Unforced Errors

This post sort of follows on from this one, and describes much the same problem.

A year or two into my assignment doing weight estimates, we had a big re-organisation which meant I was dealing mainly with offshore facilities and more closely involved with cost estimations (rather than purely weight estimations). One of the principle ways the cost of a facility was estimated is to take various parameters – total liquid processing capacity, oil production rate, gas processing rate, etc. – and use that to work out the topsides weight. This is what they did, and as far as I know they still do.

One day we invited an American chap to visit us from a company which specialises in the design and operation of certain installations. We wanted his feedback on previous work we’d done with him, and his advice for future projects. He was very open, and I found the meeting fascinating. He highlighted the various technical requirements unique to our company which made our installations more expensive than they ought to be, with other clients happy to accept less stringent requirements or use industry standards. He went into detail on this, and in several instances it was the case that technology had moved on and our standards hadn’t yet caught up. For example, if you want to send an intelligent pig down a line you’d have to put 5D bends in (i.e. the bend radius is 5 times the pipe diameter), but nowadays the pigs can generally handle 3D bends. Our standards still required 5D bends, which take up a lot more space in a crowded facility. That was just one example of several, which as an engineer I found very interesting.

Not so my colleagues. After the meeting I raised these points as possible areas in which we could save costs, and the response was:

“Oh, that was all bullsh*t, he was just telling us that to try to get the next contract.”

Not for the first time has an expert in a particular technical field been invited into an oil company to share knowledge and been treated like he’s the dumbest one in the building.

Anyway, one of the things the American chap said was his company had found no relationship between the liquid production rates and the facility topsides weight. There were just too many other variables which affect it, such as the degree to which you want to remove certain contaminants. He even said his company had teamed up with a university to research this relationship, but after a couple of years they’d given up. What this fellow said effectively consigned our entire estimation methodology to the dustbin, because it relied entirely on a perceived association between production rates and topsides weights. This either went straight over the heads of the assembled staff sat in front of him, or they chose to ignore it. Either way, nobody mentioned it again.

Just for fun, once I’d been taught statistical analysis techniques last semester I ran some figures to see whether the methodology we’d been applying back then was mathematically sound. It turned out there was a correlation between equipment weight and topsides weight, but it was a lot weaker than I’d expected. But more importantly, there was no association between production rates and equipment weight, or indeed between any of the parameters we used and weights. So the American was right, then.

Now had I known these techniques when I still worked there, and demonstrated to those in charge of the methodology that we shouldn’t be assuming an association between X and Y when none exists, they’d have said:

“This is the methodology we are using. Your job is to follow it without asking questions.”

In fact, a short while before the reorganisation someone suggested I get involved in cost estimations and apparently one of the managers said:

“Oh, we don’t want him, he’ll just find things wrong with our methodology.”

Major corporations, people. Next time you hear about something like this or this, you’ll know how they happen.


8 thoughts on “Unforced Errors

  1. “Major corporations, people. Next time you hear about something like this or this, you’ll know how they happen.”

    Yep. And such goes double/triple for government contractors. Especially defense contractors. Not so much in that those types of disasters are common, but more so in the huge cost overruns on support programs like logistics and such. Though one other “this” that would be relevant here is the Challenger Space Shuttle disaster. Have you read Richard Feynman’s book on that?

  2. Have you read Richard Feynman’s book on that?

    I haven’t, but I suspect I should.

  3. The whole book isn’t about it, just the second half. Though the first half is amusingly interesting as well.


    What stuck with me was that during his work with the investigation committee, how he had to maneuver around management to speak to the actual engineers and how a certain general on the committee had to point him in a certain direction without explicitly telling him what he thought the problem was because it would be politically dicey for the general himself to raise the subject about the O-ring failures.

  4. We’ve always done it this way

    It’s worse than that. It’s a case of people being moved around from one position to another every few years regardless of expertise, who inherit a set of procedures they don’t understand. They see their job as to follow the process and do what management say regardless of context, but of course such people also have the tendency too meddle and “improve” things. After a few iterations like this, the whole process is completely corrupted but nobody cares provided it’s followed.

    At least the old boy who says “we’ve always done it this way” is basing his actions on something. The new style of managers just make everything up on the fly.

  5. In my case, whenever I request access to certain applications or functions within applications it always gets rejected as being contrary to SOX compliance. I’ve got to the point where I now request them to provide documentary evidence of this.

    Seems to me that most internal security and audit policies are merely efficiency prevention exercises.

  6. Tim,
    This may be a bit off message but it is interesting that the recent job loss announcements at Ford and JLR are in areas such as marketing and sales. I.e. things that can either be outsourced or may be cut without too much impact on revenues.
    Having gone through the big R in the 1990s at a bank, one can see the pattern of such activities. Hard (or possibly, not so hard) decisions are made about who should go. The only mitigation is politics. I worked for a US bank in London. The further away from HQ in NYC, the more likely the redundancies.
    Still, I suspect many organisations have, as you indicate, a lot of fat that–when times get tough–can be trimmed.
    According to the press reports it is lack of demand for diesel vehicles (I wonder why?) and a general downturn after many happy years of automobile sales seem the proximate reasons. Brexit, not so much.

  7. Not just the general anti-diesel policies, the Conservative (!!) government recently increased the taxes on new cars, especially diesels and up-markets cars.
    High taxes reduce demand!!! Who knew?

    The government should be proudly claiming these job losses as proof their tax increase policies are working. Another success!! /sarc.
    John Redwood has some pretty scathing things to say about this, and the massive stamp duty increases ‘helping’ the housing market. But he remains a member, so must accept the responsibility for his party’s policies.

Comments are closed.