Enbridge's Kalamazoo Disaster: Straight from the Files

This is the first in a series of articles culled from documents compiled by the U.S. National Transportation Safety Board (NTSB) during its investigation into the rupture of Enbridge Pipeline 6b near Marshall, Michigan, two years ago, which fouled 38 miles of the Kalamazoo River and surrounding areas, causing residents to be evacuated for their health and safety.

The NTSB posted 376 documents totaling over 14,000 pages of reports, interviews, telephone transcripts, etc. This evidence formed the basis of the NTSB's final report, released to the public on July 26, 2012. The Tyee has replaced some persons' names with more general descriptors.

1. Enbridge vice president Richard Adams: 'Our response time... can be almost instantaneous'

Ten days before the Marshall spill, on July 15, 2010, Enbridge's VP of U.S. operations Richard Adams testified:

"Our response time from our control center can be almost instantaneous, and our large leaks are typically detected by our control center personnel."

-- Congressional Hearing before the Committee on Transportation and Infrastructure, Sept. 15, 2010, p. 76.

Line 6b ruptured at 5:58 pm EDT on Sunday, July 25, 2010. During the next 17 hours, personnel at Enbridge’s Control Center (ECC) in Edmonton, Alberta, some 1,500 miles away, failed to recognize that they had a spill. Control center staff restarted the line twice during the overnight shift, pumping about 683,000 gallons of diluted bitumen, in addition to the dilbit which was spilled during the initial rupture. The spill was finally noticed by a gas employee in Marshall who phoned ECC's emergency line at 11:17 the following morning.

2. ECC operators: 'It'll be fun'

More than 10 hours after the rupture, this is part of a telephone conversation between two Enbridge operators at the Edmonton control center. They begin speaking at 4:24 a.m. EDT on Monday, July 26, 2010, a few minutes after the first restart of Line 6b:

CONTROL CENTER: Howdy.

UNIDENTIFIED SPEAKER: Hey, man. Pretty slow going at startup, hey?

CONTROL CENTER: Yes, very slow. That's why I'm just wondering either they really drained it out, which I think they did, because I don't have any pressure farther down the line.

UNIDENTIFIED SPEAKER: Yeah, they must have because I'm still trying to hold 150 and it's just sitting there.

CONTROL CENTER: Yeah, exactly. Or else I'm -- or else I'm leaking. One of the two.

UNIDENTIFIED SPEAKER: Well, either way it'll be fun.

-- from Enbridge Edmonton Control Center Phone Transcripts, Vol. 2, pgs. 596-597

3. Sarnia operator: 'Are you actually serious?'

One ECC operator expressed concern that they may have a leak on their hands, as opposed to the presumed column separation.

The NTSB final report explains:

About 7:09 a.m., operator B1 notified the Sarnia Terminal operator that they were going to start Line 6B for a second time. The Sarnia Terminal operator expressed disbelief at the idea of a second startup. He told investigators that he had voiced his concerns about a Line 6B leak to shift leads B1 and B2 and MBS analyst B that morning. He stated that MBS analyst B had dismissed his concerns and, because he was dealing with other issues that morning, he had not pursued the matter.

Line 6B was started a second time about 7:20 a.m. [and shut down about 20 minutes later].

-- from NTSB Final Report, p. 13.

When interviewed by a NTSB investigator, the Sarnia operator affirmed what was said. [Names of individuals and line indicators have been removed by The Tyee.]

NTSB: ...there is a point in the transcripts that we have, I think, just before the second start-up -- Sarnia operator: Yeah.

NTSB: -- [Line 6b operator] calls you up and asks you, I believe, to open the valves, and I think your response to [him] is something along the lines of, yeah, okay, are you actually serious? I think that's -- those were the words you used.

Sarnia operator: Yeah. By that point in the night, I had, again, voiced what I seen to the only people I could voice to, my shift leads, and [the MBS analyst] when he was trying to decide on what to do for some reason, and it was met with a lot of, you know, we're handling it, you know, go back to your console. You know, thanks for your input, but, you know. So, by the end of the night I was like they can decide what they would like to do...

