1
00:00:00,000 --> 00:00:11,040
[Music]

2
00:00:11,040 --> 00:00:12,162
Chain of events,

3
00:00:12,162 --> 00:00:14,163
cause and effect, we analyze what went

4
00:00:14,163 --> 00:00:15,522
right and what went wrong as we

5
00:00:15,522 --> 00:00:17,042
discover that many outcomes can be

6
00:00:17,042 --> 00:00:17,764
predicted,

7
00:00:17,764 --> 00:00:20,484
planned for, and even prevented. I'm John

8
00:00:20,484 --> 00:00:21,203
Chidgey

9
00:00:21,203 --> 00:00:23,444
and this is Causality. Causality is

10
00:00:23,444 --> 00:00:24,404
supported by you:

11
00:00:24,404 --> 00:00:25,926
our listeners. If you'd like to support

12
00:00:25,926 --> 00:00:28,725
the show you can by becoming a Patron.

13
00:00:28,725 --> 00:00:30,405
Patrons have access to early release

14
00:00:30,405 --> 00:00:32,244
high quality ad free episodes as well as

15
00:00:32,244 --> 00:00:33,446
other bonus material

16
00:00:33,446 --> 00:00:35,687
you can do this via Patreon just visit

17
00:00:35,687 --> 00:00:36,888
engineered.network/

18
00:00:36,888 --> 00:00:39,048
causality to learn how you can help this

19
00:00:39,048 --> 00:00:40,887
show to continue to be made.

20
00:00:40,887 --> 00:00:45,049
Thank you! "Big Dig"

21
00:00:45,049 --> 00:00:47,370
This incident is often referred to as

22
00:00:47,370 --> 00:00:48,248
the "Big Dig"

23
00:00:48,248 --> 00:00:51,209
collapse or the "Big Dig" tunnel ceiling

24
00:00:51,209 --> 00:00:52,012
collapse

25
00:00:52,012 --> 00:00:53,771
despite the fact it had nothing directly

26
00:00:53,771 --> 00:00:57,132
to do with the Big Dig in its entirety

27
00:00:57,132 --> 00:00:59,614
but rather the incident related more to

28
00:00:59,614 --> 00:01:00,251
a single

29
00:01:00,251 --> 00:01:03,133
exit portal. The Big Dig itself was a

30
00:01:03,133 --> 00:01:05,453
large civilian construction project

31
00:01:05,453 --> 00:01:08,015
that took 15 years from start to finish,

32
00:01:08,015 --> 00:01:09,296
though the project was technically

33
00:01:09,296 --> 00:01:10,016
referred to

34
00:01:10,016 --> 00:01:13,218
as the C.A.T. Project or the Central Artery

35
00:01:13,218 --> 00:01:14,817
Tunnel Project.

36
00:01:14,817 --> 00:01:16,658
The project took twice as long as the

37
00:01:16,658 --> 00:01:18,457
original estimate. Cost

38
00:01:18,457 --> 00:01:21,136
$14.6B US Dollars which was more

39
00:01:21,136 --> 00:01:23,539
than double the 1996 $1B dollar

40
00:01:23,539 --> 00:01:24,580
estimate and 5

41
00:01:24,580 --> 00:01:26,498
times the $3B original

42
00:01:26,498 --> 00:01:28,900
project estimate in 1983.

43
00:01:28,900 --> 00:01:31,221
Traffic congestion in Boston had long

44
00:01:31,221 --> 00:01:32,901
been a problem as the city had grown

45
00:01:32,901 --> 00:01:34,821
from a time before cars existed

46
00:01:34,821 --> 00:01:36,741
with some studies in the 80s suggesting

47
00:01:36,741 --> 00:01:38,903
that by 2010 Bostons rush hour could

48
00:01:38,903 --> 00:01:41,704
last approximately 16 hours every single

49
00:01:41,704 --> 00:01:42,665
day.

50
00:01:42,665 --> 00:01:45,225
To address this a new bridge and two key

51
00:01:45,225 --> 00:01:46,985
tunnels were proposed to be built

52
00:01:46,985 --> 00:01:49,143
as a concerted set of projects which

53
00:01:49,143 --> 00:01:50,667
were collectively referred to

54
00:01:50,667 --> 00:01:53,385
as the "Big Dig". The first tunnel was

55
00:01:53,385 --> 00:01:55,145
named the Ted Williams tunnel

56
00:01:55,145 --> 00:01:57,068
and goes beneath Boston harbour taking

57
00:01:57,068 --> 00:01:58,908
interstate 90 (the I90)

58
00:01:58,908 --> 00:02:01,229
traffic from South Boston to Boston

59
00:02:01,229 --> 00:02:03,388
Logan International Airport. Construction

60
00:02:03,388 --> 00:02:05,549
began in 1991 and was completed

61
00:02:05,549 --> 00:02:08,510
in 1995. The second tunnel was located

62
00:02:08,510 --> 00:02:10,511
below the Fort Point Channel and formed

63
00:02:10,511 --> 00:02:13,232
part of an I90 connector with the I93

64
00:02:13,232 --> 00:02:14,113
interchange.

65
00:02:14,113 --> 00:02:15,393
It would effectively replace the

66
00:02:15,393 --> 00:02:17,952
elevated I93 Central Artery which had

67
00:02:17,952 --> 00:02:20,034
become deteriorated with age;

68
00:02:20,034 --> 00:02:22,352
extend the I90 also called the

69
00:02:22,352 --> 00:02:24,835
Massachusetts Turnpike or "Mass-Pike"

70
00:02:24,835 --> 00:02:26,674
to the Logan International Airport

71
00:02:26,674 --> 00:02:28,596
providing an interchange for the I90

72
00:02:28,596 --> 00:02:29,876
and the I93

73
00:02:29,876 --> 00:02:32,754
and also replace the I93 bridge over

74
00:02:32,754 --> 00:02:33,718
the Charles River.

75
00:02:33,718 --> 00:02:36,438
In 1985 the department of public works

76
00:02:36,438 --> 00:02:38,196
entered into an agreement with Bechtel

77
00:02:38,196 --> 00:02:39,637
Parsons Bricknehoff

78
00:02:39,637 --> 00:02:43,077
hence referred to as BPB, a joint venture

79
00:02:43,077 --> 00:02:44,999
formed between Bechtel Corporation and

80
00:02:44,999 --> 00:02:47,161
Parsons Bricknerhoff, Quade and Douglas

81
00:02:47,161 --> 00:02:48,758
to prepare a preliminary project

82
00:02:48,758 --> 00:02:51,322
management plan. In 1990

83
00:02:51,322 --> 00:02:54,602
congress allocated $755M US

84
00:02:54,602 --> 00:02:56,043
dollars to the massive highway

85
00:02:56,043 --> 00:02:57,403
improvement project,

86
00:02:57,403 --> 00:02:59,081
and a year later the Federal Highway

87
00:02:59,081 --> 00:03:01,082
Administration gave its approval to move

88
00:03:01,082 --> 00:03:02,044
ahead.

89
00:03:02,044 --> 00:03:04,045
A massive civil construction project

90
00:03:04,045 --> 00:03:06,445
required significant and expensive

91
00:03:06,445 --> 00:03:08,846
traffic sequencing during construction.

92
00:03:08,846 --> 00:03:10,763
Traffic sequencing is a term sometimes

93
00:03:10,763 --> 00:03:12,607
used in road construction where traffic

94
00:03:12,607 --> 00:03:14,447
is directed via various detours,

95
00:03:14,447 --> 00:03:16,287
temporary roads, road direction

96
00:03:16,287 --> 00:03:18,608
reassignments and so on in accordance

97
00:03:18,608 --> 00:03:20,528
with a traffic management plan

98
00:03:20,528 --> 00:03:22,850
for the road construction project. In

99
00:03:22,850 --> 00:03:24,767
order to coordinate the construction

100
00:03:24,767 --> 00:03:25,569
sequence

101
00:03:25,569 --> 00:03:27,331
the approved traffic plan required a

102
00:03:27,331 --> 00:03:29,171
temporary above ground ramp

103
00:03:29,171 --> 00:03:31,892
to carry traffic from D Street to the

104
00:03:31,892 --> 00:03:33,971
entrance of the Ted Williams Tunnel

105
00:03:33,971 --> 00:03:35,972
in order to provide access under Boston

106
00:03:35,972 --> 00:03:37,572
Harbour to the airport

107
00:03:37,572 --> 00:03:39,173
before work was finished on the

108
00:03:39,173 --> 00:03:41,812
remainder of the I90 connector.

109
00:03:41,812 --> 00:03:43,573
This would be achieved by constructing a

110
00:03:43,573 --> 00:03:45,175
portal under D Street

111
00:03:45,175 --> 00:03:47,254
that would eventually connect through

112
00:03:47,254 --> 00:03:48,773
but was completed first

113
00:03:48,773 --> 00:03:51,576
in 1993 well before the completion of

114
00:03:51,576 --> 00:03:53,177
either the Ted Williams Tunnel

115
00:03:53,177 --> 00:03:55,656
or the remainder of the I90 connector.

116
00:03:55,656 --> 00:03:56,935
For those not familiar

117
00:03:56,935 --> 00:03:59,419
a portal is the name used for tunnel

118
00:03:59,419 --> 00:04:00,216
entries and

119
00:04:00,216 --> 00:04:03,339
exits. The D Street portal...

