1
00:00:00,000 --> 00:00:12,160
[Music]

2
00:00:12,160 --> 00:00:13,279
Chain of events.

3
00:00:13,279 --> 00:00:15,199
Cause and effect. We analyze what went

4
00:00:15,199 --> 00:00:16,399
right, and what went wrong,

5
00:00:16,399 --> 00:00:18,239
as we discover that many outcomes can be

6
00:00:18,239 --> 00:00:19,600
predicted, planned for

7
00:00:19,600 --> 00:00:22,399
and even prevented. I'm John Chidgey, and

8
00:00:22,399 --> 00:00:24,399
this is Causality.

9
00:00:24,399 --> 00:00:26,079
Causality is supported by you: our

10
00:00:26,079 --> 00:00:27,920
listeners. If you'd like to support the

11
00:00:27,920 --> 00:00:28,239
show

12
00:00:28,239 --> 00:00:31,039
you can by becoming a patron. Patron

13
00:00:31,039 --> 00:00:31,920
subscribers

14
00:00:31,920 --> 00:00:33,679
have access to early release, high

15
00:00:33,679 --> 00:00:35,600
quality ad-free episodes as well as

16
00:00:35,600 --> 00:00:38,479
bonus episodes and to Causality Explored.

17
00:00:38,479 --> 00:00:40,878
You can do this via Patreon, just visit

18
00:00:40,878 --> 00:00:42,238
https://engineered.network.causality

19
00:00:42,238 --> 00:00:44,719
to learn how you can help this

20
00:00:44,719 --> 00:00:46,719
show to continue to be made.

21
00:00:46,719 --> 00:00:51,520
Thank you. "San Bruno"

22
00:00:51,600 --> 00:00:55,359
At 6:11pm local daylight savings time

23
00:00:55,359 --> 00:00:58,479
on Thursday the 9th of September, 2010 in

24
00:00:58,479 --> 00:00:59,359
San Bruno,

25
00:00:59,359 --> 00:01:01,840
a small suburb south of San Crancisco in

26
00:01:01,840 --> 00:01:02,960
California,

27
00:01:02,960 --> 00:01:05,438
an explosion occurred in Crestmore a

28
00:01:05,438 --> 00:01:07,599
residential neighborhood.

29
00:01:07,599 --> 00:01:10,319
The explosion registered at 1.1 on the

30
00:01:10,319 --> 00:01:11,438
Richter scale

31
00:01:11,438 --> 00:01:14,239
and it took place only 3.2km or

32
00:01:14,239 --> 00:01:15,280
2mi

33
00:01:15,280 --> 00:01:17,118
from the San Francisco International

34
00:01:17,118 --> 00:01:19,599
Airport.

35
00:01:19,599 --> 00:01:22,079
The Pacific Gas & Electric Company (PG&E)

36
00:01:22,079 --> 00:01:22,799
...

37
00:01:22,799 --> 00:01:24,799
owned and operated the gas pipeline, as

38
00:01:24,799 --> 00:01:26,239
well as multiple others,

39
00:01:26,239 --> 00:01:28,319
as well as electrical distribution

40
00:01:28,319 --> 00:01:30,078
throughout approximately t2/3rds of

41
00:01:30,078 --> 00:01:31,438
northern California.

42
00:01:31,438 --> 00:01:33,519
From Bakersfield to near each of the

43
00:01:33,519 --> 00:01:35,040
Oregon, Nevada,

44
00:01:35,040 --> 00:01:38,159
and Arizona state lines. PG&E

45
00:01:38,159 --> 00:01:41,359
operated approximately 2,900km

46
00:01:41,359 --> 00:01:42,159
or

47
00:01:42,159 --> 00:01:45,118
1,800mi of high pressure gas pipelines

48
00:01:45,118 --> 00:01:47,759
at that time.

49
00:01:47,759 --> 00:01:50,478
Before we talk about the incident, let's

50
00:01:50,478 --> 00:01:51,680
talk a little bit about how gas

51
00:01:51,680 --> 00:01:53,438
pipelines work.

52
00:01:53,438 --> 00:01:55,920
Gas is moved around pipelines from a

53
00:01:55,920 --> 00:01:56,478
source

54
00:01:56,478 --> 00:01:58,879
to a destination. Where the gas is coming

55
00:01:58,879 --> 00:02:00,478
from (a source location)

56
00:02:00,478 --> 00:02:03,200
is referred to as "Upstream," to a

57
00:02:03,200 --> 00:02:04,000
"Downstream"

58
00:02:04,000 --> 00:02:05,680
location where it's going to its

59
00:02:05,680 --> 00:02:08,159
destination. Gas is compressed

60
00:02:08,159 --> 00:02:10,318
at a compression station and then that

61
00:02:10,318 --> 00:02:12,080
pressurized gas is output

62
00:02:12,080 --> 00:02:14,719
into a pipeline. A pipeline is best

63
00:02:14,719 --> 00:02:15,598
thought of

64
00:02:15,598 --> 00:02:18,080
as similar to a balloon, but fit that

65
00:02:18,080 --> 00:02:18,719
balloon

66
00:02:18,719 --> 00:02:21,759
with a flow control valve on one side to

67
00:02:21,759 --> 00:02:23,759
counterbalance the person blowing air

68
00:02:23,759 --> 00:02:26,080
into the inlet of the balloon.

69
00:02:26,080 --> 00:02:28,080
Like a balloon if the outlet side is

70
00:02:28,080 --> 00:02:29,598
restricted too much,

71
00:02:29,598 --> 00:02:31,598
whilst pressurised gas continues to

72
00:02:31,598 --> 00:02:32,719
enter it,

73
00:02:32,719 --> 00:02:35,120
then it will continue to expand, until

74
00:02:35,120 --> 00:02:35,840
eventually

75
00:02:35,840 --> 00:02:38,719
it ruptures. Hence controlling the outlet

76
00:02:38,719 --> 00:02:40,719
flow is crucial to ensuring safe

77
00:02:40,719 --> 00:02:41,598
operation

78
00:02:41,598 --> 00:02:45,039
of a pipeline. Put too much gas in:

79
00:02:45,039 --> 00:02:46,318
it will over-pressure if you're not

80
00:02:46,318 --> 00:02:48,719
letting that amount of gas out.

81
00:02:48,719 --> 00:02:50,959
Where pipelines join or split they use a

82
00:02:50,959 --> 00:02:52,878
connecting pipe section referred to as a

83
00:02:52,878 --> 00:02:54,560
"Header."

84
00:02:54,560 --> 00:02:57,039
To ensure each downstream section gets

85
00:02:57,039 --> 00:02:58,400
the correct flow,

86
00:02:58,400 --> 00:02:59,919
and that everything is set correctly

87
00:02:59,919 --> 00:03:02,639
sometimes, the back-pressure is monitored

88
00:03:02,639 --> 00:03:05,598
and alarmed to detect reverse flow, which

89
00:03:05,598 --> 00:03:08,639
is usually undesirable.

90
00:03:08,639 --> 00:03:11,519
A pipeline operator's role is to ensure

91
00:03:11,519 --> 00:03:13,039
that the flow rates

92
00:03:13,039 --> 00:03:14,800
and pressure regulating or pressure

93
00:03:14,800 --> 00:03:16,158
control valves,

94
00:03:16,158 --> 00:03:18,878
are set such that gas is directed in the

95
00:03:18,878 --> 00:03:20,318
right quantities

96
00:03:20,318 --> 00:03:22,639
to the right destinations within the

97
00:03:22,639 --> 00:03:23,919
operational limits

98
00:03:23,919 --> 00:03:26,639
of the pipeline network. High pressure

99
00:03:26,639 --> 00:03:28,560
pipelines are normally made out of steel

100
00:03:28,560 --> 00:03:30,318
or concrete lined steel

101
00:03:30,318 --> 00:03:32,878
and whilst they are made of a steel they

102
00:03:32,878 --> 00:03:33,680
still have a

103
00:03:33,680 --> 00:03:36,000
rupturing pressure limit referred to as

104
00:03:36,000 --> 00:03:37,199
the MaOP

105
00:03:37,199 --> 00:03:40,318
(or Maximum Operating Pressure).

106
00:03:40,318 --> 00:03:42,959
Steel fundamentally has 2 problems:

107
00:03:42,959 --> 00:03:43,919
Corrosion

108
00:03:43,919 --> 00:03:46,959
and fracturing. Steel is made from Iron

109
00:03:46,959 --> 00:03:49,120
mixed with Carbon which turns Iron into

110
00:03:49,120 --> 00:03:50,959
a significantly stronger material by

111
00:03:50,959 --> 00:03:53,039
introducing barriers to dislocations

112
00:03:53,039 --> 00:03:54,239
attempting to traverse through the

113
00:03:54,239 --> 00:03:56,158
materials crystalline structure.

114
00:03:56,158 --> 00:03:58,158
When dislocations propagate and build

115
00:03:58,158 --> 00:03:59,919
over a long period of time they can lead

116
00:03:59,919 --> 00:04:00,878
to fracturing

117
00:04:00,878 --> 00:04:02,959
and this effect is often referred to as

118
00:04:02,959 --> 00:04:04,400
fatigue stress.

119
00:04:04,400 --> 00:04:06,639
For a steel pipeline as gas pressure

120
00:04:06,639 --> 00:04:08,719
increases and then decreases again,

121
00:04:08,719 --> 00:04:11,360
day-in, day-out, for a long period of time

122
00:04:11,360 --> 00:04:13,360
those dislocations will inevitably lead

123
00:04:13,360 --> 00:04:15,598
to weaknesses in the steel pipeline,

124
00:04:15,598 --> 00:04:18,879
increasing the risk of a rupture. Hence a

125
00:04:18,879 --> 00:04:20,560
pipeline has a limited lifetime

126
00:04:20,560 --> 00:04:23,120
after which it cannot be used safely and

127
00:04:23,120 --> 00:04:24,639
exceeding the MaOP

128
00:04:24,639 --> 00:04:26,959
even for a young steel pipeline might

129
00:04:26,959 --> 00:04:29,439
not lead to a rupture at that moment,

130
00:04:29,439 --> 00:04:31,680
but each time there is a pressure cycle

131
00:04:31,680 --> 00:04:32,800
over that limit,

132
00:04:32,800 --> 00:04:35,120
it will significantly shorten the usable

133
00:04:35,120 --> 00:04:37,918
life of that pipeline.

134
00:04:37,918 --> 00:04:39,360
Pipeline pressures are generally

135
00:04:39,360 --> 00:04:41,120
referred to as Gauge

136
00:04:41,120 --> 00:04:43,600
Pressure as the suffix of the unit of

137
00:04:43,600 --> 00:04:44,639
measure.

138
00:04:44,639 --> 00:04:46,639
In this incident it was in the United

139
00:04:46,639 --> 00:04:48,000
States: all

140
00:04:48,000 --> 00:04:50,240
pressure measurements are in psi (or

141
00:04:50,240 --> 00:04:51,839
pounds per square inch)

142
00:04:51,839 --> 00:04:55,360
and referenced to gauge, hence psi gauge

143
00:04:55,360 --> 00:04:56,959
(or psig).

144
00:04:56,959 --> 00:05:00,879
PG&E had identified that their existing

145
00:05:00,879 --> 00:05:03,680
UPS (Uninterruptable Power Supply) system

146
00:05:03,680 --> 00:05:05,839
which was supplying backup for the 120V

147
00:05:05,839 --> 00:05:06,160
...

148
00:05:06,160 --> 00:05:08,639
AC distribution panels, throughout their

149
00:05:08,639 --> 00:05:09,199
network

150
00:05:09,199 --> 00:05:12,160
needed to be systematically upgraded for

151
00:05:12,160 --> 00:05:14,000
a multitude of reasons,

152
00:05:14,000 --> 00:05:15,839
and in order to upgrade it that it would

153
00:05:15,839 --> 00:05:17,360
first be necessary

154
00:05:17,360 --> 00:05:20,000
to move individual loads off the UPS

155
00:05:20,000 --> 00:05:21,519
distribution board,

156
00:05:21,519 --> 00:05:24,399
to individual standalone UPSs'. The

157
00:05:24,399 --> 00:05:26,160
project consisted of multiple

158
00:05:26,160 --> 00:05:28,319
smaller outages where individual loads

159
00:05:28,319 --> 00:05:30,639
were transferred over one by one.

160
00:05:30,639 --> 00:05:33,279
Approval for the UPS project work was

161
00:05:33,279 --> 00:05:35,279
granted on the 27th of August,

162
00:05:35,279 --> 00:05:39,120
2010. By the 9th of September,

163
00:05:39,120 --> 00:05:40,800
some loads had been transferred in the

164
00:05:40,800 --> 00:05:42,319
first stage of the project,

165
00:05:42,319 --> 00:05:44,160
and several more were planned for that

166
00:05:44,160 --> 00:05:45,519
day.

167
00:05:45,519 --> 00:05:47,759
The Milpitas Terminal plays a pivotal

168
00:05:47,759 --> 00:05:49,519
role in this incident.

169
00:05:49,519 --> 00:05:51,519
The impacted section of the Milpitas

170
00:05:51,519 --> 00:05:52,959
Terminal took 2 gas

171
00:05:52,959 --> 00:05:56,240
lines in: 300A and 131,

172
00:05:56,240 --> 00:05:57,918
into a common set of headers numbered

173
00:05:57,918 --> 00:05:59,439
3, 4, 5, 6

174
00:05:59,439 --> 00:06:02,800
and 7. From these headers line 300B

175
00:06:02,800 --> 00:06:06,160
and 327 feed downstream stations with

176
00:06:06,160 --> 00:06:07,680
several connectors dropping the pressure

177
00:06:07,680 --> 00:06:09,439
to header 2.

