Desiccation of clay under elevated temperature tank

[monopoly]

Hi all,
We design foundations of the oil tanks (diameters vary from 30-55m, height 12-13m) with concrete ring walls (the ring wall is required for uplift resisstance). Inside the ring wall the bottom of the tank is rested on the 80mm minus run gravel compacted to 100% SPMDD, with thickness of 1m. under the gravel layer is the engneered fill of about 1.5m thick, then the stiff to very stiff native clay. And we intended to install the high temperature geotextile for secondary liner under the tank.
The problem now is the temperature of the oil inside the tank is up to 90 degree celsius. Geotechnical engineer for the project said that the 2.5m of gravel and engineered fill above the clay is enough for the dissipated of the temperature when it transered to clay. but still warry about the desiccation settlelemnt of clay down bellow.
Any document or reference to check that problem. or any experience with that kind of design?

 

[SlideRuleEra]

Since this is a heat transfer problem, see thread486-74627 for a good discussion of the thermal conductance of soil.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[JoeTank]

I've "heard" from others that a detailed thermodynamic model of a large asphalt AST revealed that the bulk temperatures of the product (250F+) penetrated to a depth of approx one-half the tank diameter. This was the temperature along the vertical centerline of the tank. I have no info on soil properties used in the evaluation. Similar statements have been made by others whom have done simialr analyses. I think they stated that the steady state conditions occurred about 10 to 20 months are being placed into service. I am not aware of any field verification of the results, but "steaming" has been observed arouns a number of asphalt tanks at a number of sites.

Joe Tank

 

[oldestguy]

Interesting situation.

I'd accept the condition of heated soil under the tank. I'd then look at the various moisture transfer mechanisms to develop a procedure for negating the moisture loss situation.

This reminds me of how I have successfully prevented drying of soil (and shrinkage) due to fast growing trees near buildings by a simple means of wetting the soil and leaving the trees there.. That may not work here, however.

For instance if vapor transfer from warm to cold is present, can you reverse the temperature gradient some place as a barrier?

You know if you dry clay soil far enuff, it won't shrink any more. Would pre-drying be in the picture?

 

[escrowe]

Given the stiff to very stiff nature of the clay, shrinkage due to dessication may not be significant.

I would run a few Atterberg shrinkage limit tests.
Potential shrinkage volume could be estimated from the difference
between the natural (in-situ) moisture content and the shrinkage limit.

I think an estimate of total settlement under these conditions is complicated by three factors:

1) The zone of influence for consolidation foundation settlement is likely larger than the zone of expected temperature increase.

2) Potential dessication settlement may be more or less than consolidation settlement in the same location.

3) The rate of dessication may be greater or less than the rate of consolidation.

I think a conservative approach would be to simply add the expected dessication related settlement to the expected consolidation settlement.

Otherwise, crank up those diff-eqs.

 

[aeoliantexan]

I disagree with your geotech and agree with joetank that the soil heating will go very deep. This is based on experiences with refrigerated product tanks such as anhydrous ammonia and with freezer warehouses. Even with a good layer of styrofoam or foamglass insulation under the floor, if the subgrade heating system failed, freezing penetrated the soil to 20 - 30 feet in a few years. Heating requires a much smaller quantity of heat than freezing (no phase change), so it should progress much faster. You need insulation and a heat sink such as water coils.

 

[monopoly]

Thanks for all the info. now the problem come to decide if we should use the high temperature geotextile as a secondary liner of the tank , which can resist more than 100 degree celcius for long period, or use the normal geotextile , as the manufacturer say can only resist not more than 60 celsius degree. From engineering wise, I am sure that we have to use the high temperture geotextile, but the owner complain it is expensive and seeking for alternative cheaper solution. and we engineer have to prove to them.
The geotextile would be install about 1.5m (5') below the tank bottom, and on top is the structural gravel fill minus 40mm. The concern only if below the 5 feet depep of fill, should the temperature drop from 90 degree to below 60. I did some literature review but have yet found any info about earth fill properties to pratically estimate. any info about the references would be really appreciate.

 

[aeoliantexan]

I know there are programs to model heat flow in soil, including the axisymetric tank problem, and there is probably pretty good information on thermal properties of soil. Some very old but widely used data on soil conductivity were developed by Miles Kersten in the 40's and 50' and published in Highway Research Board proceedings and bulletins. Conductivity of gravel has probably been studied by Cold Regions Research Laboratory.

Heat flow is similar to water flow in soil. Imagine a flow net with flow paths beginning at the tank floor and ending at the ground surface outside the tank. If the heat source were a long strip, the problem becomes 2-dimensional and easier to imagine. The flow paths are then semicircular, starting straight down (at right angles to the boundary) and ending straight up. The longest path is from the center, and its length is 3.14 times the width of the strip. Very crudely, then, for a tank, the longest path is 3.14 R in length, or 47 M for a 30M diameter tank. If the average annual air temperature outside is 15 degrees C, the average gradient is (90 - 15)/47 or 1.6 degrees C per meter. The temperature at the bottom of the gravel fill will be roughly 87 degrees C. Correcting this to radial flow will change the answer considerably, but surely the temperature will not be below 80 degrees at the geotextile.