120
00:04:03,339 --> 00:04:05,740
(yes Boston has a series of streets that

121
00:04:05,740 --> 00:04:07,421
are just an alphabetical letter...

122
00:04:07,421 --> 00:04:10,621
A, B, C, D Street and so on so there's that...)

123
00:04:10,621 --> 00:04:11,021
The D

124
00:04:11,021 --> 00:04:13,181
Street portal had a heavily reinforced

125
00:04:13,181 --> 00:04:14,219
concrete roof

126
00:04:14,219 --> 00:04:16,860
slab varying between 1.5 meters to 2

127
00:04:16,860 --> 00:04:18,541
meters (that's 5 to 7ft)

128
00:04:18,541 --> 00:04:21,422
thick in order to accommodate the

129
00:04:21,422 --> 00:04:21,823
weight

130
00:04:21,823 --> 00:04:23,745
of a parking deck that was originally

131
00:04:23,745 --> 00:04:25,024
intended to be built

132
00:04:25,024 --> 00:04:28,066
over the portal. With that background

133
00:04:28,066 --> 00:04:31,025
let's talk about the incident itself. At

134
00:04:31,025 --> 00:04:32,706
approximately 11pm

135
00:04:32,706 --> 00:04:34,787
local Daylight Savings Time on Monday

136
00:04:34,787 --> 00:04:36,868
the 10th of July, 2006

137
00:04:36,868 --> 00:04:39,507
an entire row consisting of 20 anchor

138
00:04:39,507 --> 00:04:41,268
bolts pulled free from a suspended

139
00:04:41,268 --> 00:04:42,229
ceiling segment

140
00:04:42,229 --> 00:04:44,468
in the D Street portal. The ceiling

141
00:04:44,468 --> 00:04:45,988
consisted of a series of four

142
00:04:45,988 --> 00:04:49,270
3.6m (that's 12ft) wide by 2.4m

143
00:04:49,270 --> 00:04:50,950
(that's 8ft) long

144
00:04:50,950 --> 00:04:53,752
by 100mm or 4" thick

145
00:04:53,752 --> 00:04:54,632
3 tonnes

146
00:04:54,632 --> 00:04:58,152
or 4,700lbs of reinforced concrete

147
00:04:58,152 --> 00:04:59,513
and steel panels.

148
00:04:59,513 --> 00:05:01,433
They provided segregation between the

149
00:05:01,433 --> 00:05:03,674
roadway and the ventilation system above

150
00:05:03,674 --> 00:05:04,875
that roadway.

151
00:05:04,875 --> 00:05:06,954
With an entire row no longer supporting

152
00:05:06,954 --> 00:05:08,635
that side of the ceiling panels they

153
00:05:08,635 --> 00:05:09,676
began to rotate

154
00:05:09,676 --> 00:05:11,595
downwards with the additional load of

155
00:05:11,595 --> 00:05:13,355
the rotated panels causing rapid

156
00:05:13,355 --> 00:05:15,355
fracturing of the concrete.

157
00:05:15,355 --> 00:05:18,078
One minute later a 1991 black Buick

158
00:05:18,078 --> 00:05:18,717
sedan

159
00:05:18,717 --> 00:05:21,838
driven by Mr Angel Del Valle, aged 46

160
00:05:21,838 --> 00:05:24,240
with the sole passenger his wife Melina

161
00:05:24,240 --> 00:05:25,118
aged 38,

162
00:05:25,118 --> 00:05:26,959
was traveling eastbound in the

163
00:05:26,959 --> 00:05:29,200
I90 connector, en route to the

164
00:05:29,200 --> 00:05:31,280
Logan International airport to pick up

165
00:05:31,280 --> 00:05:33,202
Angel's brother and sister-in-law

166
00:05:33,202 --> 00:05:35,442
who returning from a trip to Puerto Rico.

167
00:05:35,442 --> 00:05:37,042
As the vehicle approached the exit

168
00:05:37,042 --> 00:05:37,683
portal

169
00:05:37,683 --> 00:05:39,684
four panels of the suspended concrete

170
00:05:39,684 --> 00:05:41,602
ceiling detached from the tunnel roof

171
00:05:41,602 --> 00:05:44,085
and fell onto the vehicle. The panels

172
00:05:44,085 --> 00:05:44,644
crushed

173
00:05:44,644 --> 00:05:46,325
the right-hand side of the vehicle's

174
00:05:46,325 --> 00:05:48,485
roof as the car came to rest against the

175
00:05:48,485 --> 00:05:49,924
Northern tunnel wall.

176
00:05:49,924 --> 00:05:52,725
Approximately 26 tonnes of concrete and

177
00:05:52,725 --> 00:05:53,285
steel

178
00:05:53,285 --> 00:05:55,606
had fallen onto both the vehicle and the

179
00:05:55,606 --> 00:05:57,208
roadway.

180
00:05:57,208 --> 00:05:59,449
Mr Del Valle was able to escape through the

181
00:05:59,449 --> 00:06:01,366
driver's side window sustaining only

182
00:06:01,366 --> 00:06:03,127
minor injuries.

183
00:06:03,127 --> 00:06:06,810
Melina was crushed and died instantly.

184
00:06:06,810 --> 00:06:08,810
Following the incident Mr Del Valle

185
00:06:08,810 --> 00:06:10,490
stated that although he saw the concrete

186
00:06:10,490 --> 00:06:12,091
panels starting to come down from the

187
00:06:12,091 --> 00:06:12,652
roof,

188
00:06:12,652 --> 00:06:15,130
he was unable to avoid them in time.

189
00:06:15,130 --> 00:06:16,731
Immediately following the incident

190
00:06:16,731 --> 00:06:18,412
the underground sections were closed to

191
00:06:18,412 --> 00:06:20,012
all traffic pending a complete

192
00:06:20,012 --> 00:06:21,293
inspection.

193
00:06:21,293 --> 00:06:23,053
The first ramp to fully reopen was on

194
00:06:23,053 --> 00:06:25,613
the 9th of August, but the entirety of

195
00:06:25,613 --> 00:06:27,373
all of the underground sections weren't

196
00:06:27,373 --> 00:06:28,414
fully reopened

197
00:06:28,414 --> 00:06:30,415
until nearly 12 months after the

198
00:06:30,415 --> 00:06:32,496
incident occurred.

199
00:06:32,496 --> 00:06:35,616
So what went wrong? The work scopes

200
00:06:35,616 --> 00:06:38,258
were split between many companies with

201
00:06:38,258 --> 00:06:40,258
the primary oversight being Bechtel

202
00:06:40,258 --> 00:06:41,779
Parsons Bricknerhoff and the

203
00:06:41,779 --> 00:06:43,779
Massachusetts Turnpike Authority, that's

204
00:06:43,779 --> 00:06:44,820
the MTA,

205
00:06:44,820 --> 00:06:46,979
who took an increasing oversight role

206
00:06:46,979 --> 00:06:48,739
towards the end of the project,

207
00:06:48,739 --> 00:06:50,901
as each component was gradually handed

208
00:06:50,901 --> 00:06:52,020
over.

209
00:06:52,020 --> 00:06:53,541
Bechtel Parsons Bricknerhoff were

210
00:06:53,541 --> 00:06:55,380
essentially the secondary approver for

211
00:06:55,380 --> 00:06:56,020
the design

212
00:06:56,020 --> 00:06:58,343
of the project as a whole as well as its

213
00:06:58,343 --> 00:07:00,502
constituent components.

214
00:07:00,502 --> 00:07:02,182
The D Street portal itself was

215
00:07:02,182 --> 00:07:04,583
constructed by Kiewit/Perini/Atkinson/

216
00:07:04,583 --> 00:07:05,225
Cashman,

217
00:07:05,225 --> 00:07:07,384
a joint venture based in Boston, which

218
00:07:07,384 --> 00:07:09,065
was contracted by the Massachusetts

219
00:07:09,065 --> 00:07:10,906
department of public works to construct

220
00:07:10,906 --> 00:07:13,467
the land-side West tunnel approach to

221
00:07:13,467 --> 00:07:15,145
the Ted Williams Tunnel.

222
00:07:15,145 --> 00:07:16,827
The tunnel consisted of approximately

223
00:07:16,827 --> 00:07:19,228
800 meters that's 2600ft

224
00:07:19,228 --> 00:07:21,628
of cut and cover and one section of

225
00:07:21,628 --> 00:07:23,068
depressed open highway

226
00:07:23,068 --> 00:07:25,148
as well as the D Street portal and a

227
00:07:25,148 --> 00:07:26,110
temporary ramp

228
00:07:26,110 --> 00:07:27,870
required for traffic sequencing during

229
00:07:27,870 --> 00:07:30,271
construction of the larger tunnels.

230
00:07:30,271 --> 00:07:32,272
The section design consultant for this

231
00:07:32,272 --> 00:07:34,191
was HDR Engineering Incorporated

232
00:07:34,191 --> 00:07:36,833
headquartered in Omaha, Nebraska. The

233
00:07:36,833 --> 00:07:38,273
roadway in the Eastbound

234
00:07:38,273 --> 00:07:40,113
D Street portal tunnel at the accident

235
00:07:40,113 --> 00:07:41,711
site consisted of two

236
00:07:41,711 --> 00:07:44,674
12 foot or 3.6 meter wide travel lanes

237
00:07:44,674 --> 00:07:46,514
and along the South side of the tunnel

238
00:07:46,514 --> 00:07:48,994
an 18 foot or 5.4 meter wide

239
00:07:48,994 --> 00:07:50,995
acceleration lane.

