Derating of underground cable

[rockman7892]

I am working on building a new plant where most the majority of the cable is run underground. This includes both 4.16kV power distribution cable, as well as 480V power distributin and branch circuit cable.

The engineers for the project have severely de-rated the cable and it has led to numerous problems with cable being too large form termination on breakers, motors, etc... This has led to many arguments in the field as well as to who was responsible for solving and solving these problems (Strange contract situation)

I understand the practice and theory behind the need to de-rate cables that are run underground and in close proximity to other cables for heat dissipation. I know there there is the McGrath model and vaious software programs such as Amp Calc to assist with such calculations.

I was wondering if there have ever been many (if any) documented underground cable failures as a result of failing to derate cables that have pushed or are pushing this practice. Are there any cases that back up the requirement for this to be an absolute necessity when performing cable sizing? In my case there are gross oversizing the in some cases seems unessescary and like I said is leading to problems.

 

[magoo2]

What total amps at 480 V are you trying to deliver?

What size conductors are used?

Are they copper?

How many sets are proposed?

I've used CYME's cable ampacity program before and have found problems when the installation had one or two sets less that what they recommend.

 

[rockman7892]

magoo2

There are many different cases throughout the plant with many different loads.

One particular example if a parallel set of 500MCM's (copper) for a 200hp motor at 480V. Obviously these 500 MCM's are difficult to terminate at both the starter and the motor.

This is a typical example of a problem that we run into as a reult of gross oversizing.

 

[EEJaime]

I guarantee you that depending on the quantity of cable runs and the soil conditions there are VERY good reasons to derate cables.

We had a plastics plant where in we had routed a conduit bank for several 1600A and 2000A feeders at 480V,3ph in conduit banks. We overs sized these feeds based on calculations and evidence that the local soil was hot. We had two failures within a year of the commissioning. It turns out the Contractor down sized the feeders to "code minimums" and this was not caught by the Inspector.

After much legal wrangling and finger pointing, we had a testing and forensic company come in to confirm that due to the difference in sizing, (and the fact that the contractor also eliminated spacers, concrete, etc..., which decreased the spacing between the conduit.

The Contractor was stuck with the costs.

Thanks,
EEJaime

 

[jghrist]

Is the oversizing a result of derating or is it for voltage drop concerns? Are these long circuits? If the oversizing is for voltage drop, you might be able to reduce the size before terminating. This might mean an additional junction box.

 

[davidbeach]

Even if the over sizing is heat reasons in the underground run, you can come out of the underground run and splice smaller conductors to the larger conductors.

If you are convinced that it is gross over sizing, write a letter to the engineer absolving him of all liability and then install the sizes you think are "right". Hang onto the original plans though so you know what to replace the failed conductor with in the future.

 

[EEJaime]

Whether the upsizing is due to voltage drop or to concerns of over-heating, if the minimum feeder cross-sectional area that the feeder breaker can protect is maintained, then the conductors may be connected with reducers at the terminals, (do not allow reduction of the stranding). This must be ok'd by the AHJ, but most we've dealt with are ok with this.

Good luck

 

[dpc]

For 480 V duct banks, the NEC does not specifically mandate any derating based on mutual heating. For 4160 V, it does.

But as mentioned, the laws of thermodynamics apply to 480 V cable as well as MV, so there may be good reason for derating the 480 V feeder as well due to local conditions - we're not in a position to second-guess that.

Another point to consider, if you are using 90 deg C insulation, you can use the 90 deg rating as the starting point for the ampacity derating of the cable.



"An 'expert' is someone who has made every possible mistake in a very narrow field of study." -- Edward Teller

 

[vnn]

We selects the cable based on the following criteria:

- short circuit withstand.
- Full load current. (de-rating factor is included here)
- voltage drop.

In most cases of low voltage cable installation to a motor, the numbers of cable and cable sizes are decided by the voltage drop, especially when we start the motor, not the full load current.

For the case, you mentioned above, one solution can be applied as follows:
At the motor: You should provide one junction box, near by the motor. The cable from the junction box to the motor is very short and we need a cable 3x185mm2 only that can transfer a full load current around 350A.
At the MCC, I think it is easier to install your cables with an addition of a plate (or bus bar) to the outgoing terminal of MCC.

Best regard,

 

[rockman7892]

Thanks for all of the responses guys. The cables are mostly derated for heating but I'm sure voltage drop is part of the oversizing as well.

As the client represenative, I am not looking to challenge the engineering of these sizes or pursue smaller cable sizes opposing what was recommended however I am simply trying to get a feel for how often this gross oversizing occors elsewhere and if it has led to similar problems. Obviously with the larger cable sizes this will add cost to the project so I'm interested to hear any thoughts on that aspect as well.

