False nuke-gauge moisture %s due to flyash in soil backfill

[Scienceguy57]

I work for an engineering services firm in Colorado Springs, Colorado and I recently had to help our Pueblo branch for a month and a half at Xcel's new Comanche 3 power plant expansion. This was my first experience on a jobsite of "industrial" stature. On the backfill of two large water-cooling pipes, they decided to use "bottom ash" which was a mixture of sand and fly ash to my understanding. The fly ash would lower our nuclear gauge moisture percentage (M%) by 7 to 10% (i.e. 10% was actually 17% to 20%). This was particularly frustrating because the contractor wanted results immediately, and without knowing the true dry density (DD) we could be telling them a failing lift passed and then they may proceed to cover it with more lifts.
The way we worked around this problem was take the wet densities (WD) the nuke gauge gave us and determine the DD with a 7-10% moisture correction. Then after every test that passed this initial field equation, we would record the WD and take a sample back and dry it out over night for the M% and then determine the true DD. We were reluctant to bring a microwave to the sight to determine the M% because that test still requires another sample to be dried out over night by ASTM standards and could still fail. We also did a sand cone for every ten densities, however that could only tell us on the spot if our WDs were inline. On top of that they moaned when our sand cone and nuke gauge WDs were off by a few tenths of a p.c.f., but that's a different story.
So my questions are does anyone know what in fly ash would throw off our M%s by -7 to -10%? I figured it was left over hydrogen from the coal burning process still in the fly ash, but you would reason that extra hydrogen would raise the M% and not lower it. Also, has anyone else ever ran into fly ash used in soil backfill? I can see why a power plant would want to use it because they have an abundance of it, but do they bother to consider the problems that would arise with density testing this material? Their mediocre specifications did not address this at all. It probably would have been cheaper to just flow-filled the whole thing. Anyways, it was a great learning experience and I am glad to be done with this pain in the butt, I hope.

Thanks guys and any thoughts will be much appreciated!

 

[DMcGrath]

I don't know what in the ash would make such a wild change in MC, but the ash itself can vary widely based on the source coal and the type of combustion process. Fly ash and bottom ash are two different materials, derived from different aspects of the coal combustion process. Why not take a sample for Proctor testing, so you know what the optimum moisture and maximum density are, rather than guessing?

Power plants love to use fly ash where they can for fill material because otherwise they have to pay to landfill it. "Unfortunately," as one of my power clients likes to say, "the competition is dirt."

 

[Scienceguy57]

Oh we had a proctor, a standard proctor of 105.9 p.c.f. at 19.7 %M. The trouble was the that the DDs the gauge would give us were typically around 88 p.c.f. and the M%s were around 8-10%, however the WDs were around 118-125 p.c.f. After taking back the sample to determine the true M% overnight and recording the WD, we determined that the M%s were actually around 20% and the true DD and the percent compaction, but this was 24 hours later and they wanted results immediately so they could continue laying more lifts. We knew the WDs from the gauge were accurate because they matched within a few tenths our sand cone WDs. So that's why we did a field calculation using a 7-10% moisture correction to try and roughly figure out the DD and then we took a sample back, dry it out, and determine the true DD and percent compaction the next morning. By that time, if it turned out our moisture correction was off then we could have passed a failing lift that was then covered by a few feet now... My understanding of what bottom ash is may be mistaken. It looked like it was just sand and flyash, but they certainly called it bottom ash, so if they are not the same, then what I am talking about here is certainly bottom ash. The coal for this plant most likely comes from Wyoming if that helps.

Thank you DMcGrath. This problem is far from solved.

 

[DRC1]

Flowable fill is esstenially fly ash, sand and a little cement(plus water. Certian soils will distub a nuke guage. Check with the manufacturer. We usede to do field moistues with a balance, a mall loaf pan and abottle or two of dry gas. Weiegh the wet soil in the pan. Pour dry gas in and swish, but do not stir. light the mix. It should burn off in a few minutes. Reweigh & compute. It's crude, it's not ASTM, but it's fast and works reasonaly well.

