Global FOS 1.3 vs 1.5 for Retaining Walls 1

[LOTE]

I am curious what people use for the cutoff to increase the global factor of safety from 1.3 to 1.5 for "critical structures", and do you ever use anything between 1.3 and 1.5? It is pretty clear cut to me for highway applications, where if the wall supports a bridge you would automatically increase to 1.5. But how about for commercial sites? Is it certain wall height, front slope, back slope, supporting a certain height building, etc.?

My current issue is I am doing the internal stability for an MSE wall, and the global analysis performed by others used FOS=1.5 (rightfully so based on literature and 2H:1V toe slope, supporting a 3-story building 15' back, and 26-degree foundation). But I am scratching my head at 42' grid lengths with a 14' exposed height and 5.5' of embedment. The owner is asking if there is any way to decrease the excavation. I have confirmed the global stability factor of safety is right at 1.5 with these variables. I wouldn't propose using a lower FOS unless I could find justification for it. I have been playing around with embedment and undercuts with not much help. Unless there is a convincing reason to use FOS<1.5, I do not see an alternative (with the current wall type and without other ground modifications which are beyond my expertise).

 

[Jrit]

Those numbers (grid length and wall height) seem strange to me too. Are those grid lengths coming from your internal stability analysis or the external global stability performed by others?

 

[LOTE]

Those are from the global analysis by others.

I have replicated the analysis in my slope stability software and am getting FOS=1.5 as well. I though maybe I would find a section of the slope that wasn't accurate from the grading plan, but that was not the case.

 

[geotechguy1]

What soil properties are you using? What geogrid properties?

 

[LOTE]

The soil properties are from a geotechnical report with recommended friction and unit weights for two layers of soil under the wall. The upper layer is 26-deg and 120 pcf for 10-ft and the lower layer is 28-deg and 115 pcf.

Using sands in the reinforced zone and the recommend values for Ci and Cd from the Synteen literature.

 

[geotechguy1]

Hypothetically, what factor of safety do you have for the existing condition of the slope with those properties, prior to building your wall?

 

[MTNClimber]

Duncan & Wright's book "Soil Strength and Slope Stability" provides great guidance on determining the required safety factor for slope stability. It uses the cost and consequences of slope failure and uncertainty of analysis conditions to determine a minimum factor of safety. In that table, I would say you fall in the category of: the cost of repair is much greater than the incremental cost to construct a more conservatively designed "slope" (in your case, a wall). As far as the uncertainty of analysis conditions, it's likely considered "small" if you have borings, lab tests, and recommended soil parameters from the geotechnical engineer. With these two parameters, a minimum factor of safety of 1.5 is recommended, which is reasonable.

NCMA's Design Manual for Segmental Retaining Walls 3rd Edition recommends a minimum factor of safety for Global Stability in static condtions 1.3-1.5 and 1.1 for seismic, which is also reasonable. It's up to the engineer to decide what paramaters to use.

I usually don't mind use 1.3 for most walls. When it's supporting a building or even close to a building, I may bump it up to 1.5 if there are some concerns of uncertainity and consequences are high.

42' is very long, but those are some tough soil parameters provided by the geotech. They may or may not be warranted, but that's tough to argue with geotechnical engineers on their soil parameters. Things you can do to improve global stability are embedding the wall deeper and making the grid longer. Other than that, you have to look at alternative ground improvement, doing a combination system (driving sheet piles at the toe or maybe doing a reinforced earth slope at the base and then putting your wall on top), or changing to a totally different wall type altogether. If you have shallow groundwater in the toe slope, you can increase stability by installing toe drains.

 

[geotechguy1]

We regularly design walls and slopes in NZ with seismic FoS in the 0.5 range.

I'm very curious to know what the static FoS is using those parameters. North American geotech engineers are IMO almost criminally conservative at times in giving parameters. If the slope is stable in it's current condition and you have evidence that it's been that way for an extended period, I'd expect that the long-term static FoS is 1.2-1.3. To get those grid lengths the geotech must be giving parameters that imply a sub-1 static FoS in the existing condition or close to it.

 

[MTNClimber]

I don’t disagree with you geotechguy. I’m not sure how litigious folks are in NZ but it’s pretty bad in the US which causes a fair amount of conservatism here.

The thing with slopes is that some are slowly failing. So assigning a friction angle based on its current geometry may not be appropriate. You really need to be tapped into the local geology to understand if that’s the case, which we will not be.

That said, I’m interested in what the soil conditions are at this site where the friction angles are so low. LOTE can you provide a boring log redacting the project and geotechnical engineer?

 

[LOTE]

 

[LOTE]

I ran the existing condition (based on recommended soil properties) and got a FOS = 1.04.

 

[geotechguy1]

Here's my two cents: If you use parameters back calculated or that imply a Static FoS of 1.0, with a long-term static groundwater condition, then you will always end up with an extremely expensive wall regardless of what type of wall is designed. IME it will also be incredibly costly to try and achieve a global FoS of 1.5 instead of 1.3. If, however, you use parameters derived for 1.2 - 1.3, then a forward calc using those parameters should achieve 1.5.

If the static FoS is actually one in existing long-term conditions than I think that their should be significant evidence of ongoing movement and failures in your slope. I.e., every extreme rainfall year or snowmelt year, you should be seeing failures in the slope. I would even expect to see failures in 'normal' years because an FoS of 1.0 is really more like a 50% chance of failure, right?

What does the geological / historical evidence in your area say? Your parameters and model imply that extremely large, deep-seated rotational failures should be occuring regularly. Are they? Is there evidence of regular failures of this type on similar slopes in similar geologies in the Holocene? If you were to look at historical evidence in INSAR sattelite deformation data from the last 10-20 years if it exists, would you find evidence of significant, regular movements happening in these slopes? What does this slope and slopes in similar geologies look like - do they look like a mess of past failures vegetated over?

 

[MTNClimber]

Considering the blow counts are from an automatic hammer, only the first 8’ are relatively low. I would say the top 8’ is loose and would conservatively warrant a 28 degree friction angle, followed by 32 degrees for 8-30’, and 34 degrees for >30’. I’d start there before fiddling with the soil properties and wall geometry more if it was my project.

The top 8’ could be excavated and recompacted to increase the friction angle to 34 degrees.

If I’m still having problems at that point, I may be looking to go in a different direction.

 

[LOTE]

@geotechguy1 Those are all great questions that I do not have the local knowledge to answer.

@MTNClimber Based on the limited info, do you think the geotech was too conservative here with their friction angles?

For liability, I would not go and use higher friction angles than what the geotech recommended in their report. I try to stay away from proposing soil properties whenever possible.

 

[MTNClimber]

If a geotech recommended those values in my area, they could be argued with. If I had the client's ear, I would recommend the owner never hire that geotech again, as they are costing them money and probably preventing projects from ever being constructed.

This is based on the very limited information I have. That doesn’t mean there isn’t a reason for the very conservative values. As geotechguy stated, if there are many failures in the area, the geotechnical engineer may need to provide highly conservative values to avoid loss of property or life.

 

[geotechguy1]

If you can't get the geotech to revisit it then you are dead in the water. The grid lengths I think make sense based on the parameters provided, you are trying to achieve FS = 1.5 from a condition of FS=1 with low parameters so you will end up with long grid lengths.