178
00:06:09,439 --> 00:06:12,319
From which 4 downstream lines: Line

179
00:06:12,319 --> 00:06:13,360
100,

180
00:06:13,360 --> 00:06:18,240
Line 109, Line 101 and Line 132...

181
00:06:18,240 --> 00:06:20,879
which would be the line of the incident.

182
00:06:20,879 --> 00:06:21,680
And then

183
00:06:21,680 --> 00:06:24,399
the San Jose distribution feeder is also

184
00:06:24,399 --> 00:06:26,079
taken off from there.

185
00:06:26,079 --> 00:06:28,879
Each feeder has a Pressure Regulating

186
00:06:28,879 --> 00:06:30,160
Valve (or PRV),

187
00:06:30,160 --> 00:06:32,319
with a local independent controller

188
00:06:32,319 --> 00:06:34,000
which in turn

189
00:06:34,000 --> 00:06:35,918
can be controlled externally in

190
00:06:35,918 --> 00:06:38,000
automatic from a PLC

191
00:06:38,000 --> 00:06:40,639
(Programmable Logic Controller) Each

192
00:06:40,639 --> 00:06:42,879
feeder also has a monitor valve

193
00:06:42,879 --> 00:06:44,879
which is a fully independent,

194
00:06:44,879 --> 00:06:46,720
pneumatically operated,

195
00:06:46,720 --> 00:06:48,800
pressure controller intended to be a

196
00:06:48,800 --> 00:06:50,720
final over-pressure protection

197
00:06:50,720 --> 00:06:53,519
for downstream pipes. Monitor valves are a

198
00:06:53,519 --> 00:06:55,519
more traditional method of control.

199
00:06:55,519 --> 00:06:58,000
They're commonly slow acting therefore a

200
00:06:58,000 --> 00:06:59,519
high lag,

201
00:06:59,519 --> 00:07:01,120
and these devices sometimes can take

202
00:07:01,120 --> 00:07:03,439
several minutes to react and compensate

203
00:07:03,439 --> 00:07:05,680
as gas pressures increase or decrease

204
00:07:05,680 --> 00:07:07,360
with faster adjustments handled by the

205
00:07:07,360 --> 00:07:09,598
PLC controlled PRVs.

206
00:07:09,598 --> 00:07:11,759
Finally there are many manually operated

207
00:07:11,759 --> 00:07:13,598
valves that also allow operators to

208
00:07:13,598 --> 00:07:15,360
direct gas flow in multiple directions

209
00:07:15,360 --> 00:07:17,439
as well as fully isolate pipelines

210
00:07:17,439 --> 00:07:19,519
and headers and provide a pathway for a

211
00:07:19,519 --> 00:07:20,959
full station bypass

212
00:07:20,959 --> 00:07:25,360
if desired. With that background,

213
00:07:25,360 --> 00:07:28,000
let's talk about the incident itself. At

214
00:07:28,000 --> 00:07:29,199
2:46pm

215
00:07:29,199 --> 00:07:31,360
a gas control technician at the Milpitas

216
00:07:31,360 --> 00:07:33,279
Terminal obtained verbal work clearance

217
00:07:33,279 --> 00:07:35,439
from the SCADA center in San Francisco

218
00:07:35,439 --> 00:07:37,839
to initiate another UPS changeover as

219
00:07:37,839 --> 00:07:40,000
part of the broader project.

220
00:07:40,000 --> 00:07:42,720
At 3:36pm the Milpitas technician

221
00:07:42,720 --> 00:07:44,560
called the SCADA center and spoke with a

222
00:07:44,560 --> 00:07:46,000
second SCADA operator

223
00:07:46,000 --> 00:07:48,160
to confirm whether the valves on 2 of

224
00:07:48,160 --> 00:07:50,639
the incoming gas pipelines would close

225
00:07:50,639 --> 00:07:53,759
when power was withdrawn. Upon being told

226
00:07:53,759 --> 00:07:54,478
they would,

227
00:07:54,478 --> 00:07:56,399
the Milpitas technician then locked the

228
00:07:56,399 --> 00:07:58,000
valves open to prevent their

229
00:07:58,000 --> 00:08:01,120
accidental closure during the changeover.

230
00:08:01,120 --> 00:08:02,879
Shortly afterward the workers at the

231
00:08:02,879 --> 00:08:04,720
Milpitas terminal contacted the SCADA

232
00:08:04,720 --> 00:08:06,079
supervisor on shift

233
00:08:06,079 --> 00:08:08,160
to confirm the correct action to take

234
00:08:08,160 --> 00:08:09,519
for the PRVs,

235
00:08:09,519 --> 00:08:11,279
the Pressure Regulating Valves. The

236
00:08:11,279 --> 00:08:13,680
supervisor advised them to put the PRVs

237
00:08:13,680 --> 00:08:14,879
into manual control

238
00:08:14,879 --> 00:08:16,800
at a set position to prevent them from

239
00:08:16,800 --> 00:08:19,519
operating during the loss of power.

240
00:08:19,519 --> 00:08:22,000
At 4:03pm the Milpitas technician

241
00:08:22,000 --> 00:08:24,160
alerted the third SCADA operator

242
00:08:24,160 --> 00:08:25,759
that the installation of one of the

243
00:08:25,759 --> 00:08:28,240
smaller UPSs was about to begin.

244
00:08:28,240 --> 00:08:30,240
The PRVs were put into manual control at

245
00:08:30,240 --> 00:08:31,759
the controller itself,

246
00:08:31,759 --> 00:08:33,599
and indicated they would be returned to

247
00:08:33,599 --> 00:08:36,080
automatic once completed.

248
00:08:36,080 --> 00:08:39,120
At 4:18pm the SCADA center lost

249
00:08:39,120 --> 00:08:41,519
SCADA indication for pressures, flows

250
00:08:41,519 --> 00:08:42,719
and valve positions

251
00:08:42,719 --> 00:08:46,240
at the Milpitas terminal. At 4:32pm

252
00:08:46,240 --> 00:08:48,320
the second SCADA operator contacted the

253
00:08:48,320 --> 00:08:50,480
Milpitas terminal requesting an update

254
00:08:50,480 --> 00:08:52,399
as the works were taking much longer

255
00:08:52,399 --> 00:08:54,240
than expected.

256
00:08:54,240 --> 00:08:56,480
At 4:38pm the Milpitas technician

257
00:08:56,480 --> 00:08:58,799
contacted the SCADA centre to verify the

258
00:08:58,799 --> 00:09:00,559
SCADA indications had now returned to

259
00:09:00,559 --> 00:09:01,278
normal

260
00:09:01,278 --> 00:09:03,519
and upon verifying this, the PRV

261
00:09:03,519 --> 00:09:07,039
controllers were returned to automatic.

262
00:09:07,039 --> 00:09:09,440
Shortly after this the technicians began

263
00:09:09,440 --> 00:09:10,399
powering off

264
00:09:10,399 --> 00:09:13,200
several breakers: one of which was

265
00:09:13,200 --> 00:09:14,559
unidentified,

266
00:09:14,559 --> 00:09:16,559
and they noted that the local control

267
00:09:16,559 --> 00:09:17,600
panel was now

268
00:09:17,600 --> 00:09:20,799
no longer powered up. The technicians

269
00:09:20,799 --> 00:09:22,799
decided to leave the circuit breaker

270
00:09:22,799 --> 00:09:25,759
open and instead referred to electrical

271
00:09:25,759 --> 00:09:27,839
drawings to try and determine

272
00:09:27,839 --> 00:09:29,600
a different way they could power the

273
00:09:29,600 --> 00:09:31,120
local control panel.

274
00:09:31,120 --> 00:09:33,360
As they investigated further they noted

275
00:09:33,360 --> 00:09:35,278
that the power supply that powered some

276
00:09:35,278 --> 00:09:37,200
of the local control panel equipment was

277
00:09:37,200 --> 00:09:39,679
outputting an inconsistent voltage.

278
00:09:39,679 --> 00:09:42,240
The power supplies were 120V AC to

279
00:09:42,240 --> 00:09:44,480
24V DC supplies operating as a

280
00:09:44,480 --> 00:09:45,839
redundant pair on the DC

281
00:09:45,839 --> 00:09:48,480
bus, and that DC bus also powered the 24V

282
00:09:48,480 --> 00:09:49,120
DC

283
00:09:49,120 --> 00:09:51,519
instruments: most notably the pressure

284
00:09:51,519 --> 00:09:52,399
transmitters

285
00:09:52,399 --> 00:09:54,399
that fed the pressure process signal

286
00:09:54,399 --> 00:09:57,039
into the PRV controller.

287
00:09:57,039 --> 00:09:59,759
Since PRVs operate on a pressure

288
00:09:59,759 --> 00:10:00,399
feedback

289
00:10:00,399 --> 00:10:02,799
in a closed loop, if the input signal

290
00:10:02,799 --> 00:10:05,200
reports that the pressure is very low,

291
00:10:05,200 --> 00:10:07,759
the valve will proportionally open to

292
00:10:07,759 --> 00:10:09,278
increase the pressure.

293
00:10:09,278 --> 00:10:11,919
When the pressure doesn't increase the

294
00:10:11,919 --> 00:10:13,679
valve will open more,

295
00:10:13,679 --> 00:10:16,240
and more until the valve is inevitably

296
00:10:16,240 --> 00:10:17,600
fully opened.

297
00:10:17,600 --> 00:10:20,639
At 5:22pm the SCADA center alarm

298
00:10:20,639 --> 00:10:23,039
console displayed more than 60 alarms

299
00:10:23,039 --> 00:10:25,440
in a space of a few seconds, including

300
00:10:25,440 --> 00:10:27,039
controller error alarms,

301
00:10:27,039 --> 00:10:29,360
high differential pressure, gas backflow

302
00:10:29,360 --> 00:10:31,600
alarms and several High High pressure

303
00:10:31,600 --> 00:10:34,078
alarms from the Milpitas terminal.

304
00:10:34,078 --> 00:10:36,559
At 5 25 pm the third SCADA operator

305
00:10:36,559 --> 00:10:38,000
called the Milpitas technicians

306
00:10:38,000 --> 00:10:39,919
reporting the high pressure alarms and

307
00:10:39,919 --> 00:10:40,639
at this time

308
00:10:40,639 --> 00:10:42,639
the Milpitas technician realised that

309
00:10:42,639 --> 00:10:44,078
the pressure and regulating valve

310
00:10:44,078 --> 00:10:45,759
controller displays on the local control

311
00:10:45,759 --> 00:10:48,240
panel had lost all of their data.

312
00:10:48,240 --> 00:10:50,480
The SCADA consoles at Milpitas displayed

313
00:10:50,480 --> 00:10:52,320
constant pressures, a sign of

314
00:10:52,320 --> 00:10:56,000
"Report Last Known Value," as a result of

315
00:10:56,000 --> 00:10:58,639
the process value having been lost with

316
00:10:58,639 --> 00:10:59,839
all the downstream

317
00:10:59,839 --> 00:11:02,399
line PRVs and most of the monitor and

318
00:11:02,399 --> 00:11:04,000
incoming line valves indicating that

319
00:11:04,000 --> 00:11:04,559
they were not

320
00:11:04,559 --> 00:11:08,159
open. At 5:28pm the Milpitas technician

321
00:11:08,159 --> 00:11:10,639
called the fourth SCADA operator to ask

322
00:11:10,639 --> 00:11:12,559
what pressure values were being

323
00:11:12,559 --> 00:11:14,879
displayed on the remote SCADA console

324
00:11:14,879 --> 00:11:16,799
and discovered that the remote console

325
00:11:16,799 --> 00:11:17,759
was not receiving

326
00:11:17,759 --> 00:11:20,879
any live values at all. The remote SCADA

327
00:11:20,879 --> 00:11:23,919
console indicated 458 psig

328
00:11:23,919 --> 00:11:26,159
at the Milpitas terminal mixer. (That's

329
00:11:26,159 --> 00:11:27,679
the name for the common headers 3

330
00:11:27,679 --> 00:11:28,000
through 7)

331
00:11:28,000 --> 00:11:30,399
The fourth SCADA operator

332
00:11:30,399 --> 00:11:32,399
concluded that the PRVs and/or

333
00:11:32,399 --> 00:11:34,399
the station bypass valves may have

334
00:11:34,399 --> 00:11:36,320
been opened which was quickly confirmed

335
00:11:36,320 --> 00:11:38,078
by the technician on site.

336
00:11:38,078 --> 00:11:39,839
With all of the regulating valves fully

337
00:11:39,839 --> 00:11:41,839
open, only the monitor valves were

338
00:11:41,839 --> 00:11:42,879
limiting the pressure

339
00:11:42,879 --> 00:11:45,600
on the downstream pipelines. The monitor

340
00:11:45,600 --> 00:11:46,399
valves were set

341
00:11:46,399 --> 00:11:50,000
at 386 psig, however due to control

342
00:11:50,000 --> 00:11:51,759
lag in the monitor valve's pneumatic

343
00:11:51,759 --> 00:11:53,600
controller, the pressure in the lines

344
00:11:53,600 --> 00:11:55,559
leaving the Milpitas terminal peaked at

345
00:11:55,559 --> 00:11:57,679
396 psig

346
00:11:57,679 --> 00:12:00,799
sometime between 5:22pm and 5:25pm,

347
00:12:00,799 --> 00:12:03,679
however with a lack of transmitter

348
00:12:03,679 --> 00:12:06,000
data it's impossible to be certain

349
00:12:06,000 --> 00:12:08,320
what the peak value was and exactly when

350
00:12:08,320 --> 00:12:09,519
it occurred.