 

[Scienceguy57]

I'm using a troxler guage and this is quoted from their website "Why is my gauge moisture result higher/lower than the oven dry/speedy moisture results?
The oven dry tells the true moisture content by removing all water from the sample. The gauge measures the hydrogen present in the material; which is usually in the form of water. If the material contains naturally occurring hydrogen or bound hydrogen, the gauge will measure the moisture falsely high in many cases. (Some of these materials are: mica, lime, fly ash, cement, organic materials, gypsum, coal, phosphates, etc.) A false low reading can also occur but is less common (causes include: high salt or iron oxide content or boron, lithium or cadmium present). A moisture offset will adjust for this problem."

Looks like one of the materials in the false low part must be present.

 

[casimmons]

Scienceguy57, something doesn't add up with your nuke gauge readings. DD=88 pcf, moisture=8-10% and WD=118-125 pcf? In my experience even when the nuke gauge moisture content reading is off, the moisture density relationships (i.e. WD=DD(1+M)) still hold true. If the gauge was reading an artificially low moisture content, then the dry density would be artificially high, not low as you have indicated.

Generally, your method for determining the field correlation seems reasonable. Also, your proctor value seems high for flyash.

 

[VAD]

The problem of moisture content being erroneous is not necessarily with flyash etc, it can happen as a result of malfunctioning of the gauge and requires to be checked by the supplier/anufacturer. I presume that you checked the guage in a controlled environment re moisture on sand or clay soil. If the gauge is malfunctioning then erroneous results would be obtained regardless. I cannot remember exactly the reason but when it was sent back to the supplier, it was fixed. Something to do with moisture and the electronics, I think. Not sure

 

[Scienceguy57]

I'm sorry, I keep calling the material fly ash when it was actually bottom ash and they are not the same thing at all.

casimmons, sorry I remembered the DDs wrong. You're right the moisture density relationships still worked. Let me try and elaborate:
-For a typical failing test the numbers in the field were WDs ~110, M%s were 8-10, and DDs 100-101. After taking a sample back and drying it out for 24 hours the true M%s were 15-20% (field M%s were low by 7-10%) with DDs 91-95. The test failed on the spec. of 95% +/- 2% moisture of the standard proctor of 105.9 at 19.7%M.
-For a passing test the WDs were typically 118-125. For a WD of 118 the M%s were 8-10%, DDs of 107-109. After drying out a sample the M%s were again 15-20% and the DDs were 98-102.

VAD, the gauges weren't the problem. We used two Troxler 3440s that are only 5 years old I believe. We also had to do a sand cone every tenth density and the WDs on those usually matched within a few tenths of a p.c.f. with the WDs that the gauge gave us. We were also testing clay material on-site and doing sand cones on that and everything matched fine with the clay. Our Pueblo office was also using an older Troxler that only gave you WD and M% and that one matched the sand cones as well.

captainplasma, that's what you would think, the extra hydrocarbons left over from the coal burning should raise the M%s, but that's not the case. The M%s were typically 7-10% lower in the field then what the true M%s were after we took a sample back and dried it out. Again, the problem is that fly ash and bottom ash aren't the same thing "Bottom Ash: The most common type of coal-burning furnace in the electric utility industry is the dry, bottom pulverized coal boiler. When pulverized coal is burned in a dry, bottom boiler, about 80 percent of the unburned material or ash is entrained in the flue gas and is captured and recovered as fly ash. The remaining 20 percent of the ash is dry bottom ash, a dark gray, granular, porous, predominantly sand size minus 12.7mm (½ in) material that is collected in a water-filled hopper at the bottom of the furnace." That's from u

http://www.tfhrc.gov/hnr20/recycle/waste/cbabs1.htm

The material they were using was bottom ash, I was wrong and thought that this had fly ash in it but it does not.
I posted above according to Troxler "A false low reading can also occur but is less common (causes include: high salt or iron oxide content or boron, lithium or cadmium present). A moisture offset will adjust for this problem." So one of these must be present in the bottom ash therefore lowering our M%s in the field. And again the optimum DD was 105.9.

The main problem in the field was that there was no way to determine the true DD on the spot. Even with a moisture correction of 7-10% there was still a possibility that we could be passing a failing lift since we didn't know the true M% and true DD and percent compaction. On top of this, the M% offset varied from time to time. This was all a huge pain in the rear for me because the contractor wanted us to magically determine right on the spot if the lift passed or not so they could cover it up. They didn’t want to understand how it was impossible to determine there on the spot if the lift passed or not and they wanted us to do something about it. However, if the lift turned out to fail 24 hours later then there was a problem and no one would want the blame. To be honest this all should have been figured out before backfill started with the bottom ash and density testing was started. I was only on this site for a month and a half and I came after backfill had started with the bottom ash. Maybe this is a unique problem and has never been seen before and that’s why the process of backfilling with the bottom ash and testing it wasn’t figured out fully before backfill started. I’m not sure.

