would
be
dragging
an
extension
cord
and
a
fiber
optics
camera,
and
things
like
that.
So,
in
short,
that's
what
we're
doing.
O
There's
restrictions
on
that,
because
if
you
increase
the
pressure
on
a
on
a
pipeline
using
a
compressible
non-compressible,
fluid
water,
then
when
the
pipe
fails,
the
energy
is
dissipated
out
through
the
water
and
any
crack
or
any
flaw
that's
created
by
that
is
arrested
quicker.
If
you
use
a
pneumatic,
you
have
to
get
the
pressure
down
so
that
flaw
may
run
farther.
O
F
I
think
the
other.
The
other
concern
that
we
had
is
the
potential
with
regard
to
hydro,
testing
or
activating
a
latent
defect,
and
so
we
could
in
fact
create
an
at-risk
situation
that
wouldn't
be
visible
to
us.
And
thus
we
think
that
the
test
methodology
that
that's
selected
really
needs
to
be
tuned
to
the
pipe
in
question,
as
opposed
to
a
blanket
policy.
With
regard
to
how
to
prove.
G
For
the
whole
panel,
I'm
curious
how
you
measure
the
success
or
effectiveness
of
your
integrity
management
program
a
lot
of
times
with
processes,
it's
the
number
of
inspections
that
are
done,
etc.
How
do
you
know
it's
working?
What
concrete
metrics
are
used
to
determine
whether
the
program
is
effective
or
not?.
L
P
L
G
F
I
think
member
rose
kind.
You
raised
a
really
good
point
and
it's
a
point
that
I
think
we
need
to
pursue
there.
There
isn't
what
I
would
consider
a
large
kind
of
a
single
macro
index
or
a
way
to
frame
all
the
individual
mechanisms
that
we
would
use
to
test
effectiveness
in
in
some
integrated
holistic
fashion,
and
I
think
that
that's
worthy
of
us
pursuing
further.
F
O
O
O
We
have
records
from
the
1948
project
that
purchased
this
pipeline,
where
they
were
sampling
at
a
10
rate
using
x-ray
of
the
girth
welds
there
there
were
tests
air
tests
that
were
done
on
them,
but
typically
they
were
done
at
around
100
pounds
or
a
little
higher,
because
they're
looking
more
for
leaks
than
they
were
to
strength
test,
the
fabrication
of
the
pipe.
At
that
time,
the
quality
assurance
on
the
pipe
itself
was
a
hydra
test.
At
the
mill
between
48
and
53.
O
O
G
O
There's
portions
of
it,
I
I
didn't,
I
don't
recall,
seeing
a
list
of
strength
tests.
There
is
a
list
of
what
pipes
failed
and
how
many
joints
they
didn't
pass.
There's
a
list
of
the
chemistries
taken
from
each
of
the
the
batches
there's
yield
strengths
and
other
things
like
that
that
are
reported
in
that
in
that
report,.
G
O
G
O
O
O
So
there's
a
couple
of
places
you
can,
you
can
see
short
pieces
of
pipe.
The
most
likely
places
are
in
the
field
construction
near
near
a
valve
because
we
want
to
at
a
valve
it's
called
a
typically
a
back
welded
joint.
When
we
go
in
the
thing
we're
worried
about
with
with
valves.
Is
we
don't
want
any
of
the
product
from
the
weld
of
the
circumferential
welding
process,
the
girth
welds?
O
We
don't
want
any
of
those
pieces
to
bounce
around
and
and
lay
themselves
up
against
the
valve,
because
it
would
prohibit
the
valve
from
closing
correctly.
So
we
call
for
short
sections
of
pipe.
That
would
be
at
least
two
diameters
long,
so
that
if
any
of
that
happened
it
wouldn't
roll
into
the
valve,
the
valve
is
protected
and
then
it's
back
welded
and
then
it's
outside
well,
that's
typically,
where
we
see
that
or
for
a
90
degree
elbow
or
45.
That
kind
of
thing.
O
O
O
O
O
So
if
you
were
to
put
that
in
without
the
caps
being
ground
down,
they
would
get
deformed
the
other
thing
that
we
know-
and
this
is
speculation-
we
just
understand
it
from
the
way
mills
work
is
my
understanding
is.
It
would
not
be
out
of
the
ordinary
for
them
to
have
taken
sheets
of
plate
that
had
been
trimmed
off
for
whatever
reason
and
roll
them
into
short
sections
to
be
used,
as
templates
to
set
up
equipment
which
may
have
been
just
a
process
of.