351
00:12:09,519 --> 00:12:12,480
At 5:42pm the Milpitas technician

352
00:12:12,480 --> 00:12:14,480
called the SCADA center and reported to

353
00:12:14,480 --> 00:12:16,240
the third SCADA operator that the

354
00:12:16,240 --> 00:12:18,159
regulating valves on the primary

355
00:12:18,159 --> 00:12:19,759
feedline (300A) had

356
00:12:19,759 --> 00:12:22,240
opened fully but the remote operator was

357
00:12:22,240 --> 00:12:23,360
unable to confirm

358
00:12:23,360 --> 00:12:25,360
live pressures or valve positions from

359
00:12:25,360 --> 00:12:28,320
the Milpitas terminal on his SCADA console.

360
00:12:28,320 --> 00:12:30,639
The third operator then authorized the

361
00:12:30,639 --> 00:12:32,559
local technician to reduce the manual

362
00:12:32,559 --> 00:12:34,159
set point of the monitor valves

363
00:12:34,159 --> 00:12:38,078
from 386 psig to 370 psig to try to

364
00:12:38,078 --> 00:12:40,879
reduce the downstream line pressures.

365
00:12:40,879 --> 00:12:44,159
At 5:52pm the senior SCADA coordinator

366
00:12:44,159 --> 00:12:46,320
directed the fourth SCADA operator to

367
00:12:46,320 --> 00:12:48,000
lower the upstream pressure set points

368
00:12:48,000 --> 00:12:49,839
for the gas supplying the Milpitas

369
00:12:49,839 --> 00:12:50,399
terminal

370
00:12:50,399 --> 00:12:52,720
upstream, as this was still able to be

371
00:12:52,720 --> 00:12:54,559
remotely controlled.

372
00:12:54,559 --> 00:12:56,559
Shortly after the site technician called

373
00:12:56,559 --> 00:12:58,399
the fourth SCADA operator to report

374
00:12:58,399 --> 00:13:00,320
that the monitor valves were shut on

375
00:13:00,320 --> 00:13:01,759
line 300B,

376
00:13:01,759 --> 00:13:03,600
with the fourth SCADA operator noting

377
00:13:03,600 --> 00:13:06,399
500 psig on the common headers 3

378
00:13:06,399 --> 00:13:07,200
through 7

379
00:13:07,200 --> 00:13:09,200
and requested the on-site technician to

380
00:13:09,200 --> 00:13:12,000
place a pressure gauge on Line 132

381
00:13:12,000 --> 00:13:14,240
to get a direct reading of the pressure

382
00:13:14,240 --> 00:13:15,519
leaving the terminal.

383
00:13:15,519 --> 00:13:17,278
There was no other way at this point to

384
00:13:17,278 --> 00:13:19,120
know for sure what that pressure

385
00:13:19,120 --> 00:13:22,958
actually was. Now at 6:02pm

386
00:13:22,958 --> 00:13:25,039
the fourth SCADA operator contacted a

387
00:13:25,039 --> 00:13:27,440
SCADA operator at the PG&E Brentwood

388
00:13:27,440 --> 00:13:28,799
facility stating,

389
00:13:28,799 --> 00:13:30,958
(and I quote): "We've over-pressured the

390
00:13:30,958 --> 00:13:32,159
whole peninsula." (end quote)

391
00:13:32,159 --> 00:13:34,879
At 6:04pm the senior SCADA coordinator

392
00:13:34,879 --> 00:13:36,559
informed the supervisor at the Milpitas

393
00:13:36,559 --> 00:13:37,919
terminal that the pressure on the

394
00:13:37,919 --> 00:13:39,679
incoming lines at the milpitas terminal

395
00:13:39,679 --> 00:13:42,320
had been lowered to 370 psig as

396
00:13:42,320 --> 00:13:43,600
directed.

397
00:13:43,600 --> 00:13:44,958
Shortly after this the on-site

398
00:13:44,958 --> 00:13:46,720
technician reported to the third SCADA

399
00:13:46,720 --> 00:13:47,440
operator

400
00:13:47,440 --> 00:13:49,278
that the manually-read line pressure on

401
00:13:49,278 --> 00:13:50,799
Line 132

402
00:13:50,799 --> 00:13:54,799
had reached 396 psig.

403
00:13:54,799 --> 00:13:58,879
At some time between 6:00pm and 6:11pm

404
00:13:58,879 --> 00:14:01,360
gas began leaking from a section of pipe

405
00:14:01,360 --> 00:14:02,559
in San Bruno,

406
00:14:02,559 --> 00:14:05,360
until it found an ignition source, which

407
00:14:05,360 --> 00:14:06,320
it found

408
00:14:06,320 --> 00:14:09,839
at 6:11pm. At 6:11pm

409
00:14:09,839 --> 00:14:12,639
a 911 call reported an explosion from

410
00:14:12,639 --> 00:14:14,958
either a suspected gas station

411
00:14:14,958 --> 00:14:17,759
or an aircraft crash. With emergency

412
00:14:17,759 --> 00:14:19,198
responders dispatched to the scene

413
00:14:19,198 --> 00:14:20,879
within 1 min

414
00:14:20,879 --> 00:14:22,958
due to the San Bruno Fire Department (the

415
00:14:22,958 --> 00:14:24,078
SBFD)

416
00:14:24,078 --> 00:14:26,720
having a fire station 300 yards from

417
00:14:26,720 --> 00:14:29,278
where the explosion occurred.

418
00:14:29,278 --> 00:14:32,879
At 6:11pm Line 132 upstream of Martin

419
00:14:32,879 --> 00:14:33,600
Sstation

420
00:14:33,600 --> 00:14:35,679
registered a pressure drop from the peak

421
00:14:35,679 --> 00:14:38,639
of 386 psig.

422
00:14:38,639 --> 00:14:40,799
Within 2 minutes of the explosion, many

423
00:14:40,799 --> 00:14:42,639
residents had self-evacuated and fled

424
00:14:42,639 --> 00:14:45,360
their homes to escape the fireball.

425
00:14:45,360 --> 00:14:48,240
By 6:15pm Martin Station generated a

426
00:14:48,240 --> 00:14:50,639
low pressure alarm for Line 132

427
00:14:50,639 --> 00:14:53,120
followed 20 seconds later by an alarm at

428
00:14:53,120 --> 00:14:55,440
150 psig.

429
00:14:55,440 --> 00:14:58,320
At 6:16pm local police began diverting

430
00:14:58,320 --> 00:14:59,759
traffic away from the incident

431
00:14:59,759 --> 00:15:02,639
and assisting in organizing evacuations.

432
00:15:02,639 --> 00:15:04,159
At 6:18pm

433
00:15:04,159 --> 00:15:07,120
an off-duty PG&E employee notified the

434
00:15:07,120 --> 00:15:09,039
PG&E dispatch centre in Concord,

435
00:15:09,039 --> 00:15:09,839
California,

436
00:15:09,839 --> 00:15:11,679
about an explosion that had occurred in

437
00:15:11,679 --> 00:15:13,360
the San Bruno area.

438
00:15:13,360 --> 00:15:16,078
This was the first formal report to PG&E

439
00:15:16,078 --> 00:15:17,198
about the incident,

440
00:15:17,198 --> 00:15:20,559
7 min after it had occurred. At

441
00:15:20,559 --> 00:15:23,039
6:23pm a dispatcher sent a gas service

442
00:15:23,039 --> 00:15:24,799
representative working in Daly City,

443
00:15:24,799 --> 00:15:26,240
that's 8mi South

444
00:15:26,240 --> 00:15:28,480
of San Bruno, to confirm the report was

445
00:15:28,480 --> 00:15:30,000
accurate.

446
00:15:30,000 --> 00:15:32,799
At 6:24pm a PG&E supervisor saw the

447
00:15:32,799 --> 00:15:34,480
accident fire while driving home from

448
00:15:34,480 --> 00:15:35,198
work,

449
00:15:35,198 --> 00:15:37,519
after which he called the PG&E dispatch

450
00:15:37,519 --> 00:15:39,519
center, reported the fire and proceeded

451
00:15:39,519 --> 00:15:40,879
to the scene.

452
00:15:40,879 --> 00:15:43,839
At 6:27pm a PG&E dispatcher called the

453
00:15:43,839 --> 00:15:45,839
SCADA centre and asked the third SCADA

454
00:15:45,839 --> 00:15:47,278
operator if they had observed any

455
00:15:47,278 --> 00:15:48,399
pressure drops

456
00:15:48,399 --> 00:15:51,759
(and I quote): "...at a station in (the) area..."

457
00:15:51,759 --> 00:15:53,440
stating there had been reports of a

458
00:15:53,440 --> 00:15:55,120
flame shooting up in the air

459
00:15:55,120 --> 00:15:57,039
accompanied by a sound similar to a jet

460
00:15:57,039 --> 00:15:58,720
engine and that a PG&E

461
00:15:58,720 --> 00:16:00,799
supervisor and a GSR had been

462
00:16:00,799 --> 00:16:02,078
dispatched to the area

463
00:16:02,078 --> 00:16:03,839
to which the third operator replied that

464
00:16:03,839 --> 00:16:05,759
the SCADA center had not received any

465
00:16:05,759 --> 00:16:06,720
calls about

466
00:16:06,720 --> 00:16:10,958
any incident. At 6:29pm the senior SCADA

467
00:16:10,958 --> 00:16:11,679
coordinator

468
00:16:11,679 --> 00:16:13,679
informed a SCADA coordinator at the

469
00:16:13,679 --> 00:16:15,360
Brentwood Facility that there had been a

470
00:16:15,360 --> 00:16:16,559
gas line break,

471
00:16:16,559 --> 00:16:18,320
and further stated that there had been

472
00:16:18,320 --> 00:16:19,679
an over-pressure event

473
00:16:19,679 --> 00:16:22,799
at the Milpitas terminal earlier. By 6:30pm

474
00:16:22,799 --> 00:16:23,440
...

475
00:16:23,440 --> 00:16:26,000
some staff at the SCADA center realised

476
00:16:26,000 --> 00:16:27,278
that there had been a rupture

477
00:16:27,278 --> 00:16:30,240
along Line 132 in the San Bruno area,

478
00:16:30,240 --> 00:16:32,399
however the exact rupture location was

479
00:16:32,399 --> 00:16:35,039
still unknown.

480
00:16:35,039 --> 00:16:39,600
We're now 19min post-explosion.

481
00:16:39,600 --> 00:16:42,879
At 6:35pm an off-duty PG&E gas

482
00:16:42,879 --> 00:16:45,039
measurement and control mechanic who was

483
00:16:45,039 --> 00:16:46,159
qualified to operate

484
00:16:46,159 --> 00:16:49,198
manual mainline valves, saw media reports

485
00:16:49,198 --> 00:16:50,320
about the fire,

486
00:16:50,320 --> 00:16:52,320
and as he suspected he might be able to

487
00:16:52,320 --> 00:16:54,639
assist, he notified the PG&E dispatch

488
00:16:54,639 --> 00:16:56,559
center and proceeded to the PG&E

489
00:16:56,559 --> 00:16:58,958
Colma Maintenance Yard to obtain a

490
00:16:58,958 --> 00:16:59,600
service truck

491
00:16:59,600 --> 00:17:01,519
and tools necessary to shut off a

492
00:17:01,519 --> 00:17:04,160
mainline valve in case it was needed.

493
00:17:04,160 --> 00:17:05,838
A few minutes later a second mechanic

494
00:17:05,838 --> 00:17:07,519
called him to check on his well-being

495
00:17:07,519 --> 00:17:08,480
and agreed to meet

496
00:17:08,480 --> 00:17:11,679
at the Colmer Yard and by 6:36pm the

497
00:17:11,679 --> 00:17:13,038
Line 132 pressure

498
00:17:13,038 --> 00:17:15,119
at the Martin Station had dropped to

499
00:17:15,119 --> 00:17:18,240
only 50 psig.

500
00:17:18,240 --> 00:17:21,359
At 6:50pm both mechanics arrived at

501
00:17:21,359 --> 00:17:22,240
Colma Yard

502
00:17:22,240 --> 00:17:24,400
and after reviewing a map showing the

503
00:17:24,400 --> 00:17:26,318
location of pipeline valves in the area

504
00:17:26,318 --> 00:17:28,400
and using information from local news

505
00:17:28,400 --> 00:17:29,279
reports

506
00:17:29,279 --> 00:17:32,000
regarding the fireball, they determined

507
00:17:32,000 --> 00:17:32,720
the rupture

508
00:17:32,720 --> 00:17:36,880
was most likely occurring in Line 132.

509
00:17:36,880 --> 00:17:38,558
They then called a supervisor to

510
00:17:38,558 --> 00:17:40,558
authorize isolating the line manually

511
00:17:40,558 --> 00:17:41,599
which was approved

512
00:17:41,599 --> 00:17:44,000
after a brief telephone conversation.

513
00:17:44,000 --> 00:17:46,480
About 7:06pm the two mechanics left the

514
00:17:46,480 --> 00:17:48,160
Colma Yard and route to the first

515
00:17:48,160 --> 00:17:51,679
mainline valve at (Mile Point) MP38.49

516
00:17:51,679 --> 00:17:53,599
that they planned to close and arrived

517
00:17:53,599 --> 00:17:55,038
at 7:20pm.