 

[BigH]

You should have been able to pretty accurately determine the relationship (after a week or so) of oven dry vs your nuke moisture content. I would then have used this for the initial screening. Did you try the speedy moisture content determination to check on the moisture content? (and a proper calibration of the speedy with oven dried). Having this as well can give a bit more confidence.

 

[Scienceguy57]

Bigh, what do you mean by speedy Moist. determ.? Do you mean like a burn off in the field with dry gas or something, or a microwave burn-off? We were reluctant to try a micro. The problem was that the moisture correction varied from time to time and also that it was a 4% range. The moisture could fall anywhere in that 7-10% moisture correction. So if we tried a 7% correction in the field but it turned out to be more like 10% that day we could be wrong or if we tried a 10%... The stuff varied wildly.

 

[VAD]

Moisture Content obtained from a Speedy Moisture tester. Piece of equipment that takes a charge of carbon and known weight of soil in a cap and connected to a metal container with a dial- sort of milk bottle shaped. Certain method of shaking , tapping etc and dial gives the moisture content. Calibration chart is required for differnt soil types and oven dried moisture, though sometimes used without to get a feel for moisture. Used in the old days. You have to be as old as BigH and myself to remember.

I think you can get the full picture from doing a search on the internet or Big H can provide more description.

 

[Scienceguy57]

Ah, Humboldt sells them, kind of pricey. Don't know why our project manager wouldn't have used one of those, he's been in the industry for probably 30 years. I'll have to check with him when I see him next. Thanks VAD. What's the accuracy like on them as compared to an overnight oven sample?

 

[VAD]

Scienceguy57:

Generally for sandy and low plasticity soils within 1 percent, if I remember. Worked well on soil cement stabilized mixes. Did not use on heavy clay soils

 

[oldestguy]

The Speedy was developed for the foundry industry and has been around for years.

Maybe someone other source than a soils supply house is less costly.

It works well for all soils, but only capable of those without gravel, etc., due to small sample used. So it tests the finer fraction then.

 

[1Kat1]

I know in North country, fly ash and any type of ash additive is the last thing they would want to use. They want to stay low on the use, due to it's toxicity. Any rate, sand alone is a very dry material, that is hard to acheive specified denisities on. You have to have the right moisture. The laborer needs a hose, spraying the material, while is being compacted. And the field tech needs to be right behind the compactor taking the density tests. Even having a sand cone, using the alcohol and pan method, would have shown a failng result. In cases like this, you can either be a hard head and insist on the hose down, or you can use your best knowledgeable judgement. Does the soil look and feel tight? Is there cracks in the compacted soil? Is small stone bouncing up when it is being compacted? If the answer was yes, than it achieved the specified compaction, but just may have been lacking in moisture content. I like the MC3 gauge, as you can put a bias in it, to help determine whether the material is lacking, or over in moisture.

 

[dirtsqueezer]

I didn't read the rest of the thread, but I worked on the Big Hanaford Gas Power Plant in Washington State doing the exact same thing, using bottom ash as backfill for about a year.

We did take regular samples from the fill, and with every new test came a brand new proctor value. Most were in a range of 5-15 pcf. What we found was a small variaince of the pea-sized material (most of it was sand-sized). They were little brown/black kernels that I thought were silica, but probably contained some H+. I did try a nuke gauge on the stuff one time, but my reading was so strange that I just discounted it. It seems encouraging that your numbers were all minus and not plus, and that they were all within a solid range of the material. My sand-cone numbers were all air-tight as far as compaction goes. I remember for retests how the compaction matched the compactive effort almost perfectly. A little cumbersome, but they certainly did do the trick.

The really weird thing was- and I don't know how they discovered this, but for some reason if you had a continuous flow of water over the bottom ash while you were compacting it, and I'm talking about a firehose basically spraying the bottom of the drum, your densities would shoot up to the 95%+ range with something like 2 passes. The stuff was hard to scrape out.

Anyway, a little lore from my time and bottom ash. It was a little fun seeing the material vary, but unfortunately I'm now the expert on sandcones. Fun, fun!