Seismic Site-Specific Analysis Info, please 1

[david32]

Could someone who is familiar with the site-specific procedure for determining ground motion accelerations provide some details about the analysis?

The IBC says, "A site-specific study shall account for the regional seismicity and geology; the expected recurrence rates and maximum magnitudes of events on known faults and source zones; the location of the site with respect to these; near source effects if any; and the characteristics of subsurface site conditions."

This sounds very complex to me and I didn't think that it was done very frequently. But evidently, it is performed more than I thought. I've heard that geotechs are charging around $2000-$3000 for the analysis, so it must be less in depth than I originally envisioned it.

The USGS's acceleration contours that account for the regional seismicity and geology are developed from a complex process including a large database of earthquakes, many attenuation relationships, paleoseismic investigations, and other complicated methods. Can the geotech's $3000 analysis of the seismicity be more "accurate" than the USGS's method?

Is the procedure really just a more detailed analysis of the site coefficients (Fa & Fv)? Does the software used by the geotech just 'tweak' the data from the USGS? What software is used? Are the ground accelerations determined by a site-specific analysis lower than the IBC's because less conservative models and assumptions are used?

Thanks

 

[Focht3]

The IBC says, "... and the characteristics of subsurface site conditions."

The USGS data doesn't take into account the site-specific geotechnical data, groundwater level, etc. These can (and will) affect the site coefficients. For example, the Mexico City earthquake of September 19, 1985 clearly showed the effects of soft clays in altering the site coefficients. The buildings acted as though they were sitting on a deep bowl of Jell-O, with much lower ground motion frequencies.

To me, $3,000 is too damn cheap for the effort and services provided. It is money well spent - even if you end up with about the same numbers as the USGS analysis indicates.



Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora.

 

[jheidt2543]

Never hold a dime so close you can't see the dollar behind it! $3,000 is a small percentage of the total project cost and should be looked at as NECESSARY insurance.

 

[mrwgeo]

"site specific" means different things to different people. $2000-$3000 likely means they are using a program like FRISKSP to perform a probabilistic assessment. The only consideration of site geology will be in selection of the attenuation relationship. Probabilistic studies can be very good. The one for the Yucca Mtn. nuclear waste repository cost something like $3M. A better site specific method requires on site boring data analyzed with a program like SHAKE; depending of the depth & difficulty of drilling you are looking at at least $6-10K. My experience is that most purchasers don't have a clue what they are getting, they just want to "check a box." The question you need to discuss with your geotech and the project owner,is how much difference a higher quality study will make in your design. A big part of the decision is how much risk is acceptable: acceleration with 10% probability of exceedance in 50 yrs? 2% probability in 100 yrs? For most projects (e.g. typical water tanks) my opinon is that the probabilistic assessment from USGS website is sufficient.

 

[Mccoy]

David 32,

Is the detailed on-site seismic analysis requested by regulations for the specific envisaged project?
Do the project's charachteristics (cost, height, geometrical pattern, strategic importance, open to the public) make it sensitive to ground acceleration?

In my place (Central-Italy, higly seismic) 2 procedures are adopted:
-the Nakamura method, which studies the weak signal (microtremors) coming from traffic and other human activities

- The computer simulations, using as input the project seismic acceleration (signal unaffected by loose soil alterations, i.e. measured on bedrock, subsoil geometry and properties. Tipically used softwares are SHAKE and QUAD, accurate results require accurate soil investigation

Then you have to apply factors related to topography, subsoil geometry and more

You end up with a site amplification of the basic seismic acceleration (measured on bedrock), which may be positive or negative. Mexico City is a conspicuos example of positive site amplification.

The complete procedure is costly and is rarely applied in practice.

In a few months new seismic regulations will require the geologist to produce (for structural engineers' use) seismic site spectral ampliìfications
based on local geology plus parametric classification of soil. It's my guess that people will apply the more conservative spectra to save on site investigation, spending a lot of $ in RC overdesigning.
The detailed procedure will be required only for sensitive projects and areas

As you see, this is much more than basic probabilistic magnitude contours

 

[david32]

jheidt,
I didn't mean to infer that $3000 is too much for a site-specific study. What I meant was that I thought $3000 sounded low considering the complexity of the analysis. The IBC says that you can get up to a 20% reduction in ground accelerations from those determined by the code, if justified by a site-specific study. Obviously, this could save a lot of money if the job is big enough. And even if the site-specific ground accelerations end up being higher, that is insurance that I would appreciate too.

 

[david32]

It's interesting that the code (IBC and maybe others) doesn't go into any detail as to how in depth this site-specific procedure should go. It only indicates what factors should be considered in somewhat vague terms.

It is my guess that this section of the code will be developed more in future editions. Some code requirements a century ago consisted only of statements such as, "the building's structure shall be sound." So much dependence on engineering judgement, rather than code guidance, seems to be on its way out.

The vagueness of this code language leads to different interpretations of what is adequate for a site-specific analysis. So far, I've heard cost estimates ranging from a couple thousand dollars, to 6 to 10 thousand, to 3 million for the Yukka Mountain investigation.

I would really appreciate it if other geotech's could post some info on the site-specific studies they've done. Perhaps, the size of the project, the approximate number of hours worked, the approximate fee, and the degree of complexity undertaken for the analysis??

Thanks!

 

[jas2]

david32:

I am glad to see that you have an extended interest in the geotechnical engineer's role in your project. When it comes to site analysis for seismic purposes we typically measure the shear wave velocity so that the sturctural can assign the proper site classification based on IBC, NEHRP or UBC criteria. On occasion we will extend this service to determine site specific ground motion parameters such as acceleration and velocity.

The codes consider the upper 30 meters significant in determining the ground motion at a particular site although the 30 meter depth is probably arbitrary and for most projects borings do not extend that deep and the geotechnical engineer has to guess. We measure shear wave velocity using geophysical methods that are quick and easy allowing us to provide the V30 for a fairly low fee starting at about $2000. We provide this service for almost every geotechnical investigation we do.

We practice in the Northwestern US (mostly Oregon and Washington) and find that it is not always easy to sell improved knowledge of ground conditions for the purpose of improving the structural's seismic design regardless of the project size.

Recently we were not chosen for a 5 story parking structure investigation primarily because we insisted on determining the seismic site classification rather than guessing. Guessing is still the traditional approach in our area and an accepted one yet there are far better methods. The trouble is that many geotechnical engineers are not well trained in geophysics and do not have the equipment. I think that this will change.

It is faily easy to take the shear-wave data along with appropriately selected seismic records and carry out a Shake analysis but it does require some research and some actual thinking and for that it is not uncommon to boost the fee to $8,000 or $10,000.

Hope this helps.

y.t.

jas

 

[shemp]

Thanks JAS, that's very helpful.