518
00:17:55,038 --> 00:17:57,919
At 7:27pm the supervisor who was with

519
00:17:57,919 --> 00:17:59,200
the two mechanics at

520
00:17:59,200 --> 00:18:02,000
MP38.49 requested the SCADA center

521
00:18:02,000 --> 00:18:04,558
close 2 valves at the Martin Station

522
00:18:04,558 --> 00:18:08,319
to reduce the flow of gas. At 7:29pm the

523
00:18:08,319 --> 00:18:09,839
fourth SCADA operator had remotely

524
00:18:09,839 --> 00:18:11,759
closed the 2 valves downstream of the

525
00:18:11,759 --> 00:18:12,400
rupture,

526
00:18:12,400 --> 00:18:14,319
stopping gas flow from North to South

527
00:18:14,319 --> 00:18:16,000
into Line 132.

528
00:18:16,000 --> 00:18:18,558
South to North however was still flowing.

529
00:18:18,558 --> 00:18:19,919
By 7:30pm

530
00:18:19,919 --> 00:18:21,759
the 2 mechanics had manually closed

531
00:18:21,759 --> 00:18:24,720
the mainline valve at MP38.49

532
00:18:24,720 --> 00:18:26,640
upstream of the rupture, stopping the gas

533
00:18:26,640 --> 00:18:28,079
flow into 132

534
00:18:28,079 --> 00:18:31,679
from that location. By 7:42pm

535
00:18:31,679 --> 00:18:34,720
now 91min after the rupture,

536
00:18:34,720 --> 00:18:36,880
the gas flow had decreased enough such

537
00:18:36,880 --> 00:18:38,400
that the intensity of the fire had

538
00:18:38,400 --> 00:18:39,759
decreased to a point

539
00:18:39,759 --> 00:18:41,279
where firefighters could approach the

540
00:18:41,279 --> 00:18:43,839
site and begin containment efforts.

541
00:18:43,839 --> 00:18:46,720
By 7:46pm the 2 mechanics had

542
00:18:46,720 --> 00:18:49,200
manually closed 2 additional valves

543
00:18:49,200 --> 00:18:50,640
downstream of the rupture

544
00:18:50,640 --> 00:18:54,960
at MP40.05 and MP40.05-2 at

545
00:18:54,960 --> 00:18:56,640
Healey Station.

546
00:18:56,640 --> 00:18:59,200
At this point, and by closing these final

547
00:18:59,200 --> 00:19:00,160
2 valves,

548
00:19:00,160 --> 00:19:02,640
the ruptured section of Line 132 was

549
00:19:02,640 --> 00:19:03,599
finally,

550
00:19:03,599 --> 00:19:07,038
fully isolated. 1hr

551
00:19:07,038 --> 00:19:09,440
and 35min after the loss of

552
00:19:09,440 --> 00:19:11,119
containment.

553
00:19:11,119 --> 00:19:14,400
Firefighters declared 75% of all active

554
00:19:14,400 --> 00:19:16,000
fires to be contained at about 4:24am

555
00:19:16,000 --> 00:19:19,599
the following morning. At 8:00pm

556
00:19:19,599 --> 00:19:22,720
on the 11th of September, 2010 all fires

557
00:19:22,720 --> 00:19:24,079
had finally

558
00:19:24,079 --> 00:19:27,279
been extinguished. As a result of the

559
00:19:27,279 --> 00:19:28,400
pipeline rupture

560
00:19:28,400 --> 00:19:31,839
and ensuing fires, 48 people sustained

561
00:19:31,839 --> 00:19:33,440
minor injuries,

562
00:19:33,440 --> 00:19:36,640
10 people sustained serious injuries, and

563
00:19:36,640 --> 00:19:39,440
8 people died.

564
00:19:39,440 --> 00:19:42,240
James Franco aged 58: a resident of

565
00:19:42,240 --> 00:19:43,759
Glenview Drive.

566
00:19:43,759 --> 00:19:47,279
Jacqueline Greig aged 44. Janessa Greig

567
00:19:47,279 --> 00:19:51,440
aged 13. Elizabeth Torres aged 81.

568
00:19:51,440 --> 00:19:54,480
Jessica Morales, aged 20.

569
00:19:54,480 --> 00:19:58,558
Lavonne Bullis, aged 87. Greg Bullis,

570
00:19:58,558 --> 00:20:02,480
aged 80. Will Bullis, aged 17.

571
00:20:02,480 --> 00:20:06,720
The rupture released an estimated 47.6

572
00:20:06,720 --> 00:20:09,599
million standard cubic feet of natural

573
00:20:09,599 --> 00:20:12,640
gas, that's about 136GJ (Giga-Joules)

574
00:20:12,640 --> 00:20:16,640
by energy. It left a 72ft by 26ft

575
00:20:16,640 --> 00:20:20,159
crater that's 22m by 8m, and ejected

576
00:20:20,159 --> 00:20:23,519
a 28ft long or 8.5m piece of

577
00:20:23,519 --> 00:20:24,480
pipe

578
00:20:24,480 --> 00:20:29,440
weighing 3,000lbs that's 1.36 tonnes.

579
00:20:29,440 --> 00:20:31,599
108 houses in the neighborhood were

580
00:20:31,599 --> 00:20:33,038
affected by the blast

581
00:20:33,038 --> 00:20:36,319
and fire, of which 53 had some damage,

582
00:20:36,319 --> 00:20:39,519
17 had severe or moderate damage and 38

583
00:20:39,519 --> 00:20:42,720
were completely destroyed.

584
00:20:42,880 --> 00:20:46,960
So what went wrong? There are seemingly

585
00:20:46,960 --> 00:20:50,000
3 components to explore in what went

586
00:20:50,000 --> 00:20:50,798
wrong.

587
00:20:50,798 --> 00:20:53,200
The UPS works impact on the pipeline

588
00:20:53,200 --> 00:20:54,798
pressure controllers,

589
00:20:54,798 --> 00:20:57,839
the design of the pipe itself, and the

590
00:20:57,839 --> 00:20:58,960
response

591
00:20:58,960 --> 00:21:02,000
to the incident. The response

592
00:21:02,000 --> 00:21:05,119
to the incident was slow. Mis-informed and

593
00:21:05,119 --> 00:21:06,319
ultimately

594
00:21:06,319 --> 00:21:08,798
the meaningful isolations were all done

595
00:21:08,798 --> 00:21:10,400
manually.

596
00:21:10,400 --> 00:21:13,200
Why were they done manually? There

597
00:21:13,200 --> 00:21:14,159
weren't enough

598
00:21:14,159 --> 00:21:16,558
remotely controlled valves to allow for

599
00:21:16,558 --> 00:21:18,319
a better isolation of the pipeleg

600
00:21:18,319 --> 00:21:21,679
segment, so actually the speed

601
00:21:21,679 --> 00:21:24,720
that PG&E could respond was severely

602
00:21:24,720 --> 00:21:25,519
hindered

603
00:21:25,519 --> 00:21:28,480
by the pipeline's design. So actually

604
00:21:28,480 --> 00:21:29,119
then

605
00:21:29,119 --> 00:21:31,200
the response problems are just as much

606
00:21:31,200 --> 00:21:33,200
about the design.

607
00:21:33,200 --> 00:21:36,240
So let's start there.

608
00:21:36,558 --> 00:21:38,880
There are several aspects to consider.

609
00:21:38,880 --> 00:21:40,720
The mechanical aspects of the pipe

610
00:21:40,720 --> 00:21:41,200
itself:

611
00:21:41,200 --> 00:21:44,480
why did it fail? The positioning of the

612
00:21:44,480 --> 00:21:46,960
isolation valves and their operability,

613
00:21:46,960 --> 00:21:49,519
and the path of the pipeline relative to

614
00:21:49,519 --> 00:21:51,919
where people had their houses built.

615
00:21:51,919 --> 00:21:53,679
So first things first let's talk about

616
00:21:53,679 --> 00:21:55,440
the pipe itself.

617
00:21:55,440 --> 00:21:58,480
Line 132 was constructed in phases,

618
00:21:58,480 --> 00:22:01,919
during 1944 and an extension in 1948

619
00:22:01,919 --> 00:22:04,000
which included the segment that failed

620
00:22:04,000 --> 00:22:07,919
in this incident. Line 132 consists of 24",

621
00:22:07,919 --> 00:22:11,359
30", 34" and 36" diameter segments of

622
00:22:11,359 --> 00:22:13,119
different steel grades with different

623
00:22:13,119 --> 00:22:14,640
longitudinal seam welds

624
00:22:14,640 --> 00:22:17,599
including: Double Submerged Arc-Welded

625
00:22:17,599 --> 00:22:18,720
(DSAW),

626
00:22:18,720 --> 00:22:21,038
Electric Resistance Welded (ERW) and

627
00:22:21,038 --> 00:22:21,919
Seamless

628
00:22:21,919 --> 00:22:26,000
(SMLS). To clarify, a longitudinal weld

629
00:22:26,000 --> 00:22:27,519
is one that traverses the length of the

630
00:22:27,519 --> 00:22:29,519
pipe to join the bent piece of steel

631
00:22:29,519 --> 00:22:30,798
into a pipe,

632
00:22:30,798 --> 00:22:32,640
and a girth weld is one that joins 2

633
00:22:32,640 --> 00:22:34,558
pieces of pipe together in a circle

634
00:22:34,558 --> 00:22:37,839
around the pipe. Pipe welds are commonly

635
00:22:37,839 --> 00:22:40,079
welded both on the inside and outside

636
00:22:40,079 --> 00:22:42,400
seams to improve their strength. It's a

637
00:22:42,400 --> 00:22:44,640
labor-intensive and tiring job

638
00:22:44,640 --> 00:22:46,159
and to ensure the welds are done

639
00:22:46,159 --> 00:22:48,319
properly they are inspected

640
00:22:48,319 --> 00:22:49,839
after they're completed by different

641
00:22:49,839 --> 00:22:51,679
means including x-rays

642
00:22:51,679 --> 00:22:53,839
and old-fashioned gas or water

643
00:22:53,839 --> 00:22:55,440
pressurization tests.

644
00:22:55,440 --> 00:22:58,159
Records from the 1948 project included

645
00:22:58,159 --> 00:23:00,640
logs for 209 radiographs

646
00:23:00,640 --> 00:23:02,960
and between the two project stages about

647
00:23:02,960 --> 00:23:03,759
10%

648
00:23:03,759 --> 00:23:06,798
of all welds were X-rayed. That's not

649
00:23:06,798 --> 00:23:08,880
very many but keep in mind that in those

650
00:23:08,880 --> 00:23:10,720
days it was an expensive procedure

651
00:23:10,720 --> 00:23:13,119
and X-raying...the entire pipe from

652
00:23:13,119 --> 00:23:14,079
end to end was

653
00:23:14,079 --> 00:23:17,200
unheard of, though 10% isn't very much

654
00:23:17,200 --> 00:23:18,400
admittedly.

655
00:23:18,400 --> 00:23:20,240
According to construction records as the

656
00:23:20,240 --> 00:23:22,240
final check before introducing gas to

657
00:23:22,240 --> 00:23:22,798
the 30"

658
00:23:22,798 --> 00:23:24,960
portion of the line, it was checked

659
00:23:24,960 --> 00:23:26,798
for leaks and breaks,

660
00:23:26,798 --> 00:23:28,640
though the exact method that was used

661
00:23:28,640 --> 00:23:30,480
wasn't described in any detail,

662
00:23:30,480 --> 00:23:32,319
it was assumed by investigators to

663
00:23:32,319 --> 00:23:34,720
consist of a soap and water

664
00:23:34,720 --> 00:23:36,880
mixture on girth welds as it was

665
00:23:36,880 --> 00:23:38,000
pressurised.

666
00:23:38,000 --> 00:23:41,839
In 1956 however PG&E were required

667
00:23:41,839 --> 00:23:47,119
to relocate 1,851ft, that's 564m

668
00:23:47,119 --> 00:23:49,519
of Line 132 that had originally been

669
00:23:49,519 --> 00:23:50,159
installed

670
00:23:50,159 --> 00:23:53,519
in 1948, and this section actually

671
00:23:53,519 --> 00:23:56,159
contained the segment that failed.

672
00:23:56,159 --> 00:23:58,319
They moved this because the existing

673
00:23:58,319 --> 00:24:01,679
elevation of Line 132 was incompatible

674
00:24:01,679 --> 00:24:04,079
with land grading that had been done at

675
00:24:04,079 --> 00:24:04,720
the time

676
00:24:04,720 --> 00:24:06,880
in connection with residential housing

677
00:24:06,880 --> 00:24:07,839
being built

678
00:24:07,839 --> 00:24:10,240
in a new residential area. This

679
00:24:10,240 --> 00:24:13,119
relocation re-routed Line 132 from the

680
00:24:13,119 --> 00:24:15,200
East-side to the West-side of Glenview

681
00:24:15,200 --> 00:24:15,839
Drive,

682
00:24:15,839 --> 00:24:19,839
and was done entirely in-house by PG&E.

683
00:24:19,839 --> 00:24:22,000
The investigation found that there were

684
00:24:22,000 --> 00:24:23,038
no design

685
00:24:23,038 --> 00:24:25,759
material construction specifications, no

686
00:24:25,759 --> 00:24:26,960
inspection records,

687
00:24:26,960 --> 00:24:30,079
no as-built drawings and no X-ray

688
00:24:30,079 --> 00:24:32,880
reports for any of the works conducted

689
00:24:32,880 --> 00:24:36,159
from the 1956 relocation.