240
00:07:50,995 --> 00:07:53,075
All tunnels beyond a set length require

241
00:07:53,075 --> 00:07:55,155
ventilation to ensure air quality is

242
00:07:55,155 --> 00:07:56,756
maintained and to correctly handle

243
00:07:56,756 --> 00:07:59,238
airflow during a fire emergency.

244
00:07:59,238 --> 00:08:01,396
During the original letting of the

245
00:08:01,396 --> 00:08:02,597
portal section

246
00:08:02,597 --> 00:08:04,037
no allowance had been made for

247
00:08:04,037 --> 00:08:05,799
ventilation as this component of the

248
00:08:05,799 --> 00:08:07,559
design was intended to be added in a

249
00:08:07,559 --> 00:08:08,921
later work scope.

250
00:08:08,921 --> 00:08:11,802
The portal design was, therefore, done in

251
00:08:11,802 --> 00:08:12,121
two

252
00:08:12,121 --> 00:08:14,441
parts and was executed by two different

253
00:08:14,441 --> 00:08:15,720
companies.

254
00:08:15,720 --> 00:08:17,482
The aforementioned joint venture

255
00:08:17,482 --> 00:08:19,322
constructed the cut and cover tunnel

256
00:08:19,322 --> 00:08:21,323
but the I90 connector tunnel

257
00:08:21,323 --> 00:08:23,323
"finishes" section design consultant

258
00:08:23,323 --> 00:08:25,084
which included the D Street portal

259
00:08:25,084 --> 00:08:25,723
finishing

260
00:08:25,723 --> 00:08:27,805
which included the ventilation system,

261
00:08:27,805 --> 00:08:29,723
was Gannett Fleming Incorporated

262
00:08:29,723 --> 00:08:32,285
headquartered in Camp Hill, Pennsylvania.

263
00:08:32,285 --> 00:08:33,725
For the suspended ceiling Gannett

264
00:08:33,725 --> 00:08:34,925
Fleming engineers followed the

265
00:08:34,925 --> 00:08:35,726
directions set

266
00:08:35,726 --> 00:08:37,568
in the design policy memorandum number

267
00:08:37,568 --> 00:08:39,328
107 by Fay,

268
00:08:39,328 --> 00:08:41,406
Spofford and Throndike, Inc

269
00:08:41,406 --> 00:08:42,848
(that's FST for short)

270
00:08:42,848 --> 00:08:45,329
Howard, Needles, Tammen & Bergendoff

271
00:08:45,329 --> 00:08:46,370
(HNTB

272
00:08:46,370 --> 00:08:48,687
for short) using the embedded steel

273
00:08:48,687 --> 00:08:49,807
channel inserts

274
00:08:49,807 --> 00:08:52,050
instead of roof girders for the roof

275
00:08:52,050 --> 00:08:53,169
attachments.

276
00:08:53,169 --> 00:08:55,572
The D Street portal slab was not

277
00:08:55,572 --> 00:08:57,732
constructed in the same way as the rest

278
00:08:57,732 --> 00:08:59,493
of the tunnel, with additional slab

279
00:08:59,493 --> 00:09:01,333
thickness and reinforcing intended for a

280
00:09:01,333 --> 00:09:01,811
different

281
00:09:01,811 --> 00:09:04,374
goal and did not have any embedded steel

282
00:09:04,374 --> 00:09:05,893
channel inserts or anchor points

283
00:09:05,893 --> 00:09:08,774
embedded into it in its initial design.

284
00:09:08,774 --> 00:09:10,775
Hence the design was adjusted to use

285
00:09:10,775 --> 00:09:13,497
adhesive anchors with an epoxy resin.

286
00:09:13,497 --> 00:09:15,656
This technique was also used in other

287
00:09:15,656 --> 00:09:17,497
sections of the I90 Connector tunnel

288
00:09:17,497 --> 00:09:19,016
and on the HOV

289
00:09:19,016 --> 00:09:21,094
ramp, that's High Occupancy Vehicle or

290
00:09:21,094 --> 00:09:22,695
HOV. The ceiling

291
00:09:22,695 --> 00:09:25,096
in the D Street portal was installed by

292
00:09:25,096 --> 00:09:27,260
Modern Continental Construction Company

293
00:09:27,260 --> 00:09:28,621
Incorporated

294
00:09:28,621 --> 00:09:30,697
(Modern Continental for short) of

295
00:09:30,697 --> 00:09:32,381
Cambridge, Massachusetts.

296
00:09:32,381 --> 00:09:34,781
It was constructed between 1999 and

297
00:09:34,781 --> 00:09:36,222
April of 2000.

298
00:09:36,222 --> 00:09:37,742
The ceiling module that collapsed in

299
00:09:37,742 --> 00:09:39,343
this incident was installed in November

300
00:09:39,343 --> 00:09:41,022
1999.

301
00:09:41,022 --> 00:09:42,942
Modern Continental installed the last

302
00:09:42,942 --> 00:09:44,701
ceiling module of the I90 Connector

303
00:09:44,701 --> 00:09:46,864
Tunnel in July 2002.

304
00:09:46,864 --> 00:09:49,024
Whilst the use of epoxy resins for

305
00:09:49,024 --> 00:09:50,944
anchoring applications was not unheard

306
00:09:50,944 --> 00:09:51,183
of

307
00:09:51,183 --> 00:09:53,985
it was unusual for a suspended load in a

308
00:09:53,985 --> 00:09:55,824
long-term application.

309
00:09:55,824 --> 00:09:58,148
Investigators used Fourier Transform

310
00:09:58,148 --> 00:10:00,228
Infrared Spectroscopy Headspace Gas

311
00:10:00,228 --> 00:10:02,548
Chromatography and Mass Spectroscopy of

312
00:10:02,548 --> 00:10:03,749
epoxy samples

313
00:10:03,749 --> 00:10:05,427
from the majority of the anchors that

314
00:10:05,427 --> 00:10:07,750
failed in the incident, as well as other

315
00:10:07,750 --> 00:10:09,670
randomly selected anchors to determine

316
00:10:09,670 --> 00:10:11,429
their chemical composition.

317
00:10:11,429 --> 00:10:13,189
In all cases it was found to be

318
00:10:13,189 --> 00:10:14,631
consistent with a fast

319
00:10:14,631 --> 00:10:17,352
set epoxy. Installation of the anchors in

320
00:10:17,352 --> 00:10:17,753
the D

321
00:10:17,753 --> 00:10:20,873
Street portal began in July 1999 using

322
00:10:20,873 --> 00:10:23,594
epoxy purchased from Newman Renner Colony.

323
00:10:23,594 --> 00:10:25,432
The investigations confirmed that the

324
00:10:25,432 --> 00:10:26,874
invoices indicate

325
00:10:26,874 --> 00:10:28,953
that Modern Continental purchased Power-

326
00:10:28,953 --> 00:10:31,596
fast, Fast Set NRC1000

327
00:10:31,596 --> 00:10:34,074
Gold Epoxy during the period when the D

328
00:10:34,074 --> 00:10:35,357
Street portal ceiling was being

329
00:10:35,357 --> 00:10:36,714
installed.

330
00:10:36,714 --> 00:10:38,556
Further testing of this epoxy showed

331
00:10:38,556 --> 00:10:40,234
that anchors installed using best

332
00:10:40,234 --> 00:10:41,196
practices,

333
00:10:41,196 --> 00:10:43,437
and I quote, "...exhibited significant and

334
00:10:43,437 --> 00:10:45,836
continued displacement (aka creep) when

335
00:10:45,836 --> 00:10:47,679
subjected to loads as low as one

336
00:10:47,679 --> 00:10:49,039
thousand pounds.

337
00:10:49,039 --> 00:10:50,959
Anchors loaded to four thousand pounds

338
00:10:50,959 --> 00:10:52,481
completely separated from their anchor

339
00:10:52,481 --> 00:10:54,881
holes before the end of the 82-day test

340
00:10:54,881 --> 00:10:55,440
period..."

341
00:10:55,440 --> 00:10:58,402
end quote. The portal ceiling segment

342
00:10:58,402 --> 00:11:00,239
anchors had an expected maximum

343
00:11:00,239 --> 00:11:02,802
load of 2,600 pounds and based on the

344
00:11:02,802 --> 00:11:03,682
above tests,

345
00:11:03,682 --> 00:11:05,604
due to its susceptibility to creep was

346
00:11:05,604 --> 00:11:07,924
not a suitable epoxy for use in any

347
00:11:07,924 --> 00:11:10,241
long-term tension load application.

348
00:11:10,241 --> 00:11:12,565
The above tests however suggest that the

349
00:11:12,565 --> 00:11:14,723
ceiling should have failed sooner than

350
00:11:14,723 --> 00:11:15,366
it did.

351
00:11:15,366 --> 00:11:17,607
The likely reason it did not fail sooner

352
00:11:17,607 --> 00:11:19,446
is that the ventilation system was not

353
00:11:19,446 --> 00:11:20,966
in operation for some time

354
00:11:20,966 --> 00:11:23,206
after the physical installation. Until

355
00:11:23,206 --> 00:11:24,808
the ventilation system was operating

356
00:11:24,808 --> 00:11:26,490
the ceiling tiles would have presented

357
00:11:26,490 --> 00:11:28,650
their own independent gravitational load

358
00:11:28,650 --> 00:11:29,847
on the support anchors,

359
00:11:29,847 --> 00:11:32,891
their so-called dead weight or dead load.

