Levees in New Orleans 13

[livingston]

After Hurricane Katrina hit Lousiana, more destruction is occurring because of levee failures around the city. Does this mean the factor of safety was too low, they were old, what? They seem to be failing at the purpose for which they were designed. Please explain.

 

[CajunCenturion]

Yes, there is. Here is a good summary.

Reengineering The Mississippi



Good Luck
--------------
As a circle of light increases so does the circumference of darkness around it. - Albert Einstein

 

[duck0601]

CajunCenturion,

Thank You. I will look closer later at the LCA site mentioned, but, estimated cost $14 Billon vs. cost of Katrina (so far) $100 Billion. Sounds comparatively low cost to me.

Chris

 

[ARenko]

more on the marshes (with prophetic story of "the big one" hitting NO)...

http://www3.nationalgeographic.com/ngm/0410/feature5/

 

[Slugger926]

Since my sister is a research engineer that will have to inspect the failed levees, and already has seen closeup videos of the walls as they failed, I will chime in with what happened.

The water overtopped the walls. If you notice the walls are built on top of earthen berms. The water started eroding the earth berms below the wall until headcut erosion started occuring. Once the headcut happened, erosion happened very fast with the amount of CFS that was flowing at these points.

The headcut erosion started undercutting the foundation of the wall itself until the foundation gave way. It was the berm and the foundation of the wall that caused the walls to fail once high volumes of water were going over the walls.

Also note that these walls were built decades ago, but would have probably failed the same way if they were built 2 years ago. The walls needed a foundation. THe earthen berms and walls needed to be higher and wider for the high volumes of water.

She will be going down there next weekend, then on to a conference in Orlando.

 

[BJC]

This weeks ENR had a good story on the dikes and what happened. They are the best source AFAIC on the subject. CNN is just a step above Fox news. Good pictures but no analysis or anything in depth.

http://www.enr.com/features/_Covers/2005/archives/default_weekly_mag.asp

 

[CajunCenturion]

Interesting Slugger926. The headcut erosion makes sense, and it means that the levees failed from the middle. The 17th street levee had a cement wall built on top of an earthen levee. The high wave action eroded away the top layer of the earthen levee, exposing the bottom of the cement wall, and opening a hole underneath the cement wall. A continuous flow of water through that hole created headcut erosion of the earthen levee below the cement wall. Once enough of the earthen levee was gone, the cement wall had no support and collapsed.

So you should either build the levee entirely out of earth, or entirely out of cement, but not both. The failure originated in the middle of the levee, at the transition point from earth to cement.

No matter how high you build a levee, nature can build a bigger wave, but you don't have to protect from all overflow. You can allow a reasonable (whatever reasonable is in this context) amount of water over the levee, which you can take care of with the pumps. But under no circumstances can you allow the levee to breach. The flooding is not caused by large wave overflow. The flooding comes from continuous sea level flow.

Good Luck
--------------
As a circle of light increases so does the circumference of darkness around it. - Albert Einstein

 

[BillPSU]

I lived in St. Joseph, MO in 1993 and experienced the Missouri river flood. Elwood, KS which was across the river was protected by earthen levees which were overtopped by the river. Immediately erosion occurred and the levee failed. Overtopping any levee just about assures destruction of the levee.

 

[Slugger926]

Bill - I worked at the same research facillity my sister is at in 93-95 in the summers. We did research on that levee erosion as well as those that were lost in the Miss. flood in 93 or 94. It is amazing once a headcut starts to how fast the erosion can occure.

Anyways, my sister isn't having to go to NO now. They will be heading straight to Orlando.

I will be heading to NO the second or third week of October tosurvey speeding the recovery of our network.

 

[unclesyd]

MSNBC (Rita Crosby Show) just some good closeup footage of the failed concrete flood walls. As I stated before the concrete doesn't impress me. One section of the flood wall was concrete incased sheet piling, looked like 6" thick on either side. From the failure mode the composite appeared to have very little lateral strength (cracking pattern). Some areas the concrete had completely broken off. Other sections showed only vertical rebar completely broken away from the concrete. The embedded depth of wall the appeared to be about half again as deep as it's height above the top of the levee. The wall shown had no base it was a straight and slim.

 

[unclesyd]

There is an interesting report from LSU Hurricane Center about the failure mode of the flood walls and levees. One of the reporters on CNN was trying to corner a COE spokesman about the report and the spokesman had the all the moves of River Dance trouper. It was the same with the next interviewee and ex Congressman who was dancing around better than a scared fly.

From the little information gleaned, the report say the walls themselves failed, this is counter to the overtopping scenerio being expounded on. This has bother me from the start about the flood walls. Not knowing the proper names, but the absence of a flange either at the bottom of the wall or at the surface in my opinion is a no no. To me what I'm seeing is a lever with the backside top of the levee as a fulcrum.

Aside from the engineering point of view this would have a tremendous effect on the insurance questions.

 

[vooter]

The floodwalls are I-walls. Concrete on sheet piles. Look for where the concrete failed...