690
00:24:36,159 --> 00:24:39,919
Line 132, pipe segment 180, was documented

691
00:24:39,919 --> 00:24:42,319
in the PG&E  GIS (that's the Geographic

692
00:24:42,319 --> 00:24:43,679
Information System)

693
00:24:43,679 --> 00:24:46,240
as a 30" diameter Seamless Steel

694
00:24:46,240 --> 00:24:47,440
pipe.

695
00:24:47,440 --> 00:24:51,038
Its code was API 5L X42,

696
00:24:51,038 --> 00:24:55,440
also known as L290 pipe under ISO3183.

697
00:24:55,440 --> 00:24:59,679
The wall thickness was supposedly 0.375"

698
00:24:59,679 --> 00:25:00,480
...

699
00:25:00,480 --> 00:25:03,359
and was installed in 1956 with the

700
00:25:03,359 --> 00:25:04,640
manufacturer noted

701
00:25:04,640 --> 00:25:08,480
as "N/A" on the data sheet.

702
00:25:08,480 --> 00:25:12,159
(Not very useful) During the investigation

703
00:25:12,159 --> 00:25:14,640
the NTSB found that the data in the GIS

704
00:25:14,640 --> 00:25:16,000
was actually imported

705
00:25:16,000 --> 00:25:19,839
in 1998 from data from a pipeline survey

706
00:25:19,839 --> 00:25:22,480
sheet that was created in 1977,

707
00:25:22,480 --> 00:25:24,640
and in digging further it was discovered

708
00:25:24,640 --> 00:25:26,319
that this section

709
00:25:26,319 --> 00:25:28,558
had no engineering records relating to

710
00:25:28,558 --> 00:25:30,400
its specification.

711
00:25:30,400 --> 00:25:32,000
The reference was taken from an

712
00:25:32,000 --> 00:25:35,119
accounting journal from 1956

713
00:25:35,119 --> 00:25:37,038
as materials were transferred between

714
00:25:37,038 --> 00:25:38,720
different construction jobs

715
00:25:38,720 --> 00:25:41,759
and not only that, the reference had been

716
00:25:41,759 --> 00:25:44,640
incorrectly entered in 1977

717
00:25:44,640 --> 00:25:47,599
as X42 when the original voucher

718
00:25:47,599 --> 00:25:48,480
actually read

719
00:25:48,480 --> 00:25:51,038
X52 which is a completely different

720
00:25:51,038 --> 00:25:52,640
pipe.

721
00:25:52,640 --> 00:25:54,159
The investigators found that it was not

722
00:25:54,159 --> 00:25:56,000
possible for this pipe to have been

723
00:25:56,000 --> 00:25:59,038
X42 as seamless pipe wasn't

724
00:25:59,038 --> 00:26:00,960
and still isn't available in diameters

725
00:26:00,960 --> 00:26:02,640
larger than 26"

726
00:26:02,640 --> 00:26:04,880
and this was a 30" pipe segment,

727
00:26:04,880 --> 00:26:06,400
apparently.

728
00:26:06,400 --> 00:26:08,798
The investigators found that this pipe

729
00:26:08,798 --> 00:26:09,519
segment was

730
00:26:09,519 --> 00:26:12,880
actually made from 6 PUPS, which is

731
00:26:12,880 --> 00:26:15,919
pipeline slang, for a short length of

732
00:26:15,919 --> 00:26:16,960
pipe - a short pipe

733
00:26:16,960 --> 00:26:20,079
segment. And each of these 6 PUPS

734
00:26:20,079 --> 00:26:22,000
had been welded together to make one

735
00:26:22,000 --> 00:26:24,319
longer section.

736
00:26:24,319 --> 00:26:27,599
None of the pups were X52 pipe, or

737
00:26:27,599 --> 00:26:30,400
X42 pipe either and they varied in

738
00:26:30,400 --> 00:26:32,720
length from 3.5ft to 4.7ft

739
00:26:32,720 --> 00:26:35,919
long each. Laboratory testing of the

740
00:26:35,919 --> 00:26:38,159
composition and the material properties

741
00:26:38,159 --> 00:26:39,440
in some of the PUPS

742
00:26:39,440 --> 00:26:42,558
didn't meet the PG&E 1948 material

743
00:26:42,558 --> 00:26:44,079
specifications or

744
00:26:44,079 --> 00:26:45,759
any industry pipeline material

745
00:26:45,759 --> 00:26:48,159
specifications for that time period.

746
00:26:48,159 --> 00:26:50,159
They also found several of the PUPS had

747
00:26:50,159 --> 00:26:52,000
partially welded longitudinal

748
00:26:52,000 --> 00:26:54,480
seams with part of the seam unwelded on

749
00:26:54,480 --> 00:26:55,839
the inside,

750
00:26:55,839 --> 00:26:57,679
and several of the girth worlds joining

751
00:26:57,679 --> 00:27:00,079
the PUPS together also contained several

752
00:27:00,079 --> 00:27:01,679
welding defects.

753
00:27:01,679 --> 00:27:05,278
Interestingly in 1995 PG&E replaced

754
00:27:05,278 --> 00:27:08,240
several sections of Line 132.

755
00:27:08,240 --> 00:27:10,679
Their replacements however ended about

756
00:27:10,679 --> 00:27:11,839
565ft

757
00:27:11,839 --> 00:27:14,960
to the South and about 610ft to

758
00:27:14,960 --> 00:27:15,599
the North

759
00:27:15,599 --> 00:27:18,000
of Segment 180 as part of a multi-year

760
00:27:18,000 --> 00:27:19,440
replacement project

761
00:27:19,440 --> 00:27:23,759
to address seismic hazards. Prior to 1961

762
00:27:23,759 --> 00:27:26,159
no regulations in California existed

763
00:27:26,159 --> 00:27:27,038
that governed

764
00:27:27,038 --> 00:27:29,599
natural gas pipeline safety. The

765
00:27:29,599 --> 00:27:32,159
voluntary national consensus standard

766
00:27:32,159 --> 00:27:36,480
ASME B31.1.8, 1955 edition,

767
00:27:36,480 --> 00:27:39,200
called for hydrostatic pressure testing

768
00:27:39,200 --> 00:27:40,480
of pipelines

769
00:27:40,480 --> 00:27:44,319
at 1.1x to 1.4x the MaOP.

770
00:27:44,319 --> 00:27:48,000
PG&E elected not to hydrostatically test

771
00:27:48,000 --> 00:27:51,038
segment 180 of Line 132,

772
00:27:51,038 --> 00:27:52,798
and whether PG&E followed the other

773
00:27:52,798 --> 00:27:54,640
guidelines of the ASME standard

774
00:27:54,640 --> 00:27:56,159
can't be demonstrated because it was

775
00:27:56,159 --> 00:27:58,000
never documented.

776
00:27:58,000 --> 00:28:01,440
In 1961 the new General Order 112

777
00:28:01,440 --> 00:28:03,519
required hydrostatic pressure testing of

778
00:28:03,519 --> 00:28:05,839
newly constructed pipelines in Class 3

779
00:28:05,839 --> 00:28:06,480
areas

780
00:28:06,480 --> 00:28:10,398
at 1.5x the MaOP.

781
00:28:10,398 --> 00:28:13,599
However as Line 132 was not technically

782
00:28:13,599 --> 00:28:15,200
a new pipeline,

783
00:28:15,200 --> 00:28:16,798
it was not required to be

784
00:28:16,798 --> 00:28:19,200
retrospectively tested.

785
00:28:19,200 --> 00:28:21,119
So let's talk a little bit more about

786
00:28:21,119 --> 00:28:23,519
MaOP and why it's important.

787
00:28:23,519 --> 00:28:26,079
When a pipe section's designed, a

788
00:28:26,079 --> 00:28:27,200
combination of its

789
00:28:27,200 --> 00:28:29,038
material properties, the construction and

790
00:28:29,038 --> 00:28:31,839
size, all dictate its maximum operating

791
00:28:31,839 --> 00:28:33,440
pressure that it can withstand.

792
00:28:33,440 --> 00:28:35,359
Depending on how much gas volume you

793
00:28:35,359 --> 00:28:37,519
want to transport over what distance,

794
00:28:37,519 --> 00:28:39,278
and taking into account how much Line

795
00:28:39,278 --> 00:28:40,558
Pack you want to keep

796
00:28:40,558 --> 00:28:42,798
to allow for supply upsets and demand

797
00:28:42,798 --> 00:28:45,759
surges, you pick the right kind of pipe

798
00:28:45,759 --> 00:28:47,839
at the design stage, before you put in

799
00:28:47,839 --> 00:28:48,798
the ground.

800
00:28:48,798 --> 00:28:50,720
Well that's the idea at least but in

801
00:28:50,720 --> 00:28:52,880
this case the pipeline wasn't

802
00:28:52,880 --> 00:28:54,640
really designed with that kind of

803
00:28:54,640 --> 00:28:56,798
methodology and if it was the section

804
00:28:56,798 --> 00:28:58,880
they fitted in 1956

805
00:28:58,880 --> 00:29:00,880
consisting of those PUPS, certainly

806
00:29:00,880 --> 00:29:02,079
wasn't.

807
00:29:02,079 --> 00:29:03,839
In terms of testing requirements though

808
00:29:03,839 --> 00:29:06,480
for pipelines constructed before 1970

809
00:29:06,480 --> 00:29:08,000
where hydrostatic testing wasn't

810
00:29:08,000 --> 00:29:10,839
required, 49 CFR

811
00:29:10,839 --> 00:29:14,159
192.619(a) (3),

812
00:29:14,159 --> 00:29:17,038
our old friend a "grandfather clause"

813
00:29:17,038 --> 00:29:19,200
allows the MaOP to be based on

814
00:29:19,200 --> 00:29:21,440
the highest actual operating pressure to

815
00:29:21,440 --> 00:29:23,200
which the segment was subjected,

816
00:29:23,200 --> 00:29:25,759
during the 5 years preceding July the

817
00:29:25,759 --> 00:29:26,278
1st,

818
00:29:26,278 --> 00:29:30,319
1970. Unlike an MaOP that's tested based

819
00:29:30,319 --> 00:29:31,919
on a hydrostatic pressure test,

820
00:29:31,919 --> 00:29:33,919
the grandfather clause doesn't specify a

821
00:29:33,919 --> 00:29:35,278
minimum amount of time that the

822
00:29:35,278 --> 00:29:37,038
historical pressure needs to have been

823
00:29:37,038 --> 00:29:37,519
held

824
00:29:37,519 --> 00:29:39,119
before it could be considered a valid

825
00:29:39,119 --> 00:29:40,880
MaOP sample.

826
00:29:40,880 --> 00:29:42,480
The thinking is that if a pipe wasn't

827
00:29:42,480 --> 00:29:44,398
designed with the target MaOP in mind

828
00:29:44,398 --> 00:29:45,440
from the start,

829
00:29:45,440 --> 00:29:47,440
if that pipeline has operated at a peak

830
00:29:47,440 --> 00:29:49,038
pressure recently in that 5 year

831
00:29:49,038 --> 00:29:49,759
period,

832
00:29:49,759 --> 00:29:52,640
it should be okay to handle it? Right?

833
00:29:52,640 --> 00:29:53,599
Well...

834
00:29:53,599 --> 00:29:57,038
not really...but that's the regulation.

835
00:29:57,038 --> 00:30:01,278
In 1987 the NTSB recommended eliminating

836
00:30:01,278 --> 00:30:03,278
the grandfather clause in Safety

837
00:30:03,278 --> 00:30:06,839
Recommendation:

838
00:30:06,839 --> 00:30:09,759
p-87-9 however this was not taken up by

839
00:30:09,759 --> 00:30:11,359
the standards boards

840
00:30:11,359 --> 00:30:14,079
following public comments recommending

841
00:30:14,079 --> 00:30:15,038
against

842
00:30:15,038 --> 00:30:18,319
its elimination. The MaOP for Line 132

843
00:30:18,319 --> 00:30:19,359
was established

844
00:30:19,359 --> 00:30:22,640
at 400 psig under the grandfather

845
00:30:22,640 --> 00:30:24,880
clause as a result of records from the

846
00:30:24,880 --> 00:30:27,200
highest operating pressure on Line 132

847
00:30:27,200 --> 00:30:29,200
during that 5 year period,

848
00:30:29,200 --> 00:30:31,679
it had reached 400 psig on October

849
00:30:31,679 --> 00:30:32,519
the 16th,

850
00:30:32,519 --> 00:30:35,679
1968. Now PG&E

851
00:30:35,679 --> 00:30:38,240
had a subtly different interpretation.

852
00:30:38,240 --> 00:30:40,079
They considered the MaOP of a pipeline to

853
00:30:40,079 --> 00:30:41,278
be the maximum pressure

854
00:30:41,278 --> 00:30:43,519
at which a pipeline system, as

855
00:30:43,519 --> 00:30:45,200
distinguished from a pipeline segment,

856
00:30:45,200 --> 00:30:46,640
can operate.

857
00:30:46,640 --> 00:30:50,880
Hence the MaOP set by PG&E for Line 132

858
00:30:50,880 --> 00:30:53,919
was only 375 psig.

859
00:30:53,919 --> 00:30:55,519
This is because when cross-ties

860
00:30:55,519 --> 00:30:57,599
connecting Line 132,

861
00:30:57,599 --> 00:31:00,159
although it has an MaOP of 400 psig,

862
00:31:00,159 --> 00:31:02,960
when it connects to Line 109 it only has

863
00:31:02,960 --> 00:31:03,919
an MaOP

864
00:31:03,919 --> 00:31:07,038
of 375 psig and once those cross-

865
00:31:07,038 --> 00:31:09,919
ties are open, the MaOP of Line 132

866
00:31:09,919 --> 00:31:13,759
is the MaOP of Line 109 which is 375 psig.

