Engineered Fill Settlement - near in-ground pump station 1

[ATSE]

Looking for industry-recognized method of calculating settlement of engineered fill near in-ground concrete tank (new pump station)

The new pump station (wet well) is a rectangular box structure about 30 feet deep and the excavation will be sloped back or stepped in 5 feet increments (1V:1H).

In order to minimize damage to the "new" concrete wall of the pump station, our drawings and specification do not allow the contractor to use heavy equipment (like a Cat 815) vibratory roller within 7 feet of the wall (also, we do not allow Rammex P33 within 5 feet). Instead contractor must use "hand equipment" (such as a Wacker BS50) close to wall. After you see a new thick concrete wall crack under resulting unknown but very large lateral pressure from compaction, you will do the same.

Instead of the typical 95% compaction we usually get using Caltrans Class 2 aggregate base (relative dry density per modified Proctor ASTM D1557), we are getting something less than 90%, but not sure how much less, and nuke testing is sparse and variable. However, let's say we get 85% compaction for discussion.

So the problem: There are concrete equipment pads with slab edges butting right up to the pump station (elevated) slab - separated by premolded expansion joint. These equipment slabs sit over 30 feet of new engineered fill, while the adjacent structure (pump station) is bearing 30 feet down on competent, over-consolidated silty clay.

This is a classic case of differential settlement waiting to happen.

Geotechnical engineers seem to like waving their hands and then state this differential settlement by writing "Approximately 1" to 2" of settlement, depending on..10 variables...5 disclaimers..."). Do these guys actually do calcs? A 2" step between slab edges after 5 years is an embarrassment for the engineers and the contractors.

For lack of a better method, I assume 1/2 % to 1% total long-term settlement for engineered fill compacted to 90%. I have no technical backing, though.

 

[cvg]

backfill around the wetwell with flowable fill especially within the 5 - 7 feet area. It might be cheaper, is certainly much less labor intensive and will reduce the differential settlement greatly. A lot of agencies require this, especially in public right of way.

 

[paddingtongreen]

Take a look here,

http://www.concretepromotion.com/education_6.html

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[ATSE]

Thanks for the response.

For the record, we require CDF (controlled density fill, which is very similar to flowable fill) for the lower zones of the excavation, but allow engineered fill in the upper half (for this project, 15 feet).

30 feet of CDF x 7 feet wide seems a bit unreasonable. For the mid-size pump station in question, that's about 1400 cubic yards. About $100,000 USD for partial backfill, then the extra cost for soil export / offhaul (soil not put back since it is displaced by the CDF), the the soil backfill for the balance. At that point the client says he's willing to accept the settlement (even though he'll change his mind after he sees it 5 years later).

So, what about settlement calcs for 15 feet to 30 feet of engineered fill, assuming the structural load above is equivalent to 2 feet of soil (i.e. self-compacting settlement of 32 feet of engineered fill)?

 

[cvg]

don't forget you are hauling off the dirt and replacing with Caltrans AB anyway. The cost is to take the Caltrans AB that you are already using, add 1 sack cement and water and pour into the trench. No compactors required.

 

[ATSE]

cvg,
Good point. Note that if the site soil is acceptable, then import goes away.
While working in ground improvement, I've mixed aggregate base and cement, but results in non-uniform base and usually stiffer than you think.
I have only seen CDF from a ready-mix truck. Do you know of any state highway departments or public works departments that permit field mixing (cement and native or cement and AB) and is there a recognized spec? Otherwise, a ready mix truck delivering the cheapest materials won't show up for less than $75 per cubic yard (vs. approx $25/ton of road base).

 

[JedClampett]

I've heard that 2% settlement times the depth of fill is a good assumption. However, if the contractor really pays attention to compaction, then you'll only get about 2%.
We run into this issue all the time. So I take the equipment pads and make them part of the structure. Run top bars out of the pump station top slab or wall and structurally cantilever the slab. For design, assume the backfill provides no support. Unless your slabs are pretty long this should work. It doesn't really add too much cost to the pads or structure.

 

[BigH]

Even if the "pads" - that are separate - settle more than the structure (of course), can't the equipment on the pads be shimmied when and if necessary?

 

[fattdad]

I'm just calling B.S. on your indictment of geotechnical engineers. To come into a forum of geotechnical engineers and say stuff like that is bad form!

Firstly, you can use a motorized compactor against the lower depths of the wall, as the permanent soil load will be pretty high (i.e., 1,200 psf horizontal at 20 ft).

Secondly, you can specify the level of compaction and also the soil modulus value of the completed backfill (earthwork contractors won't like this and you'd pay accordingly).

