IBC Seismic Site Classification 1

[kmmgeotech]

I am persistently questioned with respect to the IBC Site Classification for subgrade conditions especially when designating an "E" site (Soft Soil Profile). In many cases, we have a loose sand or soft clay that marginally does not conform to the specific IBC characteristics (liquefaction, high PI, low shear strength) for "E" or "F". On many projects, bedrock is encounterd within 20-30 ft of grade. I have been told to review the subgrade conditions within the top 100 ft (20% loose sand/80% bedrock). This would place the Site Class to possibly be a "C". It just seems to be at both ends of the Spectrum. Marginally good enough not to be an "F", likley to be an "E" but possibly being a "C" site when considering the shallow bedrock. What happened to "D"? I find the Code to be very vague and ambiguous. It was mentioned that shallow soil conditions(20-25 ft to bedrock) will not amplify as much as a deep soil deposits and for that reason I should account for the bedrock in the upper 100 ft. Someone please HELP explain this to me. Thanks.

 

[gheb]

kmm-

First off you need to treat the top 100 ft as an average for site classification (you can't say the top 20 ft are an "f" and the bottom 80 ft are a "C" it doesn't work like that.

The reason that IBC asks you to look at the full top 100 ft (30 m) is due to the fact that the local site amplification is controlled mainly by local site conditions in the near surface (shown to mostly dependent on the top 100 ft.) see (Borcherdt, R.D. (1994), “Estimates of Site-Dependent Response Spectra for Design (Methodology and Justification)”, Earthquake Spectra, Vol. 10, No. 4, pp. 617-653.) for the justification of that...

Anyway, in the case of an earthquake, the energy from the source (typically located at depths ~ 10 km or so below ground surface) propogates outward radially through the surrounding bedrock. For your fictiscious site, say 100 miles away from the source there will be two ground motions to deal with: the energy coming almost straight up from the deep bedrock directly below your site, and also the energy traveling along the earth's surface in the form of a surface or "Rayleigh" wave.

The reason that the local site conditions impact the behavior of structures at a particular location is that as the earthquake encounters the local soil there has to be a conservation of total energy as the wave is transmitted from one soil to the next. To not get to technical the energy flux needs to remain constant in the following relation: Energy Flux = density*Vs*u^2 (where u is the particle velocity - directly related to the soil movement). So as the wave travels from depth to the surface (i.e. from deep rocks and bedrock which are dense and have high Vs to surface soils which typiclaly have lower density and Vs)the particle velocity and thus particle movements need to change (typically increase) in order to keep the energy flux constant. So for a soft near surface soil you can get up to say 3-4 times the movement than at a dense near surface site.

Ok, so the reason they ask you to look at the soil properties (Vs) in the whole top 100 ft is that earthquake waves exist at long wavelengths and so the scale at which they can be influenced by local properties is controlled by changes on the order of that scale (100 ft and greater).

As for skipping over "D" to "C" your colleagues may just be saying that by properly characterizing the soil (in the whole top 100 ft) you may get find out your site is "much" better and not just a little bit better.

The book by Kramer "Geotechnical Earthquake Engineering" is a must have for anyone doing seismic site assesment and goes into the details of all you need to know.

 

[jgailla]

Good post gheb. Your response is an excellent discussion of this topic.

kmmgeotech,
I also have trouble with the seismic site class. The IBC is fairly easy to follow in most situations. However, given limited exploration data (like 50 foot borings), I am hesitant to raise the site class without other experience with deeper borings in the same area. I can't just assume conditions exist at unexplored depths on a site without experience in the area.

 

[gheb]

jgailla-

I guess that is always the main thing that must be weighed in a lot of cases. Typically sites that have sufficient high Vs material to be site classes "C" and above are typically comprised of a fair percentage of rock, or other dense materials typically not characterized in basic site characterization studies.

I'm pretty sure that you're not allowed to raise the site class without actually measuring the properties, either using direct Vs measurements or the allowed SPT-N conversion over the entire top 100 ft. The question then becomes, is the cost of conducting a seismic investigation, whether it be using downhole, crosshole, surface wave methods, etc., worth the savings that one can gain from the improved site class. I would say that in most cases if one knows the local geology well, you will be able to have a certain feel for whether the Vs measurements will raise your site class or not. Surface wave methods have advanced significantly in recent years and are probably your cheapest option, on the order of a couple thousand dollars. (Try and find a firm that uses something more modern than the classical SASW 2-sensor technique though as data confidence is greatly improved with more modern applicatoins) If you can inprove your site class to either "C" or "B", the cost of the Vs investigation will most likely be small compared to the savings in structural and foundation design and construction costs, depending on the magnitude of the structure.

There is always the possibility you'll make the measurements and find an improved site class is not warranted, but that's basically the crux with any site investigation.

 

[kmmgeotech]

Thanks for the responses. My initial concern was whether I could consider BEDROCK as part of the upper 100 ft site profile or if only SOIL conditions were to be considered. It makes a BIG difference. The IBC is not very clear in this respect.

In many areas, we have very hard bedrock typically within 20-50 ft of grade. If I consider bedrock, I may use (N=100) or a Vs of 2,500 ft/sed (Class B rock). As such, should I consider the 50-80 ft of bedrock in my analysis. My concern was many areas have loose sand or soft clay over hard bedrock.

 

[gheb]

kmm-

You in fact always need to consider all of the top 100 ft, no matter the material (e.g. soil, rock, solid waste, etc.).

Basically what you're doing is determining the travel time of the wave throuh the material. That's why you don't use a straight average of Vs but rather the representative average Vs that gives the same travel time through that length of material. As per the IBC Calculations for Vs30(m).

Also before applying a value of 2500 ft/s for bedrock you need to be able to verify the velocity of the local bedrock as rock velocities vary over a large range (say 1000 to 5000 ft/s) whether through field (preferred becuase fractures or faulting will influence velocity) or lab measurments. Blindly assuming 2500 in all cases will not provide a correct assesment of site response and does not provide a basis for improving the site class.... Just like you don't blindly assume soil properties based on characterizaiton as sand, clay, etc., you can't just asign an average value of Vs to all rocks.

The IBC does allow for the unmeasured assumption of N=100 over a rock strata but in most cases that won't be enough of an increase to raise your site class... i.e. you need the large jump in Vs from soil (on the order of 300 - 750 ft/s) to rock (1000 - 5000 ft/s) in order to raise the calculated Vs30 sufficiently.

So basically, even if you have a very loose soil at the surface, if it only extends down say 20 ft to solid bedrock your site class will probably be improved to "C" or better. However, if there are liquefaction concerns, no matter what the depth... the site class is always "F" and you're supposed to conduct a thorouh site specific seismic analysis with SHAKE or some other program.