360
00:11:32,891 --> 00:11:34,652
Once the ventilation system became

361
00:11:34,652 --> 00:11:36,248
operational the air pressure would

362
00:11:36,248 --> 00:11:38,011
increase and decrease cyclically

363
00:11:38,011 --> 00:11:39,692
due to changes in flow direction and

364
00:11:39,692 --> 00:11:41,451
pressure and would also have induced

365
00:11:41,451 --> 00:11:43,452
vibration into the material which would

366
00:11:43,452 --> 00:11:44,092
have then

367
00:11:44,092 --> 00:11:46,415
accelerated the creep. The investigators

368
00:11:46,415 --> 00:11:48,495
also tested standard set epoxy

369
00:11:48,495 --> 00:11:50,251
under the same conditions and found that

370
00:11:50,251 --> 00:11:51,776
in the 82 day test

371
00:11:51,776 --> 00:11:54,656
under a 2,000 pound load, Standard Set

372
00:11:54,656 --> 00:11:55,536
did creep

373
00:11:55,536 --> 00:11:58,253
at about 1/10th of the rate of Fast

374
00:11:58,253 --> 00:11:58,658
Set,

375
00:11:58,658 --> 00:12:02,017
and it did so in a flat linear increment.

376
00:12:02,017 --> 00:12:04,659
The Fast Set during that time exhibited

377
00:12:04,659 --> 00:12:05,619
a non-linear,

378
00:12:05,619 --> 00:12:08,018
with increasing rate of creep, after just

379
00:12:08,018 --> 00:12:10,421
55 days.

380
00:12:10,421 --> 00:12:14,181
So let's talk about the epoxy. Epoxies

381
00:12:14,181 --> 00:12:16,340
are basic components that when combined

382
00:12:16,340 --> 00:12:18,902
and cured create an epoxy resin and are

383
00:12:18,902 --> 00:12:20,022
sometimes called

384
00:12:20,022 --> 00:12:23,461
Polyepoxides hence epoxy for short.

385
00:12:23,461 --> 00:12:26,424
Epoxy resin is combined with a hardener

386
00:12:26,424 --> 00:12:26,904
also

387
00:12:26,904 --> 00:12:28,583
sometimes called a curative or a

388
00:12:28,583 --> 00:12:30,824
co-reactant and when they are mixed the

389
00:12:30,824 --> 00:12:32,906
amine groups on the hardener molecules

390
00:12:32,906 --> 00:12:35,065
begin to link with the epoxide groups on

391
00:12:35,065 --> 00:12:37,066
the resin molecules releasing heat in

392
00:12:37,066 --> 00:12:38,264
the process.

393
00:12:38,264 --> 00:12:40,345
The reaction is therefore exothermic and

394
00:12:40,345 --> 00:12:42,264
creates a thermosetting polymer

395
00:12:42,264 --> 00:12:44,509
and unlike thermoplastic it resists

396
00:12:44,509 --> 00:12:46,347
reforming under applied heat and

397
00:12:46,347 --> 00:12:47,549
chemicals.

398
00:12:47,549 --> 00:12:49,468
Most resin molecules have an epoxide

399
00:12:49,468 --> 00:12:50,988
ring at each end and the harder

400
00:12:50,988 --> 00:12:52,747
molecules have an amine group

401
00:12:52,747 --> 00:12:55,071
at each end hence they both link in an

402
00:12:55,071 --> 00:12:57,071
alternating pattern to form long polymer

403
00:12:57,071 --> 00:12:57,552
chains

404
00:12:57,552 --> 00:13:00,193
and cross-link between polymer chains.

405
00:13:00,193 --> 00:13:01,952
The more cross-links that exist between

406
00:13:01,952 --> 00:13:03,393
the polymer chains the more

407
00:13:03,393 --> 00:13:06,112
resin resists significant rearrangement

408
00:13:06,112 --> 00:13:07,555
of the polymer molecules

409
00:13:07,555 --> 00:13:09,315
so the polymer cannot be melted and

410
00:13:09,315 --> 00:13:11,316
reformed. This type of polymer is called

411
00:13:11,316 --> 00:13:12,916
a Thermoset.

412
00:13:12,916 --> 00:13:14,994
As the crosslink polymer network forms

413
00:13:14,994 --> 00:13:17,153
the viscosity of the material increases

414
00:13:17,153 --> 00:13:19,076
and eventually the reaction slows and

415
00:13:19,076 --> 00:13:20,274
finally stops,

416
00:13:20,274 --> 00:13:23,077
although any unbonded reactants may not

417
00:13:23,077 --> 00:13:24,918
have been free to migrate through the

418
00:13:24,918 --> 00:13:25,799
solution

419
00:13:25,799 --> 00:13:28,039
if not well enough mixed beforehand

420
00:13:28,039 --> 00:13:29,079
leading to some

421
00:13:29,079 --> 00:13:32,118
unbonded reactive groups. Hence the

422
00:13:32,118 --> 00:13:33,720
faster the setting of the epoxy,

423
00:13:33,720 --> 00:13:35,720
the more likely uncured sections will

424
00:13:35,720 --> 00:13:37,081
exist and there will be

425
00:13:37,081 --> 00:13:40,361
less cross-linked polymers. So

426
00:13:40,361 --> 00:13:43,241
what is creep? Creep refers to a gradual

427
00:13:43,241 --> 00:13:44,843
continuing deformation

428
00:13:44,843 --> 00:13:47,242
of a material under a sustained load.

429
00:13:47,242 --> 00:13:48,844
It's similar to dislocations we

430
00:13:48,844 --> 00:13:51,084
discussed on Episode 35 about the San

431
00:13:51,084 --> 00:13:53,326
Bruno Gas Pipeline explosion.

432
00:13:53,326 --> 00:13:55,807
As a polymer, the stiffness of an epoxy

433
00:13:55,807 --> 00:13:57,807
is time and temperature dependent, hence

434
00:13:57,807 --> 00:13:59,807
if the load is applied suddenly the

435
00:13:59,807 --> 00:14:01,648
epoxy responds like a hard

436
00:14:01,648 --> 00:14:05,007
glassy solid. If a load is static, meaning

437
00:14:05,007 --> 00:14:07,088
it's held constantly against the plastic,

438
00:14:07,088 --> 00:14:09,008
the long chain polymer molecules have

439
00:14:09,008 --> 00:14:11,008
time to rearrange and slide past one

440
00:14:11,008 --> 00:14:12,850
another and the stiffness of the epoxy

441
00:14:12,850 --> 00:14:14,531
decreases to a range where it can be

442
00:14:14,531 --> 00:14:15,408
described as

443
00:14:15,408 --> 00:14:17,971
"rubbery." I know. Technical terms like

444
00:14:17,971 --> 00:14:18,851
"rubbery."

445
00:14:18,851 --> 00:14:20,612
Anyway. Polymer materials with an

446
00:14:20,612 --> 00:14:22,693
instantaneous elastic response followed

447
00:14:22,693 --> 00:14:24,933
by a slowly increasing deformation

448
00:14:24,933 --> 00:14:27,974
are referred to as Viscoelastic. The time

449
00:14:27,974 --> 00:14:29,493
needed for the glassy to rubbery

450
00:14:29,493 --> 00:14:30,291
transition

451
00:14:30,291 --> 00:14:32,215
depends on the molecular structure and

452
00:14:32,215 --> 00:14:33,494
the lengths of the polymer chains

453
00:14:33,494 --> 00:14:34,055
between

454
00:14:34,055 --> 00:14:36,856
crosslinks and also as the temperature

455
00:14:36,856 --> 00:14:37,657
increases

456
00:14:37,657 --> 00:14:39,495
the rate of rearrangement increases

457
00:14:39,495 --> 00:14:41,736
allowing the glassy-to-rubbery transition

458
00:14:41,736 --> 00:14:44,458
to occur more rapidly. In a well-formed

459
00:14:44,458 --> 00:14:46,699
cross-linked network of a Thermoset,

460
00:14:46,699 --> 00:14:48,700
the polymer molecules are prevented from

461
00:14:48,700 --> 00:14:49,739
moving very far,

462
00:14:49,739 --> 00:14:51,739
and once the rubbery state is reached, no

463
00:14:51,739 --> 00:14:53,740
further softening occurs.

464
00:14:53,740 --> 00:14:56,139
The design for the anchor resin provided

465
00:14:56,139 --> 00:14:58,380
by Gannett Fleming to Modern Continental

466
00:14:58,380 --> 00:15:00,621
specified the following, and I quote:

467
00:15:00,621 --> 00:15:01,742
"...Provide adhesive

468
00:15:01,742 --> 00:15:03,824
consisting of two component plastic

469
00:15:03,824 --> 00:15:06,300
resin and catalyst hardener mixture,

470
00:15:06,300 --> 00:15:08,383
resin material shall remain unaffected

471
00:15:08,383 --> 00:15:10,545
by continuous humidity and by chemicals

472
00:15:10,545 --> 00:15:12,225
present in a vehicle exhaust type

473
00:15:12,225 --> 00:15:15,264
of air duct environment..." (end quote) The

474
00:15:15,264 --> 00:15:17,183
specification goes on to include minimum

475
00:15:17,183 --> 00:15:18,947
design service loads, minimum factors of

476
00:15:18,947 --> 00:15:19,505
safety,

477
00:15:19,505 --> 00:15:21,183
and installation requirements that had

478
00:15:21,183 --> 00:15:23,425
to be met. The epoxy selected

479
00:15:23,425 --> 00:15:25,507
by Modern Continental used a two-part

480
00:15:25,507 --> 00:15:26,628
epoxy material,

481
00:15:26,628 --> 00:15:29,185
resin and hardener formulated by Sika

482
00:15:29,185 --> 00:15:31,186
Corporation of Lyndhurst, New Jersey

483
00:15:31,186 --> 00:15:33,028
packaged by Powers Fasteners

484
00:15:33,028 --> 00:15:34,711
Incorporated (formerly the Rawlplug

485
00:15:34,711 --> 00:15:35,908
Company Incorporated)

486
00:15:35,908 --> 00:15:38,390
of New Rochelle, New York and distributed

487
00:15:38,390 --> 00:15:39,270
by Newman Renner

488
00:15:39,270 --> 00:15:43,189
Colony LLC of Westwood, Massachusetts.