867
00:31:13,759 --> 00:31:14,960
...

868
00:31:14,960 --> 00:31:17,519
Now whether that MaOP was correctly set

869
00:31:17,519 --> 00:31:20,240
to match the pipeline's actual capacity,

870
00:31:20,240 --> 00:31:23,278
it was probably borderline, well it would

871
00:31:23,278 --> 00:31:24,960
have been for a well-constructed segment

872
00:31:24,960 --> 00:31:26,798
but not a poorly constructed segment.

873
00:31:26,798 --> 00:31:28,240
Whether the measured pressure

874
00:31:28,240 --> 00:31:30,079
immediately prior to the loss of

875
00:31:30,079 --> 00:31:32,079
containment exceeded PG&E's

876
00:31:32,079 --> 00:31:34,960
own MaOP for Line 132 there is no

877
00:31:34,960 --> 00:31:36,319
question that it was exceeded

878
00:31:36,319 --> 00:31:37,679
significantly.

879
00:31:37,679 --> 00:31:41,200
It was clocked at 398 psig, MaOP was

880
00:31:41,200 --> 00:31:41,839
set

881
00:31:41,839 --> 00:31:46,159
to 375psig. The NTSB report's findings, well,

882
00:31:46,159 --> 00:31:47,759
there were 28 in total,

883
00:31:47,759 --> 00:31:49,278
but I'll just touch on the ones that i

884
00:31:49,278 --> 00:31:52,079
think are the most relevant.

885
00:31:52,079 --> 00:31:53,839
The pipe segment that failed did not

886
00:31:53,839 --> 00:31:55,839
conform to PG&E's own standards

887
00:31:55,839 --> 00:31:57,440
nor any other standards of the time and

888
00:31:57,440 --> 00:31:59,359
was unable to be traced to where it was

889
00:31:59,359 --> 00:32:00,079
manufactured

890
00:32:00,079 --> 00:32:02,558
or by whom, and it would have failed

891
00:32:02,558 --> 00:32:04,480
accepted industry quality control and

892
00:32:04,480 --> 00:32:06,079
welding standards from the day it was

893
00:32:06,079 --> 00:32:08,640
constructed.

894
00:32:08,798 --> 00:32:10,640
The lack of quality control during the

895
00:32:10,640 --> 00:32:13,278
1956 relocation project led to a

896
00:32:13,278 --> 00:32:15,359
defective pipe segment being installed

897
00:32:15,359 --> 00:32:16,079
that remained

898
00:32:16,079 --> 00:32:19,119
undetected until it failed 54 years

899
00:32:19,119 --> 00:32:20,480
later.

900
00:32:20,480 --> 00:32:23,440
the fracture of Line 132 Segment 180,

901
00:32:23,440 --> 00:32:25,278
originated in the partially welded

902
00:32:25,278 --> 00:32:27,359
longitudinal seam of a PUP.

903
00:32:27,359 --> 00:32:29,679
PUP number 1. Which was progressively

904
00:32:29,679 --> 00:32:32,240
weakened due to ductile crack growth

905
00:32:32,240 --> 00:32:35,278
and fatigue stress. The SCADA system

906
00:32:35,278 --> 00:32:36,960
made tracing the source of the leak

907
00:32:36,960 --> 00:32:38,398
difficult and

908
00:32:38,398 --> 00:32:40,558
a lack of remotely controllable valves

909
00:32:40,558 --> 00:32:42,240
and a lack of Automatic Line Break

910
00:32:42,240 --> 00:32:44,319
valves led to an excessive delay

911
00:32:44,319 --> 00:32:46,398
in isolating the segment, causing

912
00:32:46,398 --> 00:32:48,720
significantly greater damage.

913
00:32:48,720 --> 00:32:51,278
The report was also critical of the poor

914
00:32:51,278 --> 00:32:53,119
planning of the electrical UPS work at

915
00:32:53,119 --> 00:32:54,398
the Milpitas terminal,

916
00:32:54,398 --> 00:32:56,398
which i agree could have been done a bit

917
00:32:56,398 --> 00:32:57,440
more sensibly,

918
00:32:57,440 --> 00:32:59,278
like turning the breaker back on and

919
00:32:59,278 --> 00:33:00,720
making sure the critical instruments

920
00:33:00,720 --> 00:33:02,319
were always energised,

921
00:33:02,319 --> 00:33:04,000
valve positions and manual override and

922
00:33:04,000 --> 00:33:06,558
such. The truth is though that

923
00:33:06,558 --> 00:33:08,798
those events were just the final straw.

924
00:33:08,798 --> 00:33:10,720
This pipe leg wasn't subjected to any

925
00:33:10,720 --> 00:33:12,720
pressure it hadn't been previously

926
00:33:12,720 --> 00:33:14,558
and it was still within 10% over

927
00:33:14,558 --> 00:33:16,720
the MaOP, which in a properly constructed

928
00:33:16,720 --> 00:33:19,119
pipeline wouldn't have led to a failure.

929
00:33:19,119 --> 00:33:21,919
Had the UPS over-pressurisation not

930
00:33:21,919 --> 00:33:22,960
occurred,

931
00:33:22,960 --> 00:33:24,720
if they hadn't found the defective pipe

932
00:33:24,720 --> 00:33:26,558
segment in the next 12 to 24 months or

933
00:33:26,558 --> 00:33:27,038
so,

934
00:33:27,038 --> 00:33:28,960
it's very likely that it would have let

935
00:33:28,960 --> 00:33:31,200
go through normal operational spikes and

936
00:33:31,200 --> 00:33:32,798
surges.

937
00:33:32,798 --> 00:33:35,200
Specifically in the report though, noting

938
00:33:35,200 --> 00:33:36,319
this one:

939
00:33:36,319 --> 00:33:39,359
(and I quote): "Inline inspection technology

940
00:33:39,359 --> 00:33:41,119
is not available for use in all

941
00:33:41,119 --> 00:33:42,880
currently operating gas transmission

942
00:33:42,880 --> 00:33:44,240
pipeline systems.

943
00:33:44,240 --> 00:33:46,079
Operators do not have the benefit of a

944
00:33:46,079 --> 00:33:48,000
uniquely effective assessment tool

945
00:33:48,000 --> 00:33:49,759
to identify and assess the threat from

946
00:33:49,759 --> 00:33:52,319
critical defects in their pipelines."

947
00:33:52,319 --> 00:33:56,398
(end quote) That is a really good point.

948
00:33:56,398 --> 00:33:58,640
When i first read about this incident i

949
00:33:58,640 --> 00:34:00,000
thought immediately,

950
00:34:00,000 --> 00:34:01,679
"Why didn't they just use an inspection

951
00:34:01,679 --> 00:34:03,440
PIG?"

952
00:34:03,440 --> 00:34:07,038
PIGs are Pipeline Inspection Gauges,

953
00:34:07,038 --> 00:34:09,358
and are basically a big cylindrical (not

954
00:34:09,358 --> 00:34:10,480
quite a plug)

955
00:34:10,480 --> 00:34:12,398
that you load into a working pipeline at

956
00:34:12,398 --> 00:34:13,519
the upstream end,

957
00:34:13,519 --> 00:34:15,039
build back pressure behind it in a

958
00:34:15,039 --> 00:34:17,599
launcher, which is essentially just a

959
00:34:17,599 --> 00:34:19,119
large chamber for loading and unloading a

960
00:34:19,119 --> 00:34:21,280
PIG, and then you open the valve and let

961
00:34:21,280 --> 00:34:22,960
it fly into the pipeline.

962
00:34:22,960 --> 00:34:25,358
It's pushed through the pipeline from

963
00:34:25,358 --> 00:34:27,039
the gas pressure behind it

964
00:34:27,039 --> 00:34:28,878
and it travels down the other end of the

965
00:34:28,878 --> 00:34:30,639
pipeline to the receiver

966
00:34:30,639 --> 00:34:32,398
and they're commonly used for cleaning

967
00:34:32,398 --> 00:34:33,679
things out of the pipe like

968
00:34:33,679 --> 00:34:36,719
Tri-Ethylene Glycol (TEG), or in other cases

969
00:34:36,719 --> 00:34:38,639
they have a 360°

970
00:34:38,639 --> 00:34:41,679
X-ray machine built into them. They

971
00:34:41,679 --> 00:34:44,320
literally take a continuous X-ray of the

972
00:34:44,320 --> 00:34:45,918
inside of the pipeline looking for

973
00:34:45,918 --> 00:34:48,800
defects in welds and pipe work corrosion,

974
00:34:48,800 --> 00:34:50,079
everywhere from the launcher to the

975
00:34:50,079 --> 00:34:53,119
receiver, recording them all on board

976
00:34:53,119 --> 00:34:54,639
and the engineers pour over them when

977
00:34:54,639 --> 00:34:56,320
they're done at the end.

978
00:34:56,320 --> 00:34:59,519
Very, extremely cool stuff! Not cheap

979
00:34:59,519 --> 00:35:00,559
though.

980
00:35:00,559 --> 00:35:02,480
They're technically called an ILI, an

981
00:35:02,480 --> 00:35:04,639
In-Line Inspection PIG.

982
00:35:04,639 --> 00:35:07,280
Of course they only work if the pipeline

983
00:35:07,280 --> 00:35:09,039
is much the same size

984
00:35:09,039 --> 00:35:11,760
along its length and have pig launchers

985
00:35:11,760 --> 00:35:12,880
and receivers

986
00:35:12,880 --> 00:35:14,800
and have a maximum bend radius for all

987
00:35:14,800 --> 00:35:16,239
the sections you want the PIG to travel

988
00:35:16,239 --> 00:35:17,280
through.

989
00:35:17,280 --> 00:35:18,719
Too many corners obviously it's not

990
00:35:18,719 --> 00:35:20,719
going to work. Line 132

991
00:35:20,719 --> 00:35:23,838
didn't qualify for pigging, or certainly

992
00:35:23,838 --> 00:35:27,280
Segment 180 didn't. So leak detection

993
00:35:27,280 --> 00:35:27,679
then

994
00:35:27,679 --> 00:35:29,920
is down to pressurizing and testing for

995
00:35:29,920 --> 00:35:31,199
a loss of pressure or

996
00:35:31,199 --> 00:35:33,679
by using tracer gas and gas sniffers at

997
00:35:33,679 --> 00:35:35,199
ground level.

998
00:35:35,199 --> 00:35:37,920
Not as accurate. Not as effective. So

999
00:35:37,920 --> 00:35:40,400
despite all of this,

1000
00:35:40,400 --> 00:35:43,039
why would you build it in a residential

1001
00:35:43,039 --> 00:35:44,239
neighborhood?

1002
00:35:44,239 --> 00:35:47,760
So let's look at that. California

1003
00:35:47,760 --> 00:35:50,239
Public Utilities Commission, San Bruno

1004
00:35:50,239 --> 00:35:51,838
Explosion Report of the Independent

1005
00:35:51,838 --> 00:35:54,559
Review Panel (2011) stated (and I quote):

1006
00:35:54,559 --> 00:35:57,920
"In 1956 the city converted PG&E's

1007
00:35:57,920 --> 00:35:59,280
pipeline Right Of Way

1008
00:35:59,280 --> 00:36:02,239
from an Easement to a Franchise Right as

1009
00:36:02,239 --> 00:36:03,760
the community was growing and

1010
00:36:03,760 --> 00:36:05,838
residential subdivisions were being laid

1011
00:36:05,838 --> 00:36:09,280
throughout the area." (end quote)

1012
00:36:09,280 --> 00:36:11,119
So what's an "Easement," at least in the

1013
00:36:11,119 --> 00:36:13,440
context of California law?

1014
00:36:13,440 --> 00:36:15,199
According to the California Land Title

1015
00:36:15,199 --> 00:36:18,079
Association an Easement is a: "...real estate

1016
00:36:18,079 --> 00:36:18,639
ownership

1017
00:36:18,639 --> 00:36:20,880
right, an encumbrance on the title

1018
00:36:20,880 --> 00:36:22,320
granted to an individual

1019
00:36:22,320 --> 00:36:24,320
or entity to make a limited but

1020
00:36:24,320 --> 00:36:26,159
typically indefinite use

1021
00:36:26,159 --> 00:36:28,880
of the land of another..." It goes on to

1022
00:36:28,880 --> 00:36:29,440
state

1023
00:36:29,440 --> 00:36:32,239
that: "...statutes may prohibit locating

1024
00:36:32,239 --> 00:36:32,880
buildings

1025
00:36:32,880 --> 00:36:35,280
or undertaking any construction over a

1026
00:36:35,280 --> 00:36:37,599
pipeline easement."

1027
00:36:37,599 --> 00:36:40,719
So a Right Of Way then is: "...the legal

1028
00:36:40,719 --> 00:36:41,280
right

1029
00:36:41,280 --> 00:36:43,599
established by usage or grant to pass

1030
00:36:43,599 --> 00:36:45,039
along a specific route

1031
00:36:45,039 --> 00:36:46,719
through grounds or property belonging to

1032
00:36:46,719 --> 00:36:48,159
another."

1033
00:36:48,159 --> 00:36:51,599
The term "Franchise Right" is interesting.

1034
00:36:51,599 --> 00:36:54,159
In California, Public Utilities are

1035
00:36:54,159 --> 00:36:55,199
defined as:

1036
00:36:55,199 --> 00:36:57,119
"An entity which provides a service to

1037
00:36:57,119 --> 00:36:58,960
the public for reasonable charges

1038
00:36:58,960 --> 00:37:00,480
whether such charges be in the form of

1039
00:37:00,480 --> 00:37:02,559
taxes or rates."