-- from transcript of interview with Sarnia operator, Nov. 2011, p. 25-26.

4. ECC operators: 'Let's not worry about it anymore'

This is part of a telephone conversation between two ECC employees. Investigators subsequently determined the identities of the two speakers, both of them console operators at ECC. This conversation began at 7:54 a.m. EDT on Monday, July 26, 2010 (a few minutes after the second startup and 14 hours after the rupture):

CONTROL CENTER: This is great, eh?

UNIDENTIFIED SPEAKER: Yeah. Well, I've never seen this problem. That's kind of interesting, to be honest.

CONTROL CENTER: Yeah, this is nice. I like this.

UNIDENTIFIED SPEAKER: Have you ever done this?

CONTROL CENTER: No not like this.

UNIDENTIFIED SPEAKER: Well, neither have I. And to me like it looks like a leak.

CONTROL CENTER: For sure.

UNIDENTIFIED SPEAKER: And I'm like holy cow that's amazing. Like I've never ever heard of that where you can't get enough --

CONTROL CENTER: I can pump as hard as I want and I -- I'd never over pressure the line?

UNIDENTIFIED SPEAKER: Yeah. But doesn't it seem messed up? Like eventually the oil has to go somewhere.

CONTROL CENTER: It has to.

UNIDENTIFIED SPEAKER: Like, I don't know.

CONTROL CENTER: (indiscernible)

UNIDENTIFIED SPEAKER: I don't know. Something about this feels wrong.

CONTROL CENTER: Yup.

UNIDENTIFIED SPEAKER: Whatever. We're going home. We're off for a few days.

CONTROL CENTER: Exactly.

UNIDENTIFIED SPEAKER: Let's not worry about it anymore.

CONTROL CENTER: I'm done. Exactly. We're not going to try this again. Not on our shift.

UNIDENTIFIED SPEAKER: No.

-- from Enbridge Edmonton Control Center Phone Transcripts, Vol. 2, pgs. 662-663.

5. The 10 Minute Rule

One of the NTSB documents is titled, "Source of 10-minute rules." It examines a Minnesota spill in March, 1991 which was remarkably similar to the Marshall spill. This spill is notable not only for its similarities but for three additional reasons:

1. The 1991 spill led to the creation of Enbridge's "10 minute rule" -- an internal procedure to prevent additional oil from being pumped during a suspected leak.

2. The 10 minute rule was not adhered to during the Marshall spill, resulting in the pumping of large volumes of oil for over an hour (as in the 1991 spill).

3. Enbridge's regional manager during the 2010 Marshall spill was Lakehead's manager of operations services during the late 1990s. As the telephone transcript further down indicates, this manager, despite his awareness of the historical reasons for the 10-minute rule, gives permission for a third restart of Line 6b.

NTSB summarized the 1991 spill and the origins of Enbridge's "10 minute rule":

The 10-minute limitation was adopted as a result of the March 1991 Enbridge rupture and release that occurred on Line 3, spilling 1.7 million gallons of crude oil in Grand Rapids, Minnesota... During the 1991 accident, personnel in Enbridge's Edmonton Control Center interpreted the SCADA alarms and indications to a condition of column separation and instrument error and continued to pump oil into the ruptured 34-inch-diameter line for more than an hour until the leak was recognized.

In 1991, Enbridge stated in its response to PHMSA that a revision to the operation maintenance procedures manual was adopted stating, "If an operator experiences pressure or flow abnormalities or unexplainable changes in line conditions for which a reason cannot be established within a 10- minute period, the line shall be shut down, isolated, and evaluated until the situation is verified and or [sic] corrected."

-- NTSB Final Report, p. 52

6. 'Our aggressive, proactive and risk-based integrity management program'

In 1998, Lakehead's manager of operations services wrote to the Office of Pipeline Safety, pointing out his company's improvements and the importance of an "awareness of... history." From that letter:

Lakehead's actions subsequent to the Consent Order Agreement have significantly exceeded the requirements and directives of the Order. Our aggressive, proactive and risk-based integrity management program has addressed longitudinal seam cracks in USS SAW pipe, corrosion, operator training and pressure cycling. Lakehead has made a commitment to continue this focussed approach into the future...