489
00:15:43,189 --> 00:15:45,353
Sika Corporation supplied the epoxy

490
00:15:45,353 --> 00:15:47,674
resin and hardener in bulk to Powers,

491
00:15:47,674 --> 00:15:50,074
which packaged, marketed and distributed

492
00:15:50,074 --> 00:15:51,994
it for Newman Renner Colony as

493
00:15:51,994 --> 00:15:55,756
NRC1000 a gold epoxy, which was used by

494
00:15:55,756 --> 00:15:57,996
Modern Continental on this project.

495
00:15:57,996 --> 00:16:00,076
In order to begin installation the

496
00:16:00,076 --> 00:16:01,678
construction contractor needed their

497
00:16:01,678 --> 00:16:03,276
final design to be approved by the

498
00:16:03,276 --> 00:16:04,238
consultants

499
00:16:04,238 --> 00:16:06,157
and Modern Continental submitted a total

500
00:16:06,157 --> 00:16:08,559
of four individual submissions

501
00:16:08,559 --> 00:16:11,278
for their anchor design. The first three

502
00:16:11,278 --> 00:16:13,361
anchor design submissions were rejected

503
00:16:13,361 --> 00:16:15,197
by Gannett Fleming, with Gannet Fleming

504
00:16:15,197 --> 00:16:17,040
requesting additional information

505
00:16:17,040 --> 00:16:20,562
each time. On the 30th of December, 1999

506
00:16:20,562 --> 00:16:22,563
Modern Continental submitted the fourth

507
00:16:22,563 --> 00:16:24,402
anchor adequacy submittal.

508
00:16:24,402 --> 00:16:27,042
The anchor service load data in that

509
00:16:27,042 --> 00:16:28,004
submittal

510
00:16:28,004 --> 00:16:30,163
were the values calculated and specified

511
00:16:30,163 --> 00:16:32,245
by Gannett Fleming originally in their

512
00:16:32,245 --> 00:16:33,523
initial design.

513
00:16:33,523 --> 00:16:35,366
The anchor load capacity data was

514
00:16:35,366 --> 00:16:37,443
extracted from the powers design manual

515
00:16:37,443 --> 00:16:38,567
(second edition)

516
00:16:38,567 --> 00:16:40,323
and powers indicated the adhesive

517
00:16:40,323 --> 00:16:42,007
anchors were capable of supporting up to

518
00:16:42,007 --> 00:16:42,968
6,350

519
00:16:42,968 --> 00:16:45,204
pounds each in

520
00:16:45,204 --> 00:16:46,888
4,000 psi concrete,

521
00:16:46,888 --> 00:16:48,570
while maintaining a safety factor of

522
00:16:48,570 --> 00:16:50,489
4 against the ultimate load which in

523
00:16:50,489 --> 00:16:51,370
this case was

524
00:16:51,370 --> 00:16:54,731
25,400 pounds.

525
00:16:54,731 --> 00:16:57,049
None of the four submissions included

526
00:16:57,049 --> 00:16:59,530
any specific reference to exactly

527
00:16:59,530 --> 00:17:02,493
which epoxy formulation was going to be

528
00:17:02,493 --> 00:17:02,973
used

529
00:17:02,973 --> 00:17:05,212
for the anchor. The fourth submission

530
00:17:05,212 --> 00:17:07,611
included a copy of a draft revision of

531
00:17:07,611 --> 00:17:12,091
ICBO ER-4514 dated October 1999

532
00:17:12,091 --> 00:17:14,573
in response to a specific request by the

533
00:17:14,573 --> 00:17:16,813
Gannett Fleming engineer based on his

534
00:17:16,813 --> 00:17:18,737
review of a previous submission.

535
00:17:18,737 --> 00:17:21,216
ICBO is the International Conference of

536
00:17:21,216 --> 00:17:22,334
Building Officials,

537
00:17:22,334 --> 00:17:26,499
and ER-4514 was the ICBO's independent

538
00:17:26,499 --> 00:17:27,938
evaluation report

539
00:17:27,938 --> 00:17:30,179
for the Power-Fast adhesive anchor

540
00:17:30,179 --> 00:17:31,857
system from Powers.

541
00:17:31,857 --> 00:17:33,858
The draft report revision limited the

542
00:17:33,858 --> 00:17:36,338
use of the Power-Fast fast set epoxy to

543
00:17:36,338 --> 00:17:38,662
short-term loads such as those resulting

544
00:17:38,662 --> 00:17:40,982
from wind or earthquake forces.

545
00:17:40,982 --> 00:17:44,503
On the 17th of December, 1999 the anchor

546
00:17:44,503 --> 00:17:46,581
capacity structural calculations were

547
00:17:46,581 --> 00:17:47,381
certified

548
00:17:47,381 --> 00:17:49,382
by a registered professional engineer

549
00:17:49,382 --> 00:17:50,905
employed by Sigma Engineering

550
00:17:50,905 --> 00:17:52,745
International Incorporated of Lincoln,

551
00:17:52,745 --> 00:17:53,623
Rhode Island,

552
00:17:53,623 --> 00:17:56,426
noting that, and i quote: "...the calculations

553
00:17:56,426 --> 00:17:58,427
were performed to compare the anchor

554
00:17:58,427 --> 00:18:00,427
minimum design service loads per

555
00:18:00,427 --> 00:18:02,748
project specification with the allowable

556
00:18:02,748 --> 00:18:04,586
loads provided by the anchor bolt

557
00:18:04,586 --> 00:18:06,749
manufacturer only..."

558
00:18:06,749 --> 00:18:09,386
(end quote) In other words no accounting

559
00:18:09,386 --> 00:18:10,029
for creep

560
00:18:10,029 --> 00:18:12,347
and long-term use in this application

561
00:18:12,347 --> 00:18:13,789
was made.

562
00:18:13,789 --> 00:18:16,431
On the 7th of January, 2000 Gannett

563
00:18:16,431 --> 00:18:18,348
Fleming authorized the contract to

564
00:18:18,348 --> 00:18:20,349
proceed with the anchor installation per

565
00:18:20,349 --> 00:18:21,390
their design

566
00:18:21,390 --> 00:18:23,472
and this was later secondarily approved

567
00:18:23,472 --> 00:18:24,993
by B/PB

568
00:18:24,993 --> 00:18:27,793
in February 2000. Having said that, due

569
00:18:27,793 --> 00:18:29,953
to schedule pressures the contractor had

570
00:18:29,953 --> 00:18:31,632
already been installing the anchors

571
00:18:31,632 --> 00:18:32,995
prior to formal approval

572
00:18:32,995 --> 00:18:35,955
since July, 1999 as previously mentioned

573
00:18:35,955 --> 00:18:36,276
some

574
00:18:36,276 --> 00:18:38,594
5 months before that formal approval

575
00:18:38,594 --> 00:18:39,633
came through.

576
00:18:39,633 --> 00:18:42,437
But hang on a minute! They found problems

577
00:18:42,437 --> 00:18:45,158
in 1999.

578
00:18:45,158 --> 00:18:47,958
On the 7th of October, 1999 Modern

579
00:18:47,958 --> 00:18:50,117
Continental's project manager contacted

580
00:18:50,117 --> 00:18:50,837
B/PB

581
00:18:50,837 --> 00:18:53,160
via a formal letter that they had become

582
00:18:53,160 --> 00:18:54,920
aware of problems involving

583
00:18:54,920 --> 00:18:57,399
"...a small percentage of adhesive anchors

584
00:18:57,399 --> 00:18:59,482
in the HOV ceiling mock-up."

585
00:18:59,482 --> 00:19:01,080
The ceiling module had been built in

586
00:19:01,080 --> 00:19:03,321
August that year and by October the

587
00:19:03,321 --> 00:19:05,163
hanger plates had displaced from the

588
00:19:05,163 --> 00:19:05,802
roof

589
00:19:05,802 --> 00:19:08,361
by about 1/2 an inch or 12mm

590
00:19:08,361 --> 00:19:09,882
prompting their letter.

591
00:19:09,882 --> 00:19:12,765
B/PB responded on the 12th of October

592
00:19:12,765 --> 00:19:14,685
suggesting that the turnbuckles in the

593
00:19:14,685 --> 00:19:15,564
ceiling module

594
00:19:15,564 --> 00:19:18,205
may have been causing excessive load on

595
00:19:18,205 --> 00:19:19,404
some supports,

596
00:19:19,404 --> 00:19:21,884
leading to creep. They also suggested

597
00:19:21,884 --> 00:19:23,728
incorrect fitting of the bolts and the

598
00:19:23,728 --> 00:19:24,528
epoxy

599
00:19:24,528 --> 00:19:26,929
by the constructor as another potential

600
00:19:26,929 --> 00:19:27,968
cause.