1040
00:37:02,559 --> 00:37:04,960
So utilities are granted statutory

1041
00:37:04,960 --> 00:37:05,760
franchise

1042
00:37:05,760 --> 00:37:08,320
rights, most often in the context of

1043
00:37:08,320 --> 00:37:09,760
let's say a water mains

1044
00:37:09,760 --> 00:37:11,838
under a city street, and this provides a

1045
00:37:11,838 --> 00:37:13,519
degree of legal protections against

1046
00:37:13,519 --> 00:37:15,199
consequential costs for repairs,

1047
00:37:15,199 --> 00:37:16,239
relocations,

1048
00:37:16,239 --> 00:37:19,199
amongst other things but in this context,

1049
00:37:19,199 --> 00:37:20,960
PG&E were operating

1050
00:37:20,960 --> 00:37:23,358
as a utility and were granted statutory

1051
00:37:23,358 --> 00:37:25,280
franchise rights for that section of the

1052
00:37:25,280 --> 00:37:26,480
pipeline,

1053
00:37:26,480 --> 00:37:28,239
rather than the stricter Right Of Way

1054
00:37:28,239 --> 00:37:30,320
condition that an easement would have

1055
00:37:30,320 --> 00:37:32,239
because under a local statute that would

1056
00:37:32,239 --> 00:37:33,838
have prevented anyone from building on

1057
00:37:33,838 --> 00:37:34,800
it.

1058
00:37:34,800 --> 00:37:38,079
So in summary, the city re-classified the

1059
00:37:38,079 --> 00:37:38,800
easement

1060
00:37:38,800 --> 00:37:40,639
so they could utilise what was the

1061
00:37:40,639 --> 00:37:42,239
Right Of Way easement,

1062
00:37:42,239 --> 00:37:45,760
so they could then build houses in 1957,

1063
00:37:45,760 --> 00:37:48,480
and to make way for that they regraded a

1064
00:37:48,480 --> 00:37:50,159
section of the ground to level it for

1065
00:37:50,159 --> 00:37:52,239
those houses and additional streets

1066
00:37:52,239 --> 00:37:53,599
requiring a short section of the

1067
00:37:53,599 --> 00:37:56,320
pipeline to be moved to a different spot,

1068
00:37:56,320 --> 00:37:58,639
which was the work that PG&E did

1069
00:37:58,639 --> 00:37:59,760
themselves

1070
00:37:59,760 --> 00:38:02,880
in 1956. The suburb

1071
00:38:02,880 --> 00:38:05,599
that was built there, Crestmoor, consists

1072
00:38:05,599 --> 00:38:09,199
of 2,500 homes built between 1957

1073
00:38:09,199 --> 00:38:12,639
and 1959. That was a decade after

1074
00:38:12,639 --> 00:38:15,039
the pipeline was originally finished and

1075
00:38:15,039 --> 00:38:17,440
had been in operation for that long.

1076
00:38:17,440 --> 00:38:21,039
Only 1 year though after Segment 180

1077
00:38:21,039 --> 00:38:24,400
was reinstalled, made from those 6

1078
00:38:24,400 --> 00:38:27,920
pieces of PUPS. So technically

1079
00:38:27,920 --> 00:38:29,679
they didn't bring the pipeline to the

1080
00:38:29,679 --> 00:38:32,000
neighborhood. The neighborhood came to

1081
00:38:32,000 --> 00:38:34,719
the pipeline.

1082
00:38:34,719 --> 00:38:37,440
Let's talk about the fallout. The cost to

1083
00:38:37,440 --> 00:38:38,960
repair the pipeline

1084
00:38:38,960 --> 00:38:42,320
was estimated to be $13.5M USD.

1085
00:38:42,320 --> 00:38:43,599
...

1086
00:38:43,599 --> 00:38:45,358
About $1/4M USD

1087
00:38:45,358 --> 00:38:46,800
worth of gas,

1088
00:38:46,800 --> 00:38:50,480
went up in smoke in 90min.

1089
00:38:50,480 --> 00:38:53,760
The CPUC fined the utility $1.6B USD

1090
00:38:53,760 --> 00:38:54,320
...

1091
00:38:54,320 --> 00:38:56,880
in connection with the disaster. That's

1092
00:38:56,880 --> 00:38:58,159
the largest penalty

1093
00:38:58,159 --> 00:39:01,920
ever levied against a California Utility.

1094
00:39:01,920 --> 00:39:05,440
PG&E paid over $550M USD to

1095
00:39:05,440 --> 00:39:07,679
the victims of the explosion.

1096
00:39:07,679 --> 00:39:09,679
PG&E were required to undertake pressure

1097
00:39:09,679 --> 00:39:12,800
tests and integrity checks of their 1,800mi

1098
00:39:12,800 --> 00:39:13,440
...

1099
00:39:13,440 --> 00:39:16,480
of transmission pipelines.

1100
00:39:16,480 --> 00:39:19,760
In August 2016, a federal jury convicted

1101
00:39:19,760 --> 00:39:22,239
PG&E on 5 charges of violating

1102
00:39:22,239 --> 00:39:24,639
federal pipeline safety regulations

1103
00:39:24,639 --> 00:39:27,358
and 1 charge of obstructing the NTSB

1104
00:39:27,358 --> 00:39:28,960
investigation.

1105
00:39:28,960 --> 00:39:32,000
In January 2017, a federal judge

1106
00:39:32,000 --> 00:39:34,480
sentenced PG&E for crimes linked to the

1107
00:39:34,480 --> 00:39:35,358
explosion,

1108
00:39:35,358 --> 00:39:37,599
imposing the maximum fine of $3M

1109
00:39:37,599 --> 00:39:38,559
USD,

1110
00:39:38,559 --> 00:39:41,599
and enforcing mandatory oversight by an

1111
00:39:41,599 --> 00:39:42,800
external monitor

1112
00:39:42,800 --> 00:39:46,079
(not the PUC) and 5 years

1113
00:39:46,079 --> 00:39:49,599
probation. No employees

1114
00:39:49,599 --> 00:39:52,960
of PG&E or the PUC were individually

1115
00:39:52,960 --> 00:39:54,159
sentenced.

1116
00:39:54,159 --> 00:39:56,400
No jail time was served by anyone

1117
00:39:56,400 --> 00:39:57,519
involved in this

1118
00:39:57,519 --> 00:40:01,280
incident. In the years since PG&E were

1119
00:40:01,280 --> 00:40:03,119
also found guilty in the Camp Fire

1120
00:40:03,119 --> 00:40:04,000
incident

1121
00:40:04,000 --> 00:40:05,440
and whose equipment also started

1122
00:40:05,440 --> 00:40:08,318
multiple wildfires, with subsequent fines

1123
00:40:08,318 --> 00:40:11,358
and class actions with $30B USD

1124
00:40:11,358 --> 00:40:14,639
in liabilities PG&E filed for

1125
00:40:14,639 --> 00:40:15,599
bankruptcy

1126
00:40:15,599 --> 00:40:18,079
for a second time in its history, in

1127
00:40:18,079 --> 00:40:20,639
January of 2019.

1128
00:40:20,639 --> 00:40:24,079
On Wednesday the 1st of July, 2020

1129
00:40:24,079 --> 00:40:26,400
only a few weeks ago at the time of

1130
00:40:26,400 --> 00:40:27,920
recording this,

1131
00:40:27,920 --> 00:40:31,599
PG&E exited Chapter 11 and is once again

1132
00:40:31,599 --> 00:40:33,358
out of bankruptcy.

1133
00:40:33,358 --> 00:40:35,760
Having said that 5 years after the

1134
00:40:35,760 --> 00:40:39,679
incident, PG&E had completed 673mi of

1135
00:40:39,679 --> 00:40:41,599
pressure testing of their gas pipeline

1136
00:40:41,599 --> 00:40:42,880
network.

1137
00:40:42,880 --> 00:40:45,358
They also replaced 115mi of

1138
00:40:45,358 --> 00:40:46,639
transmission pipeline,

1139
00:40:46,639 --> 00:40:50,000
down-rated 12mi, retired 9mi,

1140
00:40:50,000 --> 00:40:53,838
upgraded 202mi of pipeline to accept

1141
00:40:53,838 --> 00:40:57,239
In-Line Inspection PIGs, and installed

1142
00:40:57,239 --> 00:41:00,318
217 automatic valves

1143
00:41:00,318 --> 00:41:02,960
to improve pipe leg isolation. So what do

1144
00:41:02,960 --> 00:41:03,519
we learn

1145
00:41:03,519 --> 00:41:05,838
from all of this? Let's start with the

1146
00:41:05,838 --> 00:41:07,679
operational issues on the day.

1147
00:41:07,679 --> 00:41:09,838
Keeping in mind they contributed as the

1148
00:41:09,838 --> 00:41:11,679
trigger, but if they hadn't, the pipe

1149
00:41:11,679 --> 00:41:13,039
would have ruptured at some point in the

1150
00:41:13,039 --> 00:41:14,559
future anyway unless they caught it

1151
00:41:14,559 --> 00:41:15,679
first in a test.

1152
00:41:15,679 --> 00:41:17,280
That said the technicians at the

1153
00:41:17,280 --> 00:41:19,280
terminal doing the UPS work,

1154
00:41:19,280 --> 00:41:20,800
they shouldn't have been trying to

1155
00:41:20,800 --> 00:41:22,719
figure out the consequences of turning

1156
00:41:22,719 --> 00:41:24,318
power off on valves during the

1157
00:41:24,318 --> 00:41:25,358
changeover.

1158
00:41:25,358 --> 00:41:28,318
Why hadn't they pre-planned that? The

1159
00:41:28,318 --> 00:41:30,159
sequence of events that jumped out to me

1160
00:41:30,159 --> 00:41:31,440
as well in all of this,

1161
00:41:31,440 --> 00:41:33,280
why was there no dedicated operator as

1162
00:41:33,280 --> 00:41:34,559
the point of contact?

1163
00:41:34,559 --> 00:41:36,318
There was no SCADA operator as the sole

1164
00:41:36,318 --> 00:41:38,480
point of contact for those works!

1165
00:41:38,480 --> 00:41:39,920
They were constantly speaking to

1166
00:41:39,920 --> 00:41:42,239
different operators, through the entire

1167
00:41:42,239 --> 00:41:45,358
process, which led to inconsistent

1168
00:41:45,358 --> 00:41:46,880
answers, a lack of common understanding

1169
00:41:46,880 --> 00:41:48,079
in the SCADA control room

1170
00:41:48,079 --> 00:41:50,318
about what was actually happening. It was

1171
00:41:50,318 --> 00:41:52,880
just poorly coordinated overall.

1172
00:41:52,880 --> 00:41:54,880
Even if they couldn't afford to assign

1173
00:41:54,880 --> 00:41:56,880
an operator to those works entirely,

1174
00:41:56,880 --> 00:41:58,079
they should have had at least 1 of the

1175
00:41:58,079 --> 00:41:59,599
4 SCADA operators as the point of

1176
00:41:59,599 --> 00:42:01,199
contact for the works.

1177
00:42:01,199 --> 00:42:04,159
Besides that though, engineering records

1178
00:42:04,159 --> 00:42:05,039
are critical.

1179
00:42:05,039 --> 00:42:07,119
Design details that can be traced and

1180
00:42:07,119 --> 00:42:08,800
linked to inspection

1181
00:42:08,800 --> 00:42:11,679
records. These are critical pieces of

1182
00:42:11,679 --> 00:42:12,400
information

1183
00:42:12,400 --> 00:42:14,159
that so many companies seem to treat as

1184
00:42:14,159 --> 00:42:15,440
an afterthought.

1185
00:42:15,440 --> 00:42:17,838
When operational companies execute works

1186
00:42:17,838 --> 00:42:19,199
entirely in-house,

1187
00:42:19,199 --> 00:42:21,119
the first things that seem to get cut

1188
00:42:21,119 --> 00:42:22,639
are the engineering records.

1189
00:42:22,639 --> 00:42:25,679
The design documents. External companies

1190
00:42:25,679 --> 00:42:26,159
develop

1191
00:42:26,159 --> 00:42:28,159
design documentations and keep test

1192
00:42:28,159 --> 00:42:30,000
records as evidence or proof

1193
00:42:30,000 --> 00:42:32,239
for progress claims as well as for legal

1194
00:42:32,239 --> 00:42:33,039
reasons,

1195
00:42:33,039 --> 00:42:35,679
but operational companies don't feel the

1196
00:42:35,679 --> 00:42:36,400
need to,

1197
00:42:36,400 --> 00:42:38,400
which may not be needed for progress

1198
00:42:38,400 --> 00:42:39,519
claims (because there aren't any it's

1199
00:42:39,519 --> 00:42:40,239
internal)

1200
00:42:40,239 --> 00:42:41,838
but they still need to be kept for legal

1201
00:42:41,838 --> 00:42:44,559
reasons. in the competition for BAU

1202
00:42:44,559 --> 00:42:47,599
(Business As Usual), and project work, all

1203
00:42:47,599 --> 00:42:48,318
too often

1204
00:42:48,318 --> 00:42:51,039
corners get cut on design. And

1205
00:42:51,039 --> 00:42:53,679
documentation on internal projects

1206
00:42:53,679 --> 00:42:55,920
was clearly a corner that was cut in the

1207
00:42:55,920 --> 00:42:57,599
1956 work

1208
00:42:57,599 --> 00:43:01,280
on Segment 180. Finally isolation

1209
00:43:01,280 --> 00:43:04,559
or rather a lack of it. When the pipeline

1210
00:43:04,559 --> 00:43:05,760
was built,

1211
00:43:05,760 --> 00:43:08,318
remotely operated valves did exist but

1212
00:43:08,318 --> 00:43:08,880
they were

1213
00:43:08,880 --> 00:43:11,280
very expensive to install and not very

1214
00:43:11,280 --> 00:43:13,119
reliable but manual valves were still

1215
00:43:13,119 --> 00:43:14,079
installed.