O
I
need
a
piece
of
pipe
to
set
it
up
to
set
the
diameters
or
to
get
the
weld
right
or
that
kind
of
thing.
Again:
it's
speculation.
Those
cans
may
have
been
lying
around
and
they
use
them.
We
know
from
the
reports
that
there
was
a
lot
of
demand
on
30
inch,
double
submerged,
dark,
welded
pipe
in
1948
and
49.
O
A
A
O
O
A
O
O
They
tend
to
be
what
we
would
call
a
nickel
high
and
a
dime
wide
and
looking
like
a
row
of
dimes,
there's
a
couple
of
these
that
don't
look
like
they're
they're,
very
flat,
that's
relevant
because
by
the
time
you
put
about
three-eighths
to
a
half
inch
of
hot
applied
asphalt
and
felt
over
that
someone
in
the
field
wouldn't
see
the
impression
of
that
girth.
Well,
so
it's
likely
they
could
have
taken
a
piece,
a
full
length
piece
from
their
perspective
and
cut
it
down.
A
O
We
haven't
discovered
the
specific
a-specific
procedure
that
that
told
them
what
to
do
and
how
often
to
do
it
there
may
there
are
some
indications
on
the
pipe
and
it's
still
being
investigated
that
may
indicate
it
at
one
joint.
There
was
a
a
girth
weld
x-ray
and
I
believe
it's
the
joint
between
pup,
4
and
pup
three,
but
other
than
that.
We've
not
we've
not
found
anything.
A
O
O
O
G
I
K
O
O
There
there
is
a
stamp
on
it
that
looks
that
meets
the
description
of
how
brands
were
made
by
consolidated
western
comes
from
the
1998.
I
believe
asme
report
on
the
history
of
pipe
manufacturing
and
we've
been
asking.
We
have
been
researching
to
see
whether
that's
the
brand
consolidated
western
use
consolidated
western,
was
absorbed
by
another
mill
in.
I
believe
1950.
K
O
So,
typically,
when
a
30
inch
or
30
foot
long
pipe
is
made,
you
have
a
plate,
that's
longer
than
30
feet,
and
then
it's
placed
in
a
series
of
presses
and
it's
bent
into
a.
U
and
then
it's
bent
more
into
a
circle,
and
then
it's
run
through
so
the
orientation
of
those
grains
through
that
process
would
be
lengthwise.
O
You
would
expect
them
to
go
the
30
foot
length,
the
orientation
of
the
grains
in
pup
1
we're
going
the
other
direction,
which
is
more
an
indication
that
somebody
may
have
taken
some
plate
and
again
they're
in
they're
trying
to
use
as
much
plate
wherever
they
could
a
plate.
That
may
have
had
a
defect
in
it
that
they
had
to
shear
off
and
they
took
that
plate
and
they
then
rolled
it.
O
It's
reasonable
to
say
that
consolidated
was
using
that
my
speculation
was,
we
know
they.
I
believe
we
know
that
they've
used
short
pieces
of
pipe
that
they've
made
to
set
up
equipment
because
they
wouldn't
want
to
get
the
settings
wrong
and
decide
and
destroy
a
full
piece
of
pipe.
So
they'd
use
short
pipes.
But
it's
as
I
said,
it's
speculation
and
I
believe
it's
still
being
researched
by
the
making.
K
I
realize
my
time
is
coming
up
I'll.
Ask
you
last
question
in
this
whole
discussion
of
pop,
and
I
only
made
few
doubts
you
must
have
used
the
word,
I
believe,
or
it's
hard
to
believe,
or
I
don't
understand,
probably
if
not
every
sentence,
every
other
sentence
and
my
concern
is
how
does
strong
believe
you
support
contention
that
maybe
we
have
a
diesel
welding
problem
here?
We
don't
really
have
any
confirmatory
facts
and
is
pg
e
doing
something
to
provide
those
facts
or
we
are
just
drawing
this
conclusion
based
on
beliefs.
I
O
That
should
tell
us
what
the
specific
issue
is.
We
didn't
want
to
not
do
anything
so
we're
taking
our
best
effort
using
the
information
we
have
and
trying
to
mitigate
the
concern
but
correct
we
don't
we
don't
have
a
root
cause
yet
and
when
we
do,
then
we'll
look
at
what
we
did
and
determine
if
that
was
enough
or
we
need
to
do
more.
C
D
C
G
C
Okay,
sorry
about
that,
I'm
really
just
trying
to
get
at
the
question
about
the
extent
to
which
your
program
looks
at
the
interaction
of
various
defects,
and
it
could
be
anything
from
subsidence
to
seismic
activity.