601
00:19:27,968 --> 00:19:29,886
Several meetings with Powers and Modern

602
00:19:29,886 --> 00:19:31,969
Continental ensued and were followed by

603
00:19:31,969 --> 00:19:33,328
a letter from Powers to Modern

604
00:19:33,328 --> 00:19:36,451
Continental on the 29th of October, 1999

605
00:19:36,451 --> 00:19:38,931
reminding of the correct installation

606
00:19:38,931 --> 00:19:40,051
procedures but

607
00:19:40,051 --> 00:19:42,211
ultimately stating that since the design

608
00:19:42,211 --> 00:19:43,493
of the ceiling plate

609
00:19:43,493 --> 00:19:45,572
module and anchors were not done by

610
00:19:45,572 --> 00:19:47,494
Powers they could not address,

611
00:19:47,494 --> 00:19:49,734
and I quote: "...the different loads that may

612
00:19:49,734 --> 00:19:51,412
be acting on these anchors..."

613
00:19:51,412 --> 00:19:54,935
(end quote). On the 8th of November, 1999

614
00:19:54,935 --> 00:19:57,496
Modern Continental wrote back to B/PB

615
00:19:57,496 --> 00:19:59,413
attaching Power's formal response and

616
00:19:59,413 --> 00:20:00,537
concluded amongst

617
00:20:00,537 --> 00:20:03,178
other comments with, and I quote: "...it is

618
00:20:03,178 --> 00:20:04,855
improbable that the anchors were

619
00:20:04,855 --> 00:20:06,299
overstressed as a result

620
00:20:06,299 --> 00:20:08,299
of erection loads or the erection

621
00:20:08,299 --> 00:20:09,740
procedures..." (end quote)

622
00:20:09,740 --> 00:20:12,459
So let's time out for a second. It's a "he

623
00:20:12,459 --> 00:20:13,578
said," "she said,"

624
00:20:13,578 --> 00:20:16,219
kind of situation. Prove that these

625
00:20:16,219 --> 00:20:18,378
anchors were all installed correctly

626
00:20:18,378 --> 00:20:19,820
despite the fact that some of the

627
00:20:19,820 --> 00:20:21,580
anchors were installed in the presence

628
00:20:21,580 --> 00:20:22,461
of a B/PB

629
00:20:22,461 --> 00:20:24,943
engineer, unless every single one was

630
00:20:24,943 --> 00:20:26,704
that's not enough proof that the anchors

631
00:20:26,704 --> 00:20:28,304
weren't installed correctly.

632
00:20:28,304 --> 00:20:31,103
During retesting several failed but

633
00:20:31,103 --> 00:20:32,464
evidence was found where

634
00:20:32,464 --> 00:20:34,706
incorrect hole cleaning was applied and

635
00:20:34,706 --> 00:20:36,383
insufficient epoxy was used

636
00:20:36,383 --> 00:20:38,786
leading B/PB to issue Deficiency Report

637
00:20:38,786 --> 00:20:39,506
001,

638
00:20:39,506 --> 00:20:41,585
and after many letters and emails

639
00:20:41,585 --> 00:20:43,187
back and forth with allegations of

640
00:20:43,187 --> 00:20:44,948
incorrect installation procedures being

641
00:20:44,948 --> 00:20:46,625
followed and in an email

642
00:20:46,625 --> 00:20:48,789
from the B/PB design manager to the

643
00:20:48,789 --> 00:20:51,269
project engineer and structural engineer

644
00:20:51,269 --> 00:20:53,589
internally to B/PB stating the following

645
00:20:53,589 --> 00:20:54,627
and I quote:

646
00:20:54,627 --> 00:20:55,990
"...we are not trying to hold up

647
00:20:55,990 --> 00:20:57,908
construction, we are trying to make a

648
00:20:57,908 --> 00:20:58,869
determination

649
00:20:58,869 --> 00:21:01,192
that the installation is safe and

650
00:21:01,192 --> 00:21:01,991
functional..."

651
00:21:01,991 --> 00:21:05,033
(end quote). Sensing frustration? It gets

652
00:21:05,033 --> 00:21:05,832
worse.

653
00:21:05,832 --> 00:21:07,671
The tension between the two companies

654
00:21:07,671 --> 00:21:09,514
escalated to the suggestion

655
00:21:09,514 --> 00:21:11,912
of completely pulling and reinstalling

656
00:21:11,912 --> 00:21:13,354
every single anchor

657
00:21:13,354 --> 00:21:15,353
in the HOV ramp area with no clear

658
00:21:15,353 --> 00:21:17,276
agreement of who should pay

659
00:21:17,276 --> 00:21:20,717
until the overarching CAT Project agreed

660
00:21:20,717 --> 00:21:23,434
in February, 2000 to pay to retest all

661
00:21:23,434 --> 00:21:25,278
anchors to a higher proof test

662
00:21:25,278 --> 00:21:27,676
and split the cost of any that required

663
00:21:27,676 --> 00:21:29,117
replacement.

664
00:21:29,117 --> 00:21:33,040
To that end: 187 anchors were tested.

665
00:21:33,040 --> 00:21:37,201
19 of them failed! Deficiency Report 001

666
00:21:37,201 --> 00:21:39,762
was considered to be closed and it was

667
00:21:39,762 --> 00:21:42,962
closed on the 26th of January, 2001.

668
00:21:42,962 --> 00:21:46,482
On the 17th of December, 2001

669
00:21:46,482 --> 00:21:48,804
a Modern Continental QA inspector filed

670
00:21:48,804 --> 00:21:50,243
a non-compliance report

671
00:21:50,243 --> 00:21:52,642
to B/PB after finding multiple anchor

672
00:21:52,642 --> 00:21:54,642
displacements in the I90 connector

673
00:21:54,642 --> 00:21:55,525
tunnel.

674
00:21:55,525 --> 00:21:58,006
The action directed by bpb was the same

675
00:21:58,006 --> 00:21:59,443
as it had been undertaken

676
00:21:59,443 --> 00:22:02,086
in the HOV tunnel, 2 years previously.

677
00:22:02,086 --> 00:22:03,288
These were executed

678
00:22:03,288 --> 00:22:06,247
and the non-compliance was closed out.

679
00:22:06,247 --> 00:22:07,848
During the investigation

680
00:22:07,848 --> 00:22:10,570
two other incidents of anchor slippage

681
00:22:10,570 --> 00:22:11,686
were found.

682
00:22:11,686 --> 00:22:14,570
In several B/PB field engineering reports

683
00:22:14,570 --> 00:22:16,571
in 2001 and once again

684
00:22:16,571 --> 00:22:20,169
in 2002, however neither were formally

685
00:22:20,169 --> 00:22:20,812
flagged

686
00:22:20,812 --> 00:22:23,370
as non-compliances or defects and no

687
00:22:23,370 --> 00:22:25,611
deficiency reports were created.

688
00:22:25,611 --> 00:22:27,852
Despite the discovery of these damaged

689
00:22:27,852 --> 00:22:28,974
anchor bolts,

690
00:22:28,974 --> 00:22:31,214
over a four year period, project

691
00:22:31,214 --> 00:22:33,215
officials did not begin

692
00:22:33,215 --> 00:22:36,413
any inspection program. So

693
00:22:36,413 --> 00:22:39,135
what are the key findings? The NTSB

694
00:22:39,135 --> 00:22:40,977
investigation delivered its report on

695
00:22:40,977 --> 00:22:43,538
the 10th of July, 2007 on the anniversary

696
00:22:43,538 --> 00:22:44,338
of the incident

697
00:22:44,338 --> 00:22:46,415
and there were 20 findings in their

698
00:22:46,415 --> 00:22:48,819
report. I'm going to focus on 4

699
00:22:48,819 --> 00:22:52,018
that I think are key. Finding #5:

700
00:22:52,018 --> 00:22:53,539
Gannet Fleming and Bechtel

701
00:22:53,539 --> 00:22:55,221
Parsons Bricknerhoff failed to account

702
00:22:55,221 --> 00:22:56,821
for the fact that polymer adhesives are

703
00:22:56,821 --> 00:22:58,418
susceptible to deformation

704
00:22:58,418 --> 00:23:00,501
(creep) under sustained load with the

705
00:23:00,501 --> 00:23:02,582
result that they made no provision for

706
00:23:02,582 --> 00:23:04,342
ensuring the long-term safe performance

707
00:23:04,342 --> 00:23:07,141
of the ceiling support anchoring system.

708
00:23:07,141 --> 00:23:10,023
Finding #10: Gannet Fleming approved the D

709
00:23:10,023 --> 00:23:11,305
Street portal anchors without

710
00:23:11,305 --> 00:23:13,304
identifying which epoxy formulation was

711
00:23:13,304 --> 00:23:14,903
being used.

