Another Sewage Treatment Wall Collapsing 7

[JedClampett]

Even though this seems to be a poorly constructed project, I've got to believe that there is a design aspect to this.
Note that the Gatlinburg Wall Collapse looked very similar to this one.

 

[msquared48]

It will be interesting to see the final reports...

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

 

[JedClampett]

Well, here's another piece of the puzzle:

 

[MikeHalloran]

There's mention of checking the design documents for hydrostatic pressure.

... but there's not a hint that they were checked for the increased effective hydrostatic pressure exerted by a load of Imported Norwegian Magic Rock.


Mike Halloran
Pembroke Pines, FL, USA

 

[hokie66]

Thanks for keeping an eye out for these, Jed. The end walls just zipped off like the one you posted earlier.

 

[a2mfk]

Nice pick up Mike, that stood out to me too. First, what the hell are these special sewage treatment rocks that we need from Norway? Last time I checked the east coast has a pretty wide variety of rocks... And like you said, did they design the walls for that additional load?

From skim reading the structural section of the report, seems like enough red flags were raised that the owner should have perhaps taken more drastic action. I know it is very expensive and they may not have many alternatives to shutting down sections of WWT plant, but look at their options NOW.

And this was just built @ 5 years ago! Not to overstate the obvious, but this should not be happening, especially in the US on municipal projects where public health and safety are paramount.

 

[JedClampett]

From the report I linked to earlier, I'm more suspicious of the inquiry about the ACI 350 design. I suspect that the invesigating engineers did some rough calculations, couldn't make them come out and wanted to see the design company's calculations to confirm their suspicions.
Of course there's almost never one single cause for a disaster like this. Several things had to go wrong.

 

[JAE]

Using 350 vs. 318 wouldn't make a difference in the basic ability of the tanks to resist the required loads. All 350 does is provide more reinforcement to keep cracks and leaks less likely.

The extra reinforcement in 350 would of course add to the capacity of the walls, etc., but if designed correctly, under 318, there shouldn't have been a collapse.

 

[JedClampett]

Except that ACI 318 allows you to use a load factor of 1.4 for Fluid Loads, where ACI 350 requires 1.7 or more.
That's what I'm wondering about. It's unlikely to be something marginal because I doubt most engineers would raise a concern on a 5% or 10% difference.

 

[JAE]

ACI 350-06 has 1.4F. ??

 

[JedClampett]

ACI 318 still has 1.4 as far as I know.

 

[JAE]

I didn't see a 1.7 in 350...that's all I was asking. What version of 350 has the 1.7?

 

[JedClampett]

I'll have to check. I think it bottoms out at 1.7 if all the spacing of reinforcing meets requirements. I'm in MikeHalloran's neck of the woods today, so I don't have my references.

 

[hokie66]

1.4...1.7...whatever. That's not what made those end walls fall off. I don't know about the rock. That could have contributed. But like the other failure which Jed posted a few weeks ago, the detail at the wall intersections, as designed and/or as constructed, was obviously the weak link.

 

[a2mfk]

This video was in a link from the original article.

Agree Hokie, I remember the other W VA failure was similar, where the front walls met the side walls. There was a "pull-away" failure of the front wall apparently from inadequate development of the horizontal reinforcement. But that W VA wall was fairly tall, this one seems like it is more in the 10-12' range judging from the video.

The pic from the main article- look at the walls in the background, there are horiz rebar stubs visible extending out from the side walls. So from this limited info I would speculate, if I must,that there were no hooks and the wall simply pulled away from the side wall (I can't imagine them detailing it like this!), or they did provide hooked bars at that connection and they failed in tension.

WWTP tanks are not my area of expertise, but square box tank structures with uniform pressures don't seem THAT complicated... Any other theories from some tank designers? If there were that many leaks as the peer review report indicated, could that be from rapid corrosion?

 

[hokie66]

Just to get the geography right, this one is in New York, and the earlier failure was in Tennessee.

 

[HeavyCivil]

a2mfk - I don't often see standard hooks provided for horizontal bars in tank walls unless geometry requires it. The need for horizontal reinforcing is obviously higher near the corner, so running corner bars the full ld or 1.3ld into the wall can be beneficial in giving you extra area of steel where its needed.

What is the gradation of those special rocks? I bet they're pretty granular - lots of surface area for microbes. If its like a coarse poorly graded filter sand or pea gravel I can imagine some really significant lateral loads developing there at saturation.

 

[a2mfk]

You are correct Hokie, I forgot where Gatlinburg is. Up near your old stomping grounds, at least that neck of the woods...

 

[JedClampett]

As far as the post from a2mfk, I design a lot of these and the main lesson I would pass on is that while these are not complicated, the forces are more "real" and unrelenting than normal building type structures. When you design a building for a 100 psf LL and a 90 mph wind, it's unlikely that they're ever going to see either one. You might get 30 square feet of bookshelves where there's 100 psf or the wind might approach 90 mph at a 60 degree angle to the wall, but neither is close to the design loads of 100 psf over every square foot of the structure plus a 90 mph wind over the whole wall of the building, plus increases for corners and edges.
Compare that to a water or wastewater tank. The pressure is 62.4 pcf times the depth of the liquid. It can't be less than that. This pressure is likely to be there for a long period, maybe years at a time.
There are some companies, big and small, that design for these loadings all the time. We have procedures in place to assure a good design product. Even we make mistakes. We learn something from every project. But on occasion, companies dabble in this work. They're the local favorites, they know somebody, they're owed the job, whatever. If they don't respect what they're doing, mistakes are made. I'm not going to say that they can't do a good job. It's just a different mindset.
When hokie66 says "1.4, 1.7 whatever?", he might be right, as long as everything else is done perfect. It's a dirty little secret, but a lot of the load factors, phi factors, cracking calculations, etc., are there to protect from normal mess-ups in construction, design and operation.
Enough of my soap box ranting. Just remember, next time you see a tank full of poop, thank people like me. And if it's running down the creek, thank someone else.

 

[ATSE]

For engineers not familiar with rectangular fluid retaining structure design, a common error (among many) is to ignore the combined tension and shear at wall intersections (Tees and corners). High tension and shear interaction, combined with peak flexure at intersections, can cause these unzipping failures.

Instead of single hooked bars, I like U-shaped bars at corners and Tees, along with vertical bars at all inside corners.
In addition to that, we always use corner bars (45 degrees in plan) to address the reduced concrete shear capacity due to the aggravating additional net tension.

As much as I dislike tort lawyers and "expert witness" engineers that usually seem to just feed on their own, they do serve a purpose in these unfortunate events. Bottom feeder engineers will steal jobs away from the experienced engineers in the short term, but also get purged from the system from time to time.