 

[unclesyd]

vooter,
I agree that some of the flood walls were the I-walls, but several sections were just vertical straight sided reinforced concrete. The sections that were only reinforced concrete appeared to have only vertical reinforcing bars, fairly large and somewhat large spacing of the bars. I have not one time seen concrete sticking to the rebar. A little information picked up this morning is that there is a CORE limit of 7' above grade and some of both types were at 11' above grade.

I have made several comments about these walls, especially the appearance of the concrete, based on snippets of visual information from fleeting TV coverage. Based on many years looking at failures I sure hope they have a geotechnical (concrete) look at failures. But they are using a lot of the wall rubble to help shore up the existing levees.

In construction of the I-walls is the two facings of concrete supposed to be reinforced and/or tied togather through the sheet piles?

 

[unclesyd]

A few more tidbits:
The report from LSU has as one of it's author/authors a gentleman named Paul Kemp. During his little time on TV it showed another problem with the flood wall. The joints between sections had varying widths and it looked like the elastomer/bitumen has deteriorated and created a leak. He stated that the failure wasn't form surge or overtopping of the walls.

vooter,
As you stated some of the walls were I-walls but during the interview it showed a small section of a failed flood wall of a different construction than has been mentioned. It was sheet piling with concrete faces as you described. This concrete appeared to be reinforced. All the concrete was

 

[unclesyd]

Fox News just showed a new levee break on the Industrial Canal. There is a good possibility that this could flood the 9th Ward again. From the pictures on TV it was a fair size break of the levee itself, there appears to be no flood wall.

 

[vooter]

I walked the wall failures three days ago (that's revealing a bit to much about myself, but...) and saw for myself that they're I-walls. They are reinforced in both directions, too. I will post my photos in the future...

There's no CORPS limit on wall height as long as the design is sound - the NYTimes article (and others) that quoted the Corps's manual regarding 7 ft vs. 11 ft is misleading.

The wall panels look like they failed as cantilevered plates; the sheet pile is in good shape, no failure I could see at all.

 

[unclesyd]

vooter,
Repeat of 9/22/05 with the missing part.
As you stated some of the walls were I-walls but during the interview it showed a small section of a failed flood wall of a different construction than has been mentioned. It was sheet piling with concrete faces as you described. This concrete appeared to be reinforced. All the concrete was above grade. The below grade sheet piling appeared to be about 3 times the length of the concrete. Is this still called an I-wall?

 

[vooter]

From

http://www.pz27.com/glossary.html:

"I-wall: A special case of a cantilevered wall consisting of sheet piling in the embedded depth and a monolithic concrete wall in the exposed height."

That's what I described, concrete on (on top of, that is) sheet piles.

How many I-walls have you designed? I've only done four, myself.

 

[unclesyd]

I have actually designed none and don't claim to understand every nuance of the design, but I'm currently looking at failed walls designed by someone who evidently wasn't well versed in the function of said walls. Having spent the majority of my career doing failure analysis work that in certain observations, as compered to statements, it is hard to overlook the obvious. As stated earlier based on early pictures one of the failed walls was solid concrete with only very large vertical rebar. Why would some of the walls have a deeper facing of concrete?


I did do some design work on a building on Pensacola Beach that has survived 5 Hurricanes with 3 storm surges including being about 20 miles East of the center of Ivan.

As far as being monolithic concrete, as I understand it is only in 30' sections and the sections are connected by a rubber seal. It is quite evident that the sections were operating independently from the pictures and the seal in many of the connections isn't in very good shape.

As far as the 7' height limit for this type construction I have seen this published, discounting the interviews on TV, by two supposedly reputable sources as being a standard.

Not saying it isn't possible, I have seen literally thousands of sheet pilings driven and have never seen a short section started on both ends to meet in the middle.

At last report the water in the lake behind this dam was 7' higher in than the water in the canals. They can't used the canals until the lake subsides. There appears to far less freeboard on the groin built around the big Industrial Canal breech that is now allowing water in.

I also understand that they “hope” to have a failure report by June 2006 and have the levees restored at the same time.

The current spin is that the flood walls were overtopped, but there is credible evidence that the case was not so. They missed a good opportunity very early on that would have answered part of the question. The empty grain barge so prominently displayed at the big breech on the Industrial Canal was first mentioned by one of evacuees as hitting and knocking down the flood wall.

 

[Slugger926]

There were multiple causes of failure depending on which failure you were looking at.

 

[Prepakt1]

Prepakt,
Seems to me that the repairs to the levees were
not given enough sealing to prevent seepage.Perhaps non permeable geotextiles or bentonite mat laid on the outside.Then a covering on top to hold in position
The repairs were then sbjected to pressure before they had time to compact properly.
If possible and had time and materials permitted the large bags should have been covered and secured by concrete not gravel. Where gravel was used bentonite mat(self sealing]
should have been laid and covered with flexible concrete matting.
Matting would prevent wash out and the flexible mat should collapse into any depressions occuring due to seepage helping to seal and both indicating a pending problem.


Intrusion Prepakt

http://www.marineconcrete.com