Thirdly, you can use a dilatometer when complete to directly measure the soil modulus (and calculate settlement).

Fourthly, you can install settlement plates and design a schedule "lag" to collect data to see just how much self-weight settlement is occurring.

I'm not so sure about using flowable fill as backfill against a 30 ft deep structure. This topic has come up in the past, but I just don't know how the transient horizontal load of the wet concrete affects the structure. I mean wet concrete would impart a horizontal load of 135 to 150 pcf. Then again, I guess you could do it a few feet at a time and be just fine.

f-de

¡papá gordo ain’t no madre flaca!

 

[ATSE]

Thanks for the replies.

BigH,
Shimming is not out of the question, but after 1 year the general contractor is dis-associated with the project and the equipment base plates have been grouted.

fattdad,
No disrespect to the better half of the geotechnical engineers out there. If you took offense, you're probably not the kind of geotech engineer that does the hand-waving trick (instead of rigorous calculations).

For you 1st point:
Calculating the instantaneous lateral pressure from a Cat 815 on a 24" thick concrete wall (with 2-way action) is a bit tricky. If you have a technical approach to this with case study backup, I'm interested in seeing it. Otherwise, the safe position is to keep such equipment far away.

For your 2nd and 3rd points:
I realize that I can assess the settlement after the fact. However, this situation I've described is relatively common, and I don't want each project to be a research effort. I'm looking for upper and lower bound settlement estimates.

Also, we limit CDF placement to 5' max vertical lifts.

America put a man on the moon 40 years ago. It seems like there should be calculation method for long-term settlement based on the variables of total engineered fill depth, gradation, water content, and compactive effort. I don't need and don't expect an exact number, but for a lower bound and upper bound would be real useful.

 

[aeoliantexan]

Regarding compaction-induced lateral pressures on retaining walls, see NAVFAC DM-7.2, page 77 and the source article by Ingold. I doubt that such pressures can be accurately predicted, but on the basis of this work, I usually recommend keeping rollers at least one roller width away from the wall and using a minimum pressure of 400 or 500 psf above the depth at which traditional methods give a higher pressure.

For some basis for estimating long-term self-weight settlement of compacted fill. see "Hydrocompression Settlement of Deep Fills" in ASCE Journal of Geotechnical Engineering, Vol 116, No 10, October 1990, and "Collapse of Compacted Clayey Sand" same journal, September 1989.

I reviewed a lift station 30 feet deep backfilled with a combination of select sandy clay and on-site fat clay presumably compacted to 95% Standard Proctor. It was tested by sampling with a drill rig every 5 feet, and the test reports indicated compliance. An equipment footing was constructed on the backfill but never loaded. At last check it had settled over 9 inches. I doubt that the compaction was a good as indicated, but that was not the design geotechnical engineer's fault.

The settlement depends on the material type, method and degree of compaction, compaction water content, and how much additional water the fill absorbs after completion. If you can tell the geotech precisely what each of those will be, he can compact samples and run consolidation tests and make a defensible estimate of settlement. If the actual performance is within 50% of his estimate, he is at the state of the art.

I have worked on a lot of pump stations. If two inches (or six)of differential settlement is a problem for you, don't support it on the backfill.

 

[muuddfun]

A couple of the things to do are to cut down your lift hight when compacting next to the wall with the smaller equipment, and to specify full time observation and testing of the backfill.

Generally when I have run remolded collapse/consol in the lab at 90 percent compaction on sandy or silty sand materials the settlement potential is between 1 to 2 percent depending on the load you are looking at. So say the soil at the bottom of the wall at 3.5 ksf overburden would give a total settlement in the 2 percent range and the higher up layers less. The majority of this settlement should already have occured by the time you have finished the backfill to the surface, so the total settlement after the fill is complete would be a lot less. I am generallizing a lot here, but from the tests I have run that is usually about what comes out.

If you have any questions about a critical project then get the geotech to run a couple of remolded soil samples and calculate it for you directly. That is not to hard nor to expensive.

 

[BigHarvey]

Another way around would be to use a diaphragm wall to build your new pump station. This would eliminate the problem of the settlement of the fill and might prove less expensive at the end.

 

[KVgeo]

@ATSE:

Geotechnical engineering is not easy. There are no codes which provide definitive absolute parameters for soils. In most cases engineering judgement is required by an experienced professional geotechnical engineer.

Given the above, when asking for geotechnical advice, you get what you pay for.

Before you shoot your mouth off about geotechnical engineers maybe you should review your proposals from geotechnical consultants a little better and establish a selection criteria which isn't just based on cost.