Thermal conductivity/resistivity of concrete used for the encasement of underground electrical ducts

[wroggent]

I am trying to chose correct cable sizes for a new underground electrical duct bank. I am using the 2011 NEC to determine the correct conductor ampacities. Since this is for 480V, I think could get away with using Table 310.15(B)(16) and be done with it, but I am not comfortable with this approach so I have opted to use the tables in Annex B. A mechanical engineer in my organization told me that we always encase underground ducts in concrete (as opposed to back filling with only earth); I would like to ask him what the thermal conductivity of the concrete that is typically used is - but he has since disappeared into a black hole and cannot be reached.

I decided to try and make a guess at what the thermal conductivity might be. This link:

http://www.engineeringtoolbox.com/thermal-conductivity-d_429.html

suggests that values of thermal resistivity (in the same units the NEC uses) ranges from 1000 for 'lightweight' concrete to 55.55 for 'dense' concrete. The NEC uses 55 as a typical value for concrete. Based on this massive range in thermal resistance, I have no confidence in guessing what the actual resistance is for the material to be used.

Do any of you use a particular value of thermal resistivity for concrete in the absence of actual data?

The mechanical engineer did tell me something about this material not being 'real' or 'structural' concrete or something along those lines; he referred to it as a 'concrete slurry'. Does anyone know if this has any relation to the density?

I looked through the following threads that had similar questions before posting this and didn't find any useful:

http://www.eng-tips.com/viewthread.cfm?qid=119019

http://www.eng-tips.com/viewthread.cfm?qid=117975

http://www.eng-tips.com/viewthread.cfm?qid=307120

I have another somewhat unrelated question:
When and how would you decide to derate ampacity in an underground ductbank due to ambient temperature? I have assumed that the phrase "ambient earth temperature" as it is used in Annex B of the NEC refers to the temperature of the soil at the depth where the duct bank is installed. Once upon a time a tried to find data for this but I wasn't very successful. I think I had found surface temperature data from an agricultural resource, but I was doubtful that it correlated to soil temperature at 24 inches, which is the depth of my duct bank, so I disregarded it. What ambient temp would you use for the central San Joaquin Valley in Ca for a duct at 24 inches below the surface? All I know about this area is that the average max annual temperature is around 100F.

 

[Compositepro]

I can't answer your questions but in a chemical plant I worked in I worked in 35 years ago, the practice was to dye the concrete red for electrical marking.

 

[dpc]

Unless things have changed, the NEC does not require any derating for mutual heating for circuits below 1000 V. This is only required at higher voltages.

The NEC ampacity tables are conservative and have been used for decades. I would not be concerned about derating due to concrete encasement. You can use 310.15(B)(16).

Regarding ambient, I doubt the soil/concrete temperature will be an issue. You would be looking at a very large thermal constant, so it is really the average daily temperature that would be a concern.

I think the original Neher-McGrath paper had some typical thermal constants - but I can't find my copy of that right now.

 

[larrym]

Average temp at 2 ft is 5-10 deg below surface temp. Solar heating could make surface temp higher than air temp during the day

 

[7anoter4]

First of all I think NEC allowable ampacity is very conservative –as dpc remarked.
In NEC Annex B is recommended Rho= 55 [oC.cm/w].Neher & Mc Grath paper recommends [Table VI Thermal Resistivity of Various Materials - C cm/watt] 85 for concrete. IEC 60287-2-1 recommends [Table 1 – Thermal resistivities of materials] 1[k.m/w=100 C.cm/w].
As dpc said if you will try to follow Neher method using NEC Rho 55 or IEC method using the same Rho you’ll never get the NEC allowable ampacity [calculated it will be more any way]

 

[wroggent]

dpc/7anoter4,

Are you saying you would never use Annex B for 480V circuits as you consider Table 310.15(B)(16) to be conservative already?

If yes, do you maintain this practice even for large duct bank arrangements, e.g., 6 or more sets of 3 conductors (3-1/C per conduit times 6 conduits)?

