Engineering Judgement vs. Codes 2

[MotorCity]

Here is a hypothetical situation which has generated much discussion among a few of my colleagues. When, if ever, should engineering judgement take precedence over legally adopted design codes or standards (i.e. building codes)? For example, if during the design process you encounter a situation where following a code would be detrimental to public safety or result in an uneconomical design are you still required to adhere to that portion of a particular code? As an engineer, could I over rule a building inspector's decision, who probably has no formal or technical education in engineering? If I am under a contract with the owner to provide design services, am I still obligated to follow the code and seal the drawings if I believe that the design is flawed due to following the code? I know that in most codes there is are statements that say something like "in lieu of a more detailed analysis based on engineering principles......." which provides a loophole, but this is not always the case.

 

[Ron]

Most model codes are written with the concept in mind that if followed by non-engineers, the result will be adequate to protect the health, safety and welfare of the public. As engineers, we know this to be a fallacy, as many of the code provisions require engineering interpretation.

Having said that, the code often becomes construed as the "minimum standard of care" and to deviate from it puts additional liability on the engineer. We all know that there are times when engineering judgment should supercede certain provisions. My suggestion would be to document both sides of the argument, so that if an issue is raised at some point in the future, then there is a clear path as to the decision process.

Also keep in mind that if the code is not followed and there is a failure, the logical assumption is that if the code HAD BEEN followed a failure would not have occurred. This too, is sometimes a fallacy.

 

[GregLocock]

I don't know about building to code, but with cars we /have/ to meet various legislative requirements, engineering judgement or not.

I have heard of cases where cars have been succesfully homologated, even though they did not meet all the written requirements, but not on anything safety or emissions related.

We also have internal standards which are substantially more restrictive than the legal ones, but we can design a car that breaks these standards, so long as we can demonstrate that the sub-standard performance still satisfies the customer's requirements, etc.



Cheers

Greg Locock

 

[stevenal]

Engineering codes of ethics hold public safety paramount. Economics, however, is not paramount.

 

[BJC]

A good many of the things in building codes have a dead person or two behind them. If you have a valid reason to change something in a code you can apply for a variance in most jurisdictions. You have to state your case and the reasons you want a variance. If it's logical you'll probably get it.
A lot of plans inspectors and building officials see things differently than engireers. Their experience and training is more focused on public safety than efficient or seemingly good design.
Everyone knows why doors of nightclubs open outward ( the Coconut Grove fire )but there are lots of other seamingly trivial things that can have an impact on public safety. Lots of people don't know what they are and the rules in place because of them.

 

[JAE]

MotorCity, I guess I don't see the dilema here. As an engineer, you have a duty to protect the life and safety of the public (per Ron above).

Building codes set minimum standards. If you see something in the code that to you feels TOO minimum, just increase your design accordingly. You can then show that you have met the minimum design required by the code, and you even went the extra mile to increase the design due to your judgement.

 

[EddyC]

Code is not god. Codes are created by human beings and human beings sometimes make mistakes. Part of the reason that our engineering education is structured the way it is, is to teach us to think. I meet too many engineers who blindly design to code without knowing the limitations and applicability of code. I suggest examining the code in detail, try to find out why it is the way it is, learn the limits of its applicability, find out the assumptions that are utilized. A good engineer questions everything and assumes nothing.

 

[GregLocock]

Gosh, when you were designing a steel structure did you measure the young's modulus of the steel? Did you also build the instrumentation used to emeasure the Young's modulus?

I think your final sentence is silly.









Cheers

Greg Locock

 

[JAE]

Greg - which post are you referring to?

 

[IRstuff]

He's probably referring to "A good engineer questions everything and assumes nothing."

TTFN

 

[HgTX]

Did the initial question refer to a case where the code didn't go far enough, or where a better structure would actually be in violation of the code? If it's the second, can you give a little more information?

Hg

 

[MotorCity]

Just to clarify a few points:

1. I agree 100% with the fallacies Ron described.
2. Public safety and protection is not an issue, as
everyone agrees that this should dictate the design
above all else.
3. I also agree with JAE in that providing the minimum
(or more) as required by code is the way to go when
possible.

The point I was referring to is a case when a safer structure, in the opinion of the engineer, would result if it were in violation of a particular code requirement. I think that its not a matter of if the code goes far enough, but rather if it goes too far in some cases. I do not have a particular situation in mind, it was just a question discussed among fellow engineers. I hope this makes the topic more clear. Any other thoughts?

 

[HgTX]

I'm just having trouble coming up with a hypothetical situation. Not that I probably haven't come across them, I just can't think of any at the moment.

