Use of the word "conservative" 1

[geoman77]

I recently had a senior engineer get rather miffed that I had used the word "conservative" on some foundation calcs that I did. He essentially said, "Think about what that word means. You're not being conservative; just say what you really mean."

I think what he's suggesting is that we often use the word conservative to imply that we're assigning values that we know to be too large (or too small, depending on the situation), when in reality we simply don't have an accurate handle on the "real" value. In other words, we're not actually being conservative; we're being purposefully broad to cover our lack of knowledge. Or something like that...

Any thoughts?

 

[msquared48]

All generalizations are not worth a da*^, including this one.

What he states is not true all the time. Sometimes we can do an accurate solution, but due to the budget or to save the client money, we just "err" so to say on the "conservative" side. There is nothing wrong with that approach. What matters is knowing you are on the right side of the line.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[Ron]

Mike is right, as usual.

One other consideration is that you don't want to write the word on your calcs. That can have a liability implication. If you ever have a project go bad, your calcs and everything you have in the file will be meticulously mauled over to pick up on the nuances of why you did what you did. That word would stick out, since it really has no business on the page with the calcs. Your assumptions are what they are, right or wrong. Your "conservative" might be considered by someone more persuasive to be "non-conservative"....and they might be right. If so, the credibility of the rest of your calcs goes south.

 

[msucog]

yep, usually when a geotech says "conservative", they mean that there's no money to actually do the necessary testing so we err on the side of caution and reasonalbity considering that we have no data justifying trying to shave that fine line. as mentioned above, you must be careful actually saying "conservative" especially if you don't have sound data. without the data, you don't know what you are...other than having a gut feeling governed by a cheapo budget instead of sound engineering principles.

 

[BigH]

What you are really trying to say is that you are utilizing a biased parameter that errs on the side of a higher factor of safety or an upper limit on settlement. Soils are variable - from location to location at the same "level" or within a particular stratum. I do not agree that more testing and "fancier" calculations are the answer in all cases. There is a point where the cost of obtaining "better" parameters exceeds the benefit of such parameters. Not all designs/recommendations must be optimized. This putatively is what the LRFD design procedures (risk analyses) are all about; to "quantify" the risk and reliability of an analysis.

As Ron points out, it is perhaps best not to write terms that can be construed by others in a different manner - yet, it is what we all do, isn't it?

Is one being conservative when he/she takes an umbrella out on a dark cloudy day? Mmmm

 

[Mccoy]

In the Italian practice the word 'conservative' is used pretty often.

By that you express in qualitative terms what in more rigorous technical terms you would define 'upper bound', 'lower bound', upeer-or lower- fractile, first or third quartile, and so on.

I absolutely agree that we must have a verifiable idea of what is the mean value before saying that our value is conservative, that is erring on the safe side of an unknown amount.

In Eurocode 7 the adjective 'conservative' has been substituted by the adjective 'cautious' like in: ...The characteristic value is a cautious estimate of the value which governs the occurrence of a limit state..."

In litigation the big issue may be how cautious you should be in adopting a cautious value.

I'll stick to the statistical method of defining a cautious or conservative value since I believe it's the only one which is really defensible in court.

 

[Debaser]

In UK civils we seem to prefer the term 'robust'. Drop that in a couple of times and see how this engineer reacts.

 

[Patgeotech]

We do not use the word 'conservative' in our geotech reports because of how other professionals perceive the word.

We typically use 'recommended' or 'it is our opinion' or if insufficient testing was undertaken to derive the parameters then we start the sentence as 'Although a limited amount of testing was undertaken...' or something in those lines.

We Geotechs are always conservative, it is the nature of our field/industry. You just need to find another way of expressing your conservative approach.

 

[csd72]

I understand the concern but I would think that if anyone reading the calculations relies on the interpretation of one word to understand them then they probably have no business reading them!

 

[SixDegrees]

We recently put out a report (addressed to a legal firm none the less)that went through a complex geotechnical computation pointing out all the areas where we made "conservative" assumptions.

As engineers we try not to assume anything, but in reality there are many unknowns, and approximations. While we spent a lot of money/time/effort to reduce the unknowns it is not possible for us to eliminate them all. I think it's important to know what are the truths and what are assumptions in any analysis. Identifying them in the calc's may be okay, if you back it up and explain why it is conservative, or why the actual value is not available.

 

[CrazyHorse81]

Ok. So when I use the word conservative when discussing design, I use it in the context that the design will add more of a safety factor to it; but my boss uses that the terminology in reverse...the design is liberal, because you are adding to the design.

And no matter if you like to use conservative or liberal to show an added safety factor, it means liberal spending once it comes to construction.

 

[csd72]

Never heard the word liberal used in that context.

Looking at the dictionary, one meaning is: 'Tending to give freely', which is not really a property that I like to have in my retaining wall designs

 

[rconner]

I'm not going to comment on the "legalese" of all this, but would just add in general,

The smarter some Engineers think they are, the less safety factor (or conservatism?) they think they need.

[I guess we could substitute cheaper or weaker materials etc. in place of "safety factor" and still have the same basic meaning, and I am putting REAL heavy emphasis on the cogitative adjective employed!]

 

[rconner]

Oops meant ...cogitative "word" (used twice)!

 

[BigH]

assumptions - no matter how refined the determination of the soil property, there is still an uncertainty in it. Similarly, in using the soil property in calcuation, there is also uncertainty. Has anyone looked into how many "recent" - say since 2000 - papers have been written on simply the three basic bearing capacity factors?

@reconner - correctomundo! Someday, an engineer (geotechnical, most likely oriented with dams) is going to be so smart, it collapses as he ponders his prowress over a glass of Laurent-Perrier's Brut Millesime champagne!