Also if yes, what about for banks that are deeper in the earth? Annex B mentions derating 6% per foot...(this one does seem a bit overboard to me)

To elaborate on my situation, I have the goal of installing two 800A feeders and I'm trying to decide between using 2 500MCM conductors per phase with 2 conduits (1 conductor of each phase per conduit), or 3 500MCM conductors per phase with 3 conduits. I think Table 310.15(B)(16) has 75 degree C 500MCM at 380A so with two in parallel they can be used with an 800A breaker. Whether there are 4 or 6 conduits in total, they will be installed together in the same bank. What are your thoughts on this? I don't want these cables burning up someday down the road, but at the same time these will be pretty long runs so cost is a big factor here. If it weren't for the existing equipment that is there, I would be using 12.47kV instead of 480V.(that's another story though)

I am relatively young, but IIRC I read something that said Annex B has been included and excluded as an official part of the code in multiple code cycles before my time. Are those calculated values (in Annex B) in an annex (and not the code) at this time because there is some consensus that they are stupidly conservative?

 

[resqcapt19]

The rules that are now in Annex B were in Article 310 for the 1987 code, but with an effective date of 1/1/1990. When the 1990 code code came out these rules were moved from Article 310 to Annex B, so they were never actual code text. Most of the proposals and comments that resulted in the information being moved to the Annex were based on the very conservative load calculations found in Article 220, not conservative ampacities.

There were comments that said if the load calculations were done using methods other than Article 220 or if there are multiple ducts in a duct bank the use of the ampacities from the Article 310 tables could result in conductor failure.

If you want to, you can read the proposals and comments on this issue by going to:

http://www.nfpa.org/codes-and-standards/document-information-pages?mode=code&code=70&tab=docinfo

Under the tabs, select the year of the code, and then select "archived revision information".

 

[7anoter4]

It is indeed a very interesting issue. See [for instance for more information]:

http://neher-mcgrath.com/Heating_Calculations.pdf

http://www.neher-mcgrath.com/41/neher-mcgrath_lowrho_hs.html

Seattle City Light
CONSTRUCTION STANDARD
Requirements for Duct Banks in the Public Right-of-Way

http://www.seattle.gov/light/engstd/docs2/0222.02_131121084951JEFJMY.pdf

“The encasement shall be red High Strength Fluidized Thermal Backfill (HSFTB). HFSTB is a concrete mix and is the only allowed material for encasement.”
This HSFTB concrete presents a Thermal Resistivity of 50-65 oC cm/w.
However if Low Strength is used the Rho will be 85.
If you could be more specific [Earth RHO, depth of the ducts, duct bank configuration] ampacity approximation could be verified.
If the duct bank depth will be 3 feet [around] in my opinion the sun radiation is not important. But the Load Factor will be.

 

[wroggent]

Interesting read resqcapt19. It looks like people are quite divided on this topic. I can understand that the NEC is not intended to be a design guide, and that including tables for the many possible varieties in installations would be cumbersome, but I also believe that allowing Table 310.15(B)(16) to be used in every scenario does not achieve the NEC's goal of ensuring a safe installation. I also don't like the idea that how a load is calculated would have any bearing on cable ampacity - if I pass exactly X amount of amps through a wire and X amps is the ampacity stated in the NEC table, then I expect the cable to not be damaged - regardless of what the nameplate on the load might say. As stated in the CMP's discussion, one error should not be used to compensate for another.

For those who would like to read what resqcapt19 was referring to, check out pages 213 to 219 (as shown on the pages) in:

http://www.nfpa.org/Assets/files/AboutTheCodes/70/NEC-TCRA-1989-Articles 90 to 310.pdf

Going back to the original question though - I talked to my concrete contractor and he said I can specify what thermal resistance I want. How convenient.

 

[jghrist]

Rating of Electric Power Cables, George J. Anders, IEEE Press, 1997 has a Table 9.1 (based on IEC 287-2-1, 1994) that gives 1.0 K·m/W for concrete.

 

[dpc]

I talked to my concrete contractor and he said I can specify what thermal resistance I want. How convenient.

He probably said that because there won't really be much you can do about it later.

 

[cranky108]

I have a Civil engineer that does that for me.

 

[Gragas]

You have to be careful using that annex.

The true answer is to use software similar to ETAP or similar to create a model in order to accurately model the actual configuration you are going to use.

Otherwise, I can tell you that all of the derates that occur if you followe the descriptions to a T will be VERY conservative.