Lakehead's commitment to personnel training will ensure that awareness of Line 3 issues including history and evolution of the integrity program are maintained.

Lakehead... recommends that the Consent Order be officially closed.

Sincerely,

[Manager, Operations Services]

-- Lakehead's letter to Rick Gulstad (OPS), Dec. 2, 1998 [in NTSB docket].

The author of this 1998 letter went on to become the Chicago region manager with Enbridge. Twelve years later, this same individual apparently forgot his "awareness of Line 3 issues including history" -- ie. why the 10-minute rule is necessary -- as the next document details.

7. The 10 Minute Rule, ignored

What follows is a part of the transcript of a telephone conversation between Enbridge's Chicago region manager and a shift supervisor at Enbridge's control center in Edmonton. In the NTSB transcript, the Chicago region manager is identified by his first name. As a courtesy, his name is here replaced by "CRM" (Chicago region manager).

This conversation began at 10:16 a.m. EDT on Monday, July 26, 2010 (16 hours after the rupture):

CONTROL CENTER: So normally when things [i.e. pressures] go to zero --

CRM: Yeah, you check for leaking.

CONTROL CENTER: -- suction end discharge [should read, "suction and discharge"] you'd be checking for leaks for sure.

CRM: Yeah.

CONTROL CENTER: And normally if it goes to zero it's usually something that happens right at the station.

CRM: Right.

CONTROL CENTER: All three transmitters went to zero at exactly the same time.

CRM: Right.

CONTROL CENTER: But I guess at this point we're kind of at a loss. We're looking at more numbers here right now.

CRM: Okay.

CONTROL CENTER: But initially I don't know if you guys need to check out some of the pipeline upstream and downstream of Marshall?

CRM: I wouldn't think so. I -- you know, if it's right at Marshall -- you know, it seems like there's something else going wrong either with the computer or with, with the instrumentation. And then your lost column and things go haywire, right?

CONTROL CENTER: Yeah. But, I guess --

CRM: Yeah, do you want them to check?

CONTROL CENTER: Well --

CRM: I'm not -- right now I'm not, I'm not convinced. We haven't had any phone calls. I mean it's perfect weather out here.

CONTROL CENTER: Yeah, (indiscernible).

CRM: Someone -- if it's a rupture someone's going to notice that, you know, and smell it.

CONTROL CENTER: Yeah. Yeah, for sure.

CRM: So --

CONTROL CENTER: Yeah, I guess -- okay. At this point we'll just keep looking into things here.

CRM: Yeah.

Later in the same conversation:

CONTROL CENTER: But, but yeah, our thinking was that it should have filled up downstream because we weren't taking much out. But yeah, we'll do some more digging. I guess --

CRM: Yeah, have a look. If you have --

CONTROL CENTER: So if we can't, if we can't make sense of the numbers then, yeah, we may have to give you guys a call back to --

CRM: Yeah, call us back, but --

CONTROL CENTER: -- check (indiscernible).

CRM: -- I'm okay with you guys ready to go if it looks like the numbers are fitting. [CRM appears to be giving permission for a third re-start of Line 6b.]

CONTROL CENTER: Yeah. Is [Enbridge Marshall employee] an electrician or is he a --

CRM: Yep. Yep.

CONTROL CENTER: He is, eh?

CRM: Yeah.

CONTROL CENTER: Is there any way he can check the transmitters to see that --

CRM: Sure.

CONTROL CENTER: -- the --

CRM: You want to, you want to give him call and just

CONTROL CENTER: (indiscernible) the PLC or the transmitters or --

CRM: Yep. Yeah, just give him a call and --

CONTROL CENTER: -- (indiscernible) something.

CRM: Yeah. You guys call him directly so that I'm not playing middleman and --

CONTROL CENTER: Okay. Yeah, no for sure.