Hg

 

[Cogs]

It seems to me that there are two directions to this thread, firstly if the hypothetical situation is purely that, hypothetical, then surely it is basically a philosophical debate on whether an engineer can justly disregard a section/s of code and for what reasons. If that is ever the case then the code that is in conflict with the engineers views could be in need of ammendment.

Secondly, if Motorcity (or anyone) has become aware of a situation (via hypothetical reasoning or otherwise) where there is a genuine deficiency in a code that could give rise to a safety issue then surely it then becomes a duty to bring this to the attention of whatever body is responsible for maintaining the code. Codes do evolve to reflect changing developments in materials / methods etc.

Afterall is not the hypothetical situation (if Motorcity had a specific design or structural problem in mind) a form of FMEA ? albeit informal in this case, and are not the codes (generally) the result of formal FMEA and real tragedies, written to prevent failures happening again ?.

 

[DRC1]

Codes especially design codes are devloped based on considerable research and experince. Provisions are often developed to address actual problems or failures. On the other hand, Natue is fixed, yet the code is constintly evolving, so obviously we are still figuring out exactly how buildings work. Look at the significant revisions to the seismic codes. Codes are not perfect. Further they are written in the abstract, a design guide for projects yet to be concieved. It is possible you have concived of a structure that requires an application of the code not anticipated by the writers that in your situation may not be safe. My advise would be to thoughly analyze the situation so you are convinced there is a problem. Then call the agency who wrote the code. They can probably put you in touch with some one on the committe that is responsible for the provision. Explain the problem to him and listen to the response. If you are still convinced,lay out your reasons carefully and clearly show the varrience on the plans. Then meet the building offical to explain it to him. It will be much easier for him to approve it if he has not rejected it first. My answer is that if you feel the code is wrong for your application, you should do what is prudent, but be sure you are right.

 

[HgTX]

That's what I was wondering--whether there's a situation at hand that might warrant changing the code in question.

I've gotten a number of minor code changes through various bodies by pointing out some discrepancy or other to people in the position to change things. It can be done. If what you have is a weird enough case, though, they would probably say that's not an intended application.

Most contracts, codes, specifications have an option for "the Engineer" to allow exceptions to the code, so your hypothetical situation could be handled under that provision. Just document the hell out of it--if applicable, show calcs both ways demonstrating that what you did is better, and also set down in words the overall logic of your reasoning.

I'm having a major "d'oh!" moment here--most of my job involves handling exceptions to codes, just on a more minor level than I'd thought you were talking about. And very often we'll let something go that's in violation of the spec but still structurally adequate, because the cure may be worse than the disease--for example, a small area reduced in thickness because a defect was ground out often isn't worth introducing a welded repair to the area. Dunno if that answers the original question.

Hg

 

[jsummerfield]

Over a decade ago a doctorate level chemical engineer from a different office identified a flaw in API RP 520 for sizing safety relief valves. His sizing requirements included partial differentials for the correct solution.

I supervised over 40 engineers at that time. I advised that we would follow API RP 520 which is a requirement of many of our contracts. I also recommended that the PHD type should present his concerns to API. As far as I am aware no change occured.

Still, the use of engineering judgement is also required. We often speak of codes and standards. As with many standards, the API standard mentioned is a recommended practice. It is common for standards to include verbage that permits exceptions. Document deviation well, distribute to those in authority and proceed with approval.

John

 

[JamesCG]

JOhn,

Was the API method more or less safe than that established by your "Chemical engineer doctor"?

Regards

James

 

[haggis]

Interesting post. Getting away from the structural aspects of building codes, I can give you a good example of deviation of code and applying for a variance. The handling of fuels, specifically the dispensing of gasoline, is strictly forbidden indoors. However, we at a car assembly plant dispense gasoline as the cars near the end of the assembly line to enable driving off and facilitate vehicle testing. Our case for dispensing gasoline indoors is logical and by showing that our equipment, methods and safety precautions exceed the code for the handling of fuels, we had no problem obtaining a variance. This of course is not a unique situation, as every vehicle assembly plant in the world does the same thing. I think the point is that had we not applied for a variance, regardless of our equipment, methods and safety precautions, we would have been in violation of code. Maybe the scenarios are not parallel but as codes are written by engineers and variances granted by engineers when the design and safety criteria exceeds minimums, the inspector's black and white interpretation of deviation/violation of the code means nothing.

 

[stevenal]

"Public safety and protection is not an issue.."
"The point I was referring to is a case when a safer structure, .., would result.."

Can't have it both ways. Either safety is an issue or it isn't. If it is an issue, it must come before code requirements.