712
00:23:14,903 --> 00:23:17,865
Finding #13: After unexplained anchor

713
00:23:17,865 --> 00:23:19,786
displacement was found in the interstate

714
00:23:19,786 --> 00:23:23,464
90 connector tunnel in 1999 and 2001,

715
00:23:23,464 --> 00:23:25,228
Bechtel Parsons Bricknerhoff and Modern

716
00:23:25,228 --> 00:23:26,905
Continental Construction Company

717
00:23:26,905 --> 00:23:28,749
Incorporated should have instituted a

718
00:23:28,749 --> 00:23:30,589
program to monitor anchor performance to

719
00:23:30,589 --> 00:23:32,186
ensure that the actions taken in

720
00:23:32,186 --> 00:23:33,707
response to the displacement

721
00:23:33,707 --> 00:23:36,667
were effective. Finding #15: Had the

722
00:23:36,667 --> 00:23:38,510
Massachusetts Turnpike Authority

723
00:23:38,510 --> 00:23:40,188
at regular intervals between November

724
00:23:40,188 --> 00:23:42,670
2003 and July 2006

725
00:23:42,670 --> 00:23:44,592
inspected the area above the suspended

726
00:23:44,592 --> 00:23:46,753
ceilings in the D Street portal tunnels

727
00:23:46,753 --> 00:23:48,353
the anchor creep that led to this

728
00:23:48,353 --> 00:23:50,431
accident would likely have been detected

729
00:23:50,431 --> 00:23:52,113
and action could have been taken that

730
00:23:52,113 --> 00:23:54,595
would have prevented this accident.

731
00:23:54,595 --> 00:23:56,756
The report also noted that there was no

732
00:23:56,756 --> 00:23:57,796
evidence found

733
00:23:57,796 --> 00:23:59,316
that Modern Continental had any

734
00:23:59,316 --> 00:24:01,395
information at that time to suggest that

735
00:24:01,395 --> 00:24:03,317
the epoxy it was using was susceptible

736
00:24:03,317 --> 00:24:05,636
to creep. The National Transportation and

737
00:24:05,636 --> 00:24:07,078
Safety Board (or NTSB)

738
00:24:07,078 --> 00:24:09,319
determined that and I quote: "...the probable

739
00:24:09,319 --> 00:24:10,916
cause of the July 10,

740
00:24:10,916 --> 00:24:13,157
2006 ceiling collapse in the D Street

741
00:24:13,157 --> 00:24:14,677
portal of the I90 connector

742
00:24:14,677 --> 00:24:16,361
tunnel in Boston Massachusetts

743
00:24:16,361 --> 00:24:18,520
was the use of an epoxy anchor adhesive

744
00:24:18,520 --> 00:24:20,041
with poor creep resistance.

745
00:24:20,041 --> 00:24:21,959
That is an epoxy formulation that was

746
00:24:21,959 --> 00:24:23,720
not capable of sustaining long-term

747
00:24:23,720 --> 00:24:24,363
loads.

748
00:24:24,363 --> 00:24:26,122
Over time the epoxy deformed and

749
00:24:26,122 --> 00:24:27,960
fractured until several ceiling support

750
00:24:27,960 --> 00:24:29,323
anchors pulled free and allowed a

751
00:24:29,323 --> 00:24:31,402
portion of the ceiling to collapse."

752
00:24:31,402 --> 00:24:34,124
(end quote). Let's talk about the fallout.

753
00:24:34,124 --> 00:24:36,923
In late 2008 the family of Melina Del

754
00:24:36,923 --> 00:24:37,325
Valle

755
00:24:37,325 --> 00:24:40,366
reached a $28.09M USD

756
00:24:40,366 --> 00:24:42,607
settlement with 15 parties named in

757
00:24:42,607 --> 00:24:44,526
their wrongful death lawsuit.

758
00:24:44,526 --> 00:24:46,368
Bechtel Parsons Bricknerhoff and Modern

759
00:24:46,368 --> 00:24:48,205
Continental paid the largest share of

760
00:24:48,205 --> 00:24:49,008
the settlement

761
00:24:49,008 --> 00:24:50,849
with Powers Fasteners and Gannett

762
00:24:50,849 --> 00:24:52,609
Fleming Incorporated the next largest

763
00:24:52,609 --> 00:24:53,648
following that.

764
00:24:53,648 --> 00:24:55,727
The supplier of the epoxy compound,

765
00:24:55,727 --> 00:24:57,488
Powers Fasteners, was charged

766
00:24:57,488 --> 00:24:59,011
with one count of involuntary

767
00:24:59,011 --> 00:25:00,688
manslaughter. Action

768
00:25:00,688 --> 00:25:02,291
against Powers Fasteners Incorporated

769
00:25:02,291 --> 00:25:03,812
was the first criminal indictment

770
00:25:03,812 --> 00:25:06,690
following the July, 2006 incident.

771
00:25:06,690 --> 00:25:08,613
The suspended ceiling structure in the D

772
00:25:08,613 --> 00:25:11,014
Street portal was removed.

773
00:25:11,014 --> 00:25:13,492
During subsequent tunnel evaluations, it

774
00:25:13,492 --> 00:25:14,534
was determined that

775
00:25:14,534 --> 00:25:16,212
because of the short length of that

776
00:25:16,212 --> 00:25:18,535
tunnel section and its proximity to a

777
00:25:18,535 --> 00:25:19,414
tunnel opening,

778
00:25:19,414 --> 00:25:21,577
the suspended ceiling was not necessary

779
00:25:21,577 --> 00:25:23,656
for adequate tunnel ventilation.

780
00:25:23,656 --> 00:25:27,098
The ceiling structure was never replaced.

781
00:25:27,098 --> 00:25:29,338
So what do we learn from all of this?

782
00:25:29,338 --> 00:25:31,177
Certainly don't start installing things

783
00:25:31,177 --> 00:25:32,779
until the design is approved.

784
00:25:32,779 --> 00:25:35,020
That's always a good one, but it's clear

785
00:25:35,020 --> 00:25:36,938
that neither of the reviewing companies

786
00:25:36,938 --> 00:25:38,938
looked beyond the static load strengths

787
00:25:38,938 --> 00:25:40,459
which showed that people reviewing

788
00:25:40,459 --> 00:25:41,900
within those companies didn't fully

789
00:25:41,900 --> 00:25:42,382
understand

790
00:25:42,382 --> 00:25:45,499
the nature of epoxies. Firstly though

791
00:25:45,499 --> 00:25:48,701
inspections. People don't seem to

792
00:25:48,701 --> 00:25:50,701
see that inspections are a form of

793
00:25:50,701 --> 00:25:52,461
protection; a layer of protection if you

794
00:25:52,461 --> 00:25:53,183
will.

795
00:25:53,183 --> 00:25:54,785
Beyond the basic quality assurance

796
00:25:54,785 --> 00:25:56,462
aspects during a project (which were

797
00:25:56,462 --> 00:25:57,346
lacking),

798
00:25:57,346 --> 00:25:59,106
the operational company or organization

799
00:25:59,106 --> 00:26:00,306
should be performing appropriate

800
00:26:00,306 --> 00:26:01,503
inspections regularly.

801
00:26:01,503 --> 00:26:03,106
Whether that's for concrete cracking,

802
00:26:03,106 --> 00:26:04,628
corrosion of metals, warping,

803
00:26:04,628 --> 00:26:07,588
bulging or in this case epoxy creeping,

804
00:26:07,588 --> 00:26:09,187
regular maintenance inspections are a

805
00:26:09,187 --> 00:26:11,186
catch-all for many types of premature

806
00:26:11,186 --> 00:26:12,308
failures.

807
00:26:12,308 --> 00:26:14,227
No tunnel inspections were performed to

808
00:26:14,227 --> 00:26:15,591
determine the physical and functional

809
00:26:15,591 --> 00:26:17,189
condition of the ceiling system

810
00:26:17,189 --> 00:26:18,871
from the time the I90 Eastbound

811
00:26:18,871 --> 00:26:20,550
connected tunnel was open to traffic on

812
00:26:20,550 --> 00:26:22,791
the 18th of January, 2003

813
00:26:22,791 --> 00:26:25,353
until the day of the incident. Had the

814
00:26:25,353 --> 00:26:25,990
MTA

815
00:26:25,990 --> 00:26:28,312
been inspecting post handover they would

816
00:26:28,312 --> 00:26:29,672
have seen the epoxy

817
00:26:29,672 --> 00:26:32,874
creep. No question. No doubt. Fundamentally

818
00:26:32,874 --> 00:26:33,355
though

819
00:26:33,355 --> 00:26:35,355
the evolution of the suspended ceiling

820
00:26:35,355 --> 00:26:36,953
design during the course of the project

821
00:26:36,953 --> 00:26:38,796
led to a poor choice and although

822
00:26:38,796 --> 00:26:40,637
it is possible to find adhesives that

823
00:26:40,637 --> 00:26:42,555
will hold that amount of load

824
00:26:42,555 --> 00:26:44,714
over a long period of time the design

825
00:26:44,714 --> 00:26:47,037
choice was probably still not the best.

826
00:26:47,037 --> 00:26:48,715
Had the original portal ceiling slab

827
00:26:48,715 --> 00:26:50,716
design incorporated anchor points,

828
00:26:50,716 --> 00:26:52,476
embedded in the concrete, it wouldn't

829
00:26:52,476 --> 00:26:54,236
have been an issue,

830
00:26:54,236 --> 00:26:56,079
but the way the design was split across

831
00:26:56,079 --> 00:26:57,439
different companies over a long period

832
00:26:57,439 --> 00:26:58,160
of time

833
00:26:58,160 --> 00:27:00,641
allowed this kind of oversight or gap to

834
00:27:00,641 --> 00:27:01,281
creep in

835
00:27:01,281 --> 00:27:04,081
which then necessitated a sub-optimal

836
00:27:04,081 --> 00:27:04,720
solution.