CRM: -- and just get him to check things out. Tell him that we're just -- your numbers aren't jiving and things aren't --we want to double check before we fire up [i.e. to restart Line 6b for a third time].

CONTROL CENTER: Yeah. Yeah, awesome. All right, [CRM].

CRM: You got, you got my okay to go, but give us a call if you want us to definitely check.

CONTROL CENTER: Yeah.

CRM: But we would have, we would have heard something by now.

CONTROL CENTER: Okay. No, it sounds good. So that whole Marshall area upstream and downstream is pretty populated then, correct?

CRM: Yeah.

CONTROL CENTER: Is that right?

CRM: Yeah. Well, I wouldn't say populated, but I mean there's farms --

CONTROL CENTER: Yeah.

CRM: -- and there's houses and people driving around all the time, yeah.

CONTROL CENTER: All over -- all the time, yeah.

CRM: Yes.

CONTROL CENTER: Yeah. Okay. No, it sounds good then, [CRM].

CRM: All righty?

CONTROL CENTER: We might give you call -- if we do decide to start up again we might give you a call anyways --

CRM: Yep.

CONTROL CENTER: -- just to double check, but --

CRM: No problem. We'll help you out. I'm sitting in here my office.

CONTROL CENTER: Awesome. Thanks, [CRM].

-- from Enbridge Edmonton Control Center Phone Transcripts, Vol. 2, pgs. 725-726 and 727-729.

8. Enbridge's no news is good news approach to leak detection

The NTSB made the following observations regarding Enbridge's interpretation of the absence of external notifications:

The control center staff, to some extent, and the Chicago regional manager believed that unintended product releases would be reported by outside sources (that is, either affected citizens or community officials)...

The visual confirmation of the leak did not occur until 11:17 a.m. on July 26. In the absence of that confirmation from a person located in Marshall, control center personnel discounted the possibility of a leak, largely because no external confirmation of a leak was present. Thus, the absence of information on a leak led to the belief that there was no leak, and that some other phenomenon, yet unrecognized, was causing the column separation.

Moreover, there was no evidence that any member of the control center staff sought to obtain information from anyone in the Marshall vicinity to verify the presence of a leak. Rather than actively soliciting information from sources in the Marshall area, the control center staff continued their erroneous decision-making by misinterpreting the absence of notifications from the Marshall community as actual information that there was no leak...

Therefore, the NTSB concludes that Enbridge control center staff misinterpreted the absence of external notifications as evidence that Line 6B had not ruptured.

-- from NTSB Final Report, p. 100

9. Treachur: 'There's a ****pile of it. That creek is black'

This is part of the emergency call which finally alerted the Control Center to the fact that Enbridge had a large spill. This call began at 11:17 a.m. EDT on Monday, July 26, 2010 (more than 17 hours after the initial spill):

CONTROL CENTER: Enbridge pipeline emergency phone. Go ahead. What's your pipeline emergency?

MR. TREACHUR: Yes. This is Chris Treachur. I work for Consumers Energy and I'm in Marshall. There's oil getting into the creek and I believe it's from your pipeline.

CONTROL CENTER: Okay.

MR. TREACHUR: I mean there's a lot. We're getting, we're getting like 20 gas leak calls and everything. So I found -- do you know where the address was of that creek oil? I'm trying to remember that, but anyway it's between 27 and 16 mile. We're trying to walk your line and see if we can find where it's broke.

CONTROL CENTER: Okay.

MR. TREACHUR: But it's -- I mean, there's, there's a ___pile of it. That creek is black.

CONTROL CENTER: Okay, so --

MR. TREACHUR: It's running.

CONTROL CENTER: Yeah. Your name is Chris Treachur?

-- from Enbridge Edmonton Control Center Phone Transcripts Enbridge Edmonton Control Center Phone Transcripts, Vol. 2, pgs. 771- 772.

Later this week, the next installment of this series will examine how circumstances in Marshall inhibited the ability of local authorities to correctly diagnose a major oil spill for 17 hours.