1216
00:43:14,079 --> 00:43:15,679
Driving to the sites and manually

1217
00:43:15,679 --> 00:43:18,159
closing the valves takes precious time,

1218
00:43:18,159 --> 00:43:19,920
that could have been saved if they had

1219
00:43:19,920 --> 00:43:21,760
remote actuators fitted to those manual

1220
00:43:21,760 --> 00:43:22,239
valves,

1221
00:43:22,239 --> 00:43:25,119
hence automating them. It also seems

1222
00:43:25,119 --> 00:43:26,000
unbelievable

1223
00:43:26,000 --> 00:43:27,519
that the SCADA operators didn't

1224
00:43:27,519 --> 00:43:29,760
realize that there was a rupture for so

1225
00:43:29,760 --> 00:43:31,838
long after it had occurred.

1226
00:43:31,838 --> 00:43:33,838
The readings at Martin Station dropped

1227
00:43:33,838 --> 00:43:35,119
so low, so

1228
00:43:35,119 --> 00:43:38,000
quickly, they should have realised but it

1229
00:43:38,000 --> 00:43:38,719
was

1230
00:43:38,719 --> 00:43:40,960
close to 20min before they did and

1231
00:43:40,960 --> 00:43:42,719
when they did they couldn't pinpoint

1232
00:43:42,719 --> 00:43:43,280
where

1233
00:43:43,280 --> 00:43:46,239
just somewhere on Line 132...is a bit

1234
00:43:46,239 --> 00:43:47,199
vague.

1235
00:43:47,199 --> 00:43:50,079
So if they'd automated more of those

1236
00:43:50,079 --> 00:43:50,880
mainline

1237
00:43:50,880 --> 00:43:53,199
isolation valves they would have also

1238
00:43:53,199 --> 00:43:54,079
fitted local

1239
00:43:54,079 --> 00:43:55,599
upstream and downstream pressure

1240
00:43:55,599 --> 00:43:57,519
monitoring at those valve locations,

1241
00:43:57,519 --> 00:43:59,039
that's standard practice,

1242
00:43:59,039 --> 00:44:00,480
and that would have made pin-pointing the

1243
00:44:00,480 --> 00:44:03,199
loss of containment significantly faster.

1244
00:44:03,199 --> 00:44:05,440
With automated valve actuators fitted to

1245
00:44:05,440 --> 00:44:07,358
the manual valves they closed

1246
00:44:07,358 --> 00:44:08,960
and pressure transmitters fitted at those

1247
00:44:08,960 --> 00:44:11,199
locations, they could have isolated that

1248
00:44:11,199 --> 00:44:13,519
pipeleg in only a few minutes.

1249
00:44:13,519 --> 00:44:16,880
Not 1-1/2hrs. Some countries

1250
00:44:16,880 --> 00:44:18,719
around the world have legislation that

1251
00:44:18,719 --> 00:44:20,880
restricts the size of gas pipelines that

1252
00:44:20,880 --> 00:44:21,519
can be built

1253
00:44:21,519 --> 00:44:23,838
in or near residential areas, but it's

1254
00:44:23,838 --> 00:44:26,079
inconsistent between countries and even

1255
00:44:26,079 --> 00:44:28,239
states in the same country. In an

1256
00:44:28,239 --> 00:44:29,838
environment where developers want to

1257
00:44:29,838 --> 00:44:30,480
build

1258
00:44:30,480 --> 00:44:32,719
where the flattest land is and where the

1259
00:44:32,719 --> 00:44:33,679
rules have

1260
00:44:33,679 --> 00:44:36,239
loopholes, there are pressures for local

1261
00:44:36,239 --> 00:44:38,159
governments to rezone and reclassify

1262
00:44:38,159 --> 00:44:39,679
vast tracts of land

1263
00:44:39,679 --> 00:44:41,039
that probably shouldn't have been

1264
00:44:41,039 --> 00:44:43,760
reclassified for all sorts of reasons.

1265
00:44:43,760 --> 00:44:45,679
Easements are generally made easements

1266
00:44:45,679 --> 00:44:48,159
for a good reason.

1267
00:44:48,159 --> 00:44:50,800
Then again the original mistake: the

1268
00:44:50,800 --> 00:44:52,960
poorly manufactured pipe leg segment

1269
00:44:52,960 --> 00:44:56,000
made from those 6 PUPS, was made

1270
00:44:56,000 --> 00:44:58,318
before more than half of these victims

1271
00:44:58,318 --> 00:44:59,039
were even

1272
00:44:59,039 --> 00:45:02,800
born. Like the 737MAX,

1273
00:45:02,800 --> 00:45:05,358
grandfathering provisions just need to

1274
00:45:05,358 --> 00:45:06,159
go away

1275
00:45:06,159 --> 00:45:08,318
from so many regulations including these

1276
00:45:08,318 --> 00:45:09,199
ones.

1277
00:45:09,199 --> 00:45:11,760
When you're designing a gas pipeline, you

1278
00:45:11,760 --> 00:45:13,519
have to think about containment.

1279
00:45:13,519 --> 00:45:15,519
They have a saying in the business: "Keep

1280
00:45:15,519 --> 00:45:16,559
it in the pipes."

1281
00:45:16,559 --> 00:45:19,358
That's your job. Because whatever's in

1282
00:45:19,358 --> 00:45:20,880
that pipe if it gets out

1283
00:45:20,880 --> 00:45:23,280
it can kill people, hurt people, destroy

1284
00:45:23,280 --> 00:45:24,239
property, cause

1285
00:45:24,239 --> 00:45:26,719
environmental damage, and yeah okay

1286
00:45:26,719 --> 00:45:27,838
there's money too

1287
00:45:27,838 --> 00:45:29,280
because you're going to lose money because

1288
00:45:29,280 --> 00:45:31,039
your product is getting out...never gets

1289
00:45:31,039 --> 00:45:33,039
to a customer, so you can't sell it.

1290
00:45:33,039 --> 00:45:35,280
Also if those customers are relying on

1291
00:45:35,280 --> 00:45:36,079
that product

1292
00:45:36,079 --> 00:45:39,039
for heating, sterilization in hospitals,

1293
00:45:39,039 --> 00:45:40,318
cooking and restaurants,

1294
00:45:40,318 --> 00:45:42,239
I mean you name it, there's lots of other

1295
00:45:42,239 --> 00:45:44,318
outcomes as well.

1296
00:45:44,318 --> 00:45:46,559
Gas pipelines can be built and operated

1297
00:45:46,559 --> 00:45:49,039
safely if you do it correctly,

1298
00:45:49,039 --> 00:45:52,000
but you have to inspect your pipes. You

1299
00:45:52,000 --> 00:45:53,679
have to know their condition.

1300
00:45:53,679 --> 00:45:56,960
That they're safe to continue to operate.

1301
00:45:56,960 --> 00:45:59,119
So they couldn't use an inspection PIG

1302
00:45:59,119 --> 00:46:01,280
in 2010,

1303
00:46:01,280 --> 00:46:04,079
but they could have hydro-tested it. So

1304
00:46:04,079 --> 00:46:06,000
why didn't they?

1305
00:46:06,000 --> 00:46:09,519
It's money. You have to empty the pipe

1306
00:46:09,519 --> 00:46:12,239
of the gas. Sometimes you need to purge

1307
00:46:12,239 --> 00:46:14,159
it with an inert gas:

1308
00:46:14,159 --> 00:46:16,079
some will debate that one. Fill it with

1309
00:46:16,079 --> 00:46:18,639
water. Pressurise it and test it.

1310
00:46:18,639 --> 00:46:20,960
If it fails you have to trace the leak,

1311
00:46:20,960 --> 00:46:21,838
dig it up,

1312
00:46:21,838 --> 00:46:24,639
patch it, fix it, refill the pipe, re-test

1313
00:46:24,639 --> 00:46:25,760
the pipe,

1314
00:46:25,760 --> 00:46:27,838
drain the pipe, dry the pipe, and then you

1315
00:46:27,838 --> 00:46:29,280
can put gas back in it again.

1316
00:46:29,280 --> 00:46:31,838
That whole process takes time and all

1317
00:46:31,838 --> 00:46:33,519
that time you're not moving gas,

1318
00:46:33,519 --> 00:46:36,159
you're not making money. So if you don't

1319
00:46:36,159 --> 00:46:37,599
mandate it,

1320
00:46:37,599 --> 00:46:42,318
they won't do it. When you're an engineer,

1321
00:46:42,318 --> 00:46:43,599
building something that could be an

1322
00:46:43,599 --> 00:46:46,480
operation for a very long time, a very,

1323
00:46:46,480 --> 00:46:49,760
very long time, you have to make sure

1324
00:46:49,760 --> 00:46:50,639
that it's

1325
00:46:50,639 --> 00:46:53,760
designed to the required standard.

1326
00:46:53,760 --> 00:46:56,480
That that design is documented. That it's

1327
00:46:56,480 --> 00:46:57,920
built and tested

1328
00:46:57,920 --> 00:47:01,519
against that standard, as best you can,

1329
00:47:01,519 --> 00:47:03,519
from the right materials and test it as

1330
00:47:03,519 --> 00:47:05,280
thoroughly as you can.

1331
00:47:05,280 --> 00:47:08,159
Because the mistakes made today will

1332
00:47:08,159 --> 00:47:08,960
sometimes

1333
00:47:08,960 --> 00:47:12,000
lay in wait for 50yrs,

1334
00:47:12,000 --> 00:47:16,159
before they take innocent people's lives.

1335
00:47:16,960 --> 00:47:18,400
If you're enjoying Causality and you

1336
00:47:18,400 --> 00:47:20,079
want to support the show you can

1337
00:47:20,079 --> 00:47:21,920
by becoming a Patron. You can find

1338
00:47:21,920 --> 00:47:23,760
details at https://engineered.network/

1339
00:47:23,760 --> 00:47:25,760
causality, with a thank you to all of our

1340
00:47:25,760 --> 00:47:27,358
Patrons and a special thank you to our

1341
00:47:27,358 --> 00:47:29,440
Patreon Silver Producers Mitch Bielger,

1342
00:47:29,440 --> 00:47:31,920
John Whitlow, Joseph Antonio, Kevin Koch,

1343
00:47:31,920 --> 00:47:33,679
Oliver Steele and Shane O'Neill.

1344
00:47:33,679 --> 00:47:35,358
And an extra special thank you to our

1345
00:47:35,358 --> 00:47:37,519
Patron Gold Producer known only as

1346
00:47:37,519 --> 00:47:39,838
'r'. Patrons have access to early release,

1347
00:47:39,838 --> 00:47:41,599
high quality ad-free episodes as well as

1348
00:47:41,599 --> 00:47:44,000
bonus episodes and to Causality Explored.

1349
00:47:44,000 --> 00:47:45,679
You can do this via Patreon just visit

1350
00:47:45,679 --> 00:47:47,358
https://engineered.network/causality to

1351
00:47:47,358 --> 00:47:47,838
learn how

1352
00:47:47,838 --> 00:47:49,838
you can help this show to continue to be

1353
00:47:49,838 --> 00:47:52,079
made. Of course there's lots of other

1354
00:47:52,079 --> 00:47:53,838
ways to help like favouriting this episode

1355
00:47:53,838 --> 00:47:55,358
in your podcast player app

1356
00:47:55,358 --> 00:47:57,199
or sharing the episode or the show with

1357
00:47:57,199 --> 00:47:58,960
your friends or via social.

1358
00:47:58,960 --> 00:48:01,119
Some podcast players let you share audio

1359
00:48:01,119 --> 00:48:02,559
clips of the episode so if you have a

1360
00:48:02,559 --> 00:48:04,239
few favourite segments feel free to share

1361
00:48:04,239 --> 00:48:04,880
them too.

1362
00:48:04,880 --> 00:48:06,639
All these things help others discover

1363
00:48:06,639 --> 00:48:08,559
the show and can make a big difference.

1364
00:48:08,559 --> 00:48:10,400
Causality is heavily researched and

1365
00:48:10,400 --> 00:48:12,079
links to all the materials used for the

1366
00:48:12,079 --> 00:48:13,838
creation of this episode are contained

1367
00:48:13,838 --> 00:48:14,639
in the show notes.

1368
00:48:14,639 --> 00:48:16,000
You can find them in the text of the

1369
00:48:16,000 --> 00:48:18,079
episode description, on your podcast

1370
00:48:18,079 --> 00:48:19,599
player or on our website.

1371
00:48:19,599 --> 00:48:21,519
You can follow me on the Fediverse

1372
00:48:21,519 --> 00:48:23,280
@chidgey@engineered.space

1373
00:48:23,280 --> 00:48:25,679
on Twitter @johnchidgey or

1374
00:48:25,679 --> 00:48:26,318
the network

1375
00:48:26,318 --> 00:48:28,880
@Engineered_Net. This was

1376
00:48:28,880 --> 00:48:30,079
Causality.

1377
00:48:30,079 --> 00:48:32,838
I'm John Chidgey. Thanks so much for

1378
00:48:32,838 --> 00:48:35,838
listening.

1379
00:49:03,480 --> 00:49:26,219
[Music]

1380
00:49:32,559 --> 00:49:34,639
[Music]