It sounds like as EEJamie points out that there are documented cases out there of failures as a result of not derating for underground distribution. I guess the cost of these failures as well as any legal wrangling involved will for even just one failure will justify the necessity to deal with the problems that occur due to these gross oversizing. Are there any other large scale failures that anyone is aware of. Just curious?

We have done what many have recommended as far as terminating these oversized cables. We are having to put J-Boxes everywhere and splice to a smaller cable. I am having engineering review each termaination case and sign off on the consolidated splice size and method in order to have this all documented. I am not allowing the contractor to size any of these spliees, so therefore it must all be revied by engineering. This always leads to an argument with the contractor as to who is going to supply the J-Boxes.


Dpc

I thought that the diagrams in figure B310.2 in the NEC as well as the corosponding table B.310.7 dealt with mutual heating in 480V ductbanks. Can you please explain your comment as well regarding the 90deg C tem derrating.

So in summary I am following the sizes and deratting specified by engineering however just wanted to hear others opinions or comments on similar problems this has led to for construction. Personally I do not think the engineering firm did a good job of verifying or specifiying how these oversized cables would terminate on supplied equipment which has lead to all of these problems.

 

[dpc]

Firstly, Annex B is NOT part of the NEC. Secondly, the use of this table is NOT required by the NEC - it is provided as guidance when ampacities are calculated under "engineering supervision".

For medium-voltage circuits, the duct bank information and de-rating is part of the NEC and there is no choice.

There is a fairly complicated NEC history with derating of low voltage duct banks. Suffice to say at this point, it is not **required** by the NEC.

As for the 90 deg C rating - in most cases, the 90 deg ampacity of a 600 V conductor cannot be applied since the terminations and equipment heat dissipation at both ends are based on the use of conductors rated for 75 deg C insulation. So even though a conductor may have 90 deg C insulation, in general it can only be applied up to the 75 deg C (sometimes only 60 C) rating. BUT... when derating for heating effects, high ambient, etc it is permissible to start with the 90 deg C ampacity as the basis for de-rating.

"An 'expert' is someone who has made every possible mistake in a very narrow field of study." -- Edward Teller

 

[7anoter4]

Following jghrist remarks: as 200 hp induction motor at 480 V will draw only 250 A [FLA] than according to NEC- for worst case-only 250 MCM will be enough. But, if the motor will be direct connected, the start current will be 5-8 times the FLA. EPRI proposed 16.3% starting voltage drop up to 200 hp. Then for 16% drop voltage at start one needs two 3*500 MCM cables for a distance of 400-600 ft.
I encountered this situation mainly with European Induction Motors where the maximum cable entering was very small with respect the US made motors.
So I used Burndy Mole Connector in a manhole or pull box: two 500 MCM incoming one 4/0 outgoing will be fair enough. See:

http://portal.fciconnect.com/res/en/pdffiles/brochures/I-Network_Underground.pdf

There are other connectors for direct burying. [UL486D Listed for direct burial]:

http://portal.fciconnect.com/res/en/pdffiles/brochures/BURNDY+UNITAP+-+2007.pdf

 

[7anoter4]

As an answer to your second question if it is a "documented underground cable failures as a result of failing to derate cables that have pushed or are pushing this practice".
Elevated temperature or overload can lead to chemical breakdown of the insulation, producing thermal runaway that can decrease insulation resistance and increase current leakage and failure- shortening life expectancy- and create a fire hazard by shortcircuit.
Any way, it happens usually in buildings and less in underground feeders-except in manholes of course. See for instance:

http://resources.metapress.com/pdf-preview.axd?code=l9ukla0ew9hdw5j5&size=largest

 

[Rodmcm]

The whole city of Auckland New Zealand lost power due to the failure of two 110KV underground cables. The essential reason was that the bedding was in sand as was fashionable at that time and the thermal resistivity was high in areas.As most engineers who followed this found there is more to burying cables than meets the eye, sespecially when they cross at right angles.

http://homepage.mac.com/bryanleyland/.Public/Power systems/AucklandCBDPowerFailure.pdf

 

[FreddyNurk]

On a relatively recent mining project we encountered issues with fitment of cables into terminal boxes and MCCs due to sizing of cables for voltage drop. The installation was in Australia, so NEC didn't apply, but we had managed to get around the issue through appropriate use of junction boxes at the motors, and use of relevant sizes of terminals at the MCCs. Difficulties were encountered with use of junction boxes in a solvent extraction area (was classed as hazardous, so we couldn't just use the same j-boxes as the rest of the plant) though due to an unusual contract set up for the construction of the site we didn't have the arguments over who was going to pay for the j-boxes and upgrades.

What was also difficult was a senior design engineer who insisted that we cut down the strands and insulation to allow the oversized cables to fit into the motor terminal boxes. As our upstream protection adequately covered the loads of the motors, we insisted on the junction box solution.