Grounding of two wire circuits

[mivey]

I like bidding against those who go beyond the Code.

I up-sell against those who do not.

 

[kwired]

Read case one here:
http://www.nfpa.org/assets/files//PDF/Proceedings/Sleights_paper.pdf

Not an conviction against of BX per se but certainly worth considering. Probably more damning is the fact that the NEC started requiring the bonding strip in 1959 (I think).

Now I'm curious. Could one of you guys that have the old NEC code books look at them and see if the old AC cable without the bonding strip was approved as a grounding path method or if the addition of the bonding strip make it an approved grounding path method? I really don't know what the rule was back then.

Maybe it was a FPE breaker

Sure looks to me like the testing of the breaker indicates that the breaker was likely defective to begin with. 16 seconds to open a 500 percent load, 5 minutes of load @ 200 percent did not trip it.

I won't argue that there would have been higher fault current with newer AC cable and this may have helped trip the breaker sooner, but if there was a good breaker protecting this circuit the fire likely does not start either.

 

[mivey]

Maybe it was a FPE breaker

Sure looks to me like the testing of the breaker indicates that the breaker was likely defective to begin with. 16 seconds to open a 500 percent load, 5 minutes of load @ 200 percent did not trip it.

I won't argue that there would have been higher fault current with newer AC cable and this may have helped trip the breaker sooner, but if there was a good breaker protecting this circuit the fire likely does not start either.

From IEEE Std 142-1991 (Green Book)

It should be noted that the values listed in Table 4 for Greenfield (flexible metal conduit) and BX without ground (AC cable) are excessive for the armor to be used as an equipment-grounding conductor. Current standards require that BX incorporate an internal bonding strip in contact with the metal armor over the full length. The NEC, Reference [1], requires a separate equipment grounding conductor in Greenfield except for very short sections.

It is significant to note the sharp decrease in voltage drop when an internal grounding conductor is added in parallel with the conduit. In addition, the line-to-ground fault impedance will be reduced. Thus, the use of a metallic conduit raceway as a grounding conductor, supplemented by an equipment-grounding conductor within the conduit, achieves both minimum ground fault impedance (see Reference [9]) and minimum shock-hazard voltage.

The NEC [1] provides no specific limitations on the length of tubular metal raceway or cable armor that may be used as an equipment-grounding conductor. It is apparent, though, that excessive length can result in an impedance that will limit the ability of the circuit overcurrent device to clear a ground fault as well as cause a hazardous voltage on the raceway or cable armor surface. Reference [21] contains some examples of maximum lengths as a function of overcurrent device rating {The Soares table example shows the maximum length for EMT, IMC, & RMC to be between 200 & 300 ft for various conduits & breaker sizes}. The resulting reduction of impedance when an internal grounding conductor is added in parallel with a metal raceway will permit feeder distance to be increased by up to 1.7 times the maximum feeder length without the internal conductor.

...

[9] Gienger, J. A., Davidson, O. C., and Brendell, R. W. ?Determination of Ground-Fault Current on Common AC Grounded-Neutral Systems in Standard Steel or Aluminum Conduit,? AIEE Transactions on Applications and Industry, vol. 79, May 1960, pp. 84?90.

[21] Soares, E. C. ?Grounding Electrical Distribution Systems for Safety,? Wayne, NJ: March Publishing Company, 1966.

 

[kwired]

So the real question is whether old AC cable is better off being made non continuous somehow or is whatever grounding effectiveness it has worth insuring it is as good as it can get? By this I primarily mean making sure connectors are installed properly and are tight.

Is enough impedance introduced to prevent overcurrent devices from operating, or will enough current flow to operate device just will take a little more time?

 

[al hildenbrand]

. . . Could one of you guys that have the old NEC code books look at them and see if the old AC cable without the bonding strip was approved as a grounding path method or if the addition of the bonding strip make it an approved grounding path method? . . .

Click HERE.

 

[al hildenbrand]

Could one of you guys that have the old NEC code books look at them and see if the old AC cable without the bonding strip was approved as a grounding path method or if the addition of the bonding strip make it an approved grounding path method? I really don't know what the rule was back then.

New in the code in 1959:

1959 NEC
Article 334 - Armored Cable
334-2. General. Armored cable for 600 volts or less shall conform to the following:
(a) Armor. All types of armored cables except Types ACL and ACV, in all sizes, shall have an internal bonding strip of either copper or aluminum