837
00:27:04,720 --> 00:27:07,201
The NTSB report also states and I quote:

838
00:27:07,201 --> 00:27:09,283
"In civil projects adhesive anchors are

839
00:27:09,283 --> 00:27:11,123
typically used in short-term or sheer

840
00:27:11,123 --> 00:27:12,323
load applications.

841
00:27:12,323 --> 00:27:14,001
Under these conditions even if the

842
00:27:14,001 --> 00:27:15,605
adhesive is susceptible to creep the

843
00:27:15,605 --> 00:27:17,365
displacement will likely never reveal

844
00:27:17,365 --> 00:27:17,846
itself

845
00:27:17,846 --> 00:27:19,606
and those responsible for specifying

846
00:27:19,606 --> 00:27:21,445
approving installing and testing the

847
00:27:21,445 --> 00:27:23,606
anchors will not be aware of it."

848
00:27:23,606 --> 00:27:25,288
If you're specifying something to be

849
00:27:25,288 --> 00:27:27,608
used, you really need to understand it

850
00:27:27,608 --> 00:27:29,206
since clearly most of the people

851
00:27:29,206 --> 00:27:31,049
involved in this project didn't

852
00:27:31,049 --> 00:27:33,208
really understand epoxy creep they

853
00:27:33,208 --> 00:27:34,409
shouldn't have been specifying

854
00:27:34,409 --> 00:27:36,169
it; and when you're a registered

855
00:27:36,169 --> 00:27:37,851
professional engineer and approving a

856
00:27:37,851 --> 00:27:38,890
document

857
00:27:38,890 --> 00:27:41,930
it's doing half the job to say "it meets

858
00:27:41,930 --> 00:27:43,132
the load requirements"

859
00:27:43,132 --> 00:27:44,893
but to not consider the application it

860
00:27:44,893 --> 00:27:46,249
was used for?

861
00:27:46,249 --> 00:27:48,970
That's important too. Ultimately there

862
00:27:48,970 --> 00:27:50,251
was enough blame

863
00:27:50,251 --> 00:27:52,011
to share around in this incident. There

864
00:27:52,011 --> 00:27:53,693
were multiple engineers that could have

865
00:27:53,693 --> 00:27:55,453
challenged the results of the HOV ramp

866
00:27:55,453 --> 00:27:56,415
test anchors.

867
00:27:56,415 --> 00:27:58,335
The engineers that cited problems in

868
00:27:58,335 --> 00:28:00,175
2001 and 2002

869
00:28:00,175 --> 00:28:01,695
should have raised those as formal

870
00:28:01,695 --> 00:28:02,973
defects to the project to be

871
00:28:02,973 --> 00:28:04,576
investigated and rectified.

872
00:28:04,576 --> 00:28:06,255
The MTA should have been conducting

873
00:28:06,255 --> 00:28:08,177
regular inspections too.

874
00:28:08,177 --> 00:28:09,776
Ultimately though the design, review,

875
00:28:09,776 --> 00:28:11,457
construct and assurance process on this

876
00:28:11,457 --> 00:28:12,016
project

877
00:28:12,016 --> 00:28:14,578
failed. The design was not specific

878
00:28:14,578 --> 00:28:16,977
enough on a crucial point. The review of

879
00:28:16,977 --> 00:28:17,620
the design

880
00:28:17,620 --> 00:28:20,018
was not detailed enough on that crucial

881
00:28:20,018 --> 00:28:21,061
point.

882
00:28:21,061 --> 00:28:22,979
When the warning signs came to light, the

883
00:28:22,979 --> 00:28:24,818
construction and assurance process

884
00:28:24,818 --> 00:28:27,623
incorrectly blamed the installation, when

885
00:28:27,623 --> 00:28:28,582
the true cause

886
00:28:28,582 --> 00:28:30,981
was not properly determined. Were the

887
00:28:30,981 --> 00:28:31,702
wrong people

888
00:28:31,702 --> 00:28:35,304
involved in these stages of the project?

889
00:28:35,304 --> 00:28:37,704
Most likely? Were the right people

890
00:28:37,704 --> 00:28:38,426
ignored

891
00:28:38,426 --> 00:28:40,503
due to cost and time pressure?

892
00:28:40,503 --> 00:28:42,186
Potentially.

893
00:28:42,186 --> 00:28:44,504
Could a simple visual inspection have

894
00:28:44,504 --> 00:28:46,186
found the problem and prevented this

895
00:28:46,186 --> 00:28:46,907
incident

896
00:28:46,907 --> 00:28:50,188
from occurring? Most definitely. Those

897
00:28:50,188 --> 00:28:51,948
engineering companies were held to some

898
00:28:51,948 --> 00:28:54,266
account but what about

899
00:28:54,266 --> 00:28:57,310
the MTA? My observation is that

900
00:28:57,310 --> 00:28:59,631
operations companies like the MTA can

901
00:28:59,631 --> 00:29:01,631
look at inspections as an operational

902
00:29:01,631 --> 00:29:02,268
expense

903
00:29:02,268 --> 00:29:04,591
and when nothing obvious is going wrong

904
00:29:04,591 --> 00:29:06,352
over a long time period

905
00:29:06,352 --> 00:29:08,592
they tend to become the lowest priority

906
00:29:08,592 --> 00:29:09,791
task and

907
00:29:09,791 --> 00:29:12,434
get pushed off repeatedly. Another

908
00:29:12,434 --> 00:29:13,634
suggestion I've heard

909
00:29:13,634 --> 00:29:15,635
is that because it's brand new, why would

910
00:29:15,635 --> 00:29:17,315
you need to inspect it?

911
00:29:17,315 --> 00:29:20,513
But inspections are critical! They are

912
00:29:20,513 --> 00:29:22,753
the layer of protection against the

913
00:29:22,753 --> 00:29:23,715
unforeseen

914
00:29:23,715 --> 00:29:25,554
and they aren't about what you expect to

915
00:29:25,554 --> 00:29:27,555
see...they're about finding what you don't

916
00:29:27,555 --> 00:29:28,597
expect,

917
00:29:28,597 --> 00:29:32,036
and acting on it. Had the MTA developed

918
00:29:32,036 --> 00:29:34,276
and performed those inspections as part

919
00:29:34,276 --> 00:29:36,277
of and following the project handover

920
00:29:36,277 --> 00:29:37,959
this could have been avoided.

921
00:29:37,959 --> 00:29:40,360
I'm not excusing the design errors at

922
00:29:40,360 --> 00:29:40,998
all

923
00:29:40,998 --> 00:29:43,401
but I am saying that a relatively

924
00:29:43,401 --> 00:29:45,082
inexpensive walk down

925
00:29:45,082 --> 00:29:47,083
would have detected this well before

926
00:29:47,083 --> 00:29:49,481
anything ever happened.

927
00:29:49,481 --> 00:29:51,482
So the next time you're thinking you can

928
00:29:51,482 --> 00:29:53,324
put off that overdue inspection

929
00:29:53,324 --> 00:29:56,925
a few more days...make the time...

930
00:29:56,925 --> 00:30:00,523
and just do it! If you're enjoying

931
00:30:00,523 --> 00:30:02,043
Causality and want to support the show

932
00:30:02,043 --> 00:30:02,763
you can

933
00:30:02,763 --> 00:30:04,605
by becoming a Patron! You can find

934
00:30:04,605 --> 00:30:06,447
details at engineered.network/

935
00:30:06,447 --> 00:30:08,525
causality about how you can help make

936
00:30:08,525 --> 00:30:10,367
this show continue to be made.

937
00:30:10,367 --> 00:30:12,766
You can find details at engineered.network/

938
00:30:12,766 --> 00:30:14,845
causality about how you can help this

939
00:30:14,845 --> 00:30:16,527
show to continue to be made.

940
00:30:16,527 --> 00:30:18,610
A big thank you to all of our Patrons

941
00:30:18,610 --> 00:30:20,130
and a special thank you to our Silver

942
00:30:20,130 --> 00:30:20,768
Producers

943
00:30:20,768 --> 00:30:23,009
Mitch Bielger, John Whitlow, Kevin Koch,

944
00:30:23,009 --> 00:30:24,452
Oliver Steele, Hafthor

945
00:30:24,452 --> 00:30:26,529
and Shane O'Neill with an extra special

946
00:30:26,529 --> 00:30:28,290
thank you to our Gold Producer known

947
00:30:28,290 --> 00:30:28,849
only as

948
00:30:28,849 --> 00:30:31,488
'r'. Causality is heavily researched and

949
00:30:31,488 --> 00:30:33,052
links to all materials used for the

950
00:30:33,052 --> 00:30:34,449
creation of this episode

951
00:30:34,449 --> 00:30:36,260
are contained in the show notes. You can

952
00:30:36,260 --> 00:30:37,663
find them in the text of the episode

953
00:30:37,663 --> 00:30:39,722
description of your podcast player or

954
00:30:39,722 --> 00:30:41,942
on our website. You can follow me on the

955
00:30:41,942 --> 00:30:42,850
fediverse at

956
00:30:42,850 --> 00:30:45,814
chidgey@engineered.space on Twitter

957
00:30:45,814 --> 00:30:48,818
@johnchidgey (all one word), or the network

958
00:30:48,818 --> 00:30:50,589
@engineered_net.

959
00:30:50,589 --> 00:30:53,884
This was Causality. I'm John Chidgey.

960
00:30:53,884 --> 00:30:56,967
Once again, thanks so much for listening...

961
00:31:04,975 --> 00:31:17,458
[Music]