How
does
the
assumption
regarding
the
stability
of
the
welds,
for
example?
How
is
that
influenced
by
other
potential
risk
information?
This
is
sort
of
the
integrity
management
program
requires
an
integration
of
that
information.
So
I'll
just
turn
it
over
to
you.
Bob,
probably.
I
guess.
O
So,
through
the
risk
calculations
that
equation-
I
I
mentioned
earlier-
there's
factors
associated
with
things
like
outside
force
and
external
corrosion
and
and
third-party
damage
things
like
that.
All
of
that
has
different
weightings
associated
with
it
and
specifically
I'll
talk
about
the
interaction
between
external
corrosion
concern
about
older
girth
welds
that
were
used
before
the
requirement
for
100
x-ray
and
outside
force.
California,
we
have
earthquakes,
so
we
have.
We
know
from
the
u.s
geological
maps
where
the
pipelines
run,
and
we
know
what
the
expected
acceleration
from
a
seismic
event
would
would
do
so.
O
O
So
maybe
it's
got
20
years
without
corrosion
control
on
it.
It's
got
these
girth
welds.
We
would
look
at
that.
Those
pipelines
would
come
up
in
in
priority
which
may
move
it
and
likely
would
move
it
to
the
first
part
of
the
baseline
assessment
program.
132
was
in
the
first
part
of
that
that
program.
C
Okay,
great
sorry
that
mr
salas
commented
on
and
it
has
to
do
with
aging
infrastructure.
I
think
there's
been
a
lot
of
debate
out
there
and
I
know
it's
not
as
simple
as
just
age,
but
age
is
a
consideration,
as
you
said,
I
wonder
if
you
have
any
ideas
on
how
pipeline
operators
could
improve
the
ability
to
re-qualify
their
infrastructure
as
fit
for
service.
C
F
We've
had
a
large
body
of
work
connected
with
developing
a
decision
tree
that
that
takes
pipe
of
various
attributes
through
a
series
of
steps
to
either
qualify
or,
ultimately,
you
know,
recommend
rehabilitation
or,
in
fact,
replacement,
and
we
would
be
happy
to
share
that.
You
know,
as
as
the
this
room
is
interested.
B
Thank
you
I'll
make
this
quick,
mr
solace.
This
this
question
is,
is
really
one
that
I'm
interested
in,
but
it
doesn't
doesn't
relate
directly
to
what
we're
talking
about.
So
my
apologies
but
oftentimes
the
board
members
go
out.
We
talk
to
the
industry,
we
talk
about
things
like
the
cost
of
an
accident.
B
F
Remember
somewhat,
that's
that's
a
tough
question,
I
I
would
say
the
vast
majority
of
of
senior
management
and
and
that
I
would
include
our
board
of
directors
in
that.
So
we
have
our
board.
We
have
executive
management,
corporate
management
as
well
as
senior
management
within
the
utility,
not
only
those
that
are
in
the
technical
disciplines,
but
across
a
number
of
the
other
other
organizations.
So
I'd
say
this:
this
is
consuming
a
very
material
proportion
of
our
time.
B
Yeah
and
that's
what
I
was
thinking-
and
I
appreciate
that-
there's
a
tangible
answer
there
that
we
can
take
out
to
the
industry
because
that's
sort
of
a
hidden
cost
that
that
time
is
time
that
you're
not
able
to
spend
with
the
strategic
and
tactical
direction
of
growing
the
company.
So
I
appreciate
your
answer
there.
Thank
you
very
much,
madam
chairman.
A
F
O
O
O
A
A
We
heard
about
cathodic
protection,
that's
been
added,
and
so
I
think
that
rhetorically,
we
can
see
that
each
of
these
actions
over
the
years
have
been
an
additional
layer
of
safety,
whether
it's
on
the
manufactured
product,
the
inspection
of
the
product
maintenance
of
the
product.
All
of
those
are
are
layers
that
have
been
added
on.
A
A
O
A
O
A
Very
good,
thank
you
all
so
much
for
a
very
productive
day,
thanks
to
this
panel's
witnesses,
for
mr
salas
for
being
an
encore
witness
for
us
today.
Thank
you
for
the
answers
that
you've
provided
and
your
willingness
to
appear
before
us
today.
It
has
helped
advance
our
investigation
both
to
the
witnesses
on
panel
two
as
well
as
panel
one.
We
will
reconvene
at
nine
o'clock
tomorrow
for
panel
three
on
public.