Gas Main Explosion in Massachusetts- Update 5

[Kestrel42]

Since the prior thread has been closed, I thought some might be interested to learn what the outcome was for the gas company involved in the disaster. Record 53 million dollar fine and the company, Columbia Gas, is to be sold. For some reason, residents just don’t feel safe leaving Columbia Gas in business in their neighborhoods, lol.

https://cumberlink.com/news/nationa...cle_3a729cf0-3183-5f8f-9050-4445b7c9cbed.html

 

[CWB1]

The trouble with requiring a PE is that the license does nothing to guarantee experience in niche areas such as gas distribution systems. One could argue that a PE is actually counterproductive as discussed here many times, many PEs believe it is their right on the basis of their PE to perform work with which they have zero experience and also complete work with no peer review. Having seen a small city in which I lived previously seriously consider employing a 27-year old PE as city engineer with final oversight on everything from dams to road construction to gas infrastructure, that mentality is a timebomb waiting to happen. I suspect that is why there has been no mention of culpability among the local licensed officialdom with ultimate engineering oversight - they likely had no experience and are thus unqualified for the position they're in.

If the goal is to hire competence, the only standard worth considering is past success meeting strict regulations and peer review. The fact that one or both was omitted in a heavily populated area is worrisome.

 

[waross]

I have had experience with a couple of those young PEs.
These were installations done well and to code, but failed because the PE didn't understand the code and had zero field experience.
Both times an old "up through the ranks" inspector was able to sort it out and save the PE the embarrassment of losing an appealed ruling.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[1503-44]

Well, that certainly is possible. The client would have done well to hire a PE with experience in their field. That is the usual procedure that was always in force whenever our clients selected our company to do such work. We were always selected from proposals made by a group of several companies with experience in the type of project of interest which included a rigorous iterview involving several in office and at site meetigs. Usually a (non-bid) process that was designed solely to select the engineering compamy based on the best documented experiemce in the particular type of installations of interest. After all, you don't go to a proctologist for brain surgury, right. Well unless of course you've got your head stuck in the wrong place.

 

[FacEngrPE]

I built a few monuments myself - before I became a PE. It is all part of learning how to be an engineer. Working under an experienced engineer helped considerably. It also helped that I apprenticed in a machine shop before I finished my schooling.

Fred

 

[1503-44]

School: All variables are given. They never change. You just have to solve the equations.
Work: No variables are given. All are subject to change. You just know what equations you have to solve.

 

[FacEngrPE]

Good point at work sometimes the equations are also unknown (or are too complex to be useful) and usable equations need to be figured out (or the actual expert found that has solved a similar problem).

When one of these complex jobs is finished and all of the task requirements are met it can be really satisfying.

 

[davidbeach]

School is equations. Real world is empirical.

I presented a paper at a conference 11+ years ago where I described a process where I solved a set of equations by trial and error. One equation and hundreds, if not thousands, of unknowns (zero sequence impedance of multiple lines with many sections and lots of variable conditions). I solved it by assuming a single value over the whole distance for a couple of necessary values and comparing those against an actual event, iterating until the results were reasonable. A couple of different reactions.

One was from my advisor (working on my Masters at the time) who brought a group of students to the conference and he said that the students were all aghast at the empirical approach. The other was the founder of a major supplier who declared that "this is a significant paper" in the field. Experience since then show that the empirical approach rules the roost.

All of that isn't to say that there aren't a whole lot of equations that are useful day in and day out, but all of those equations start out assuming a set of ideal conditions. If the ideal condition are a good match to reality then the standard equations are used and found useful. But when the ideal conditions don't do a good job of describing the actual conditions the solution becomes more and more empirical. Knowing which is which is when/where the grey hairs are earned.

 

[1503-44]

Which is why neural networks are making progress, essentially emulating that approach. Show them enough of the mash available and the right results and they start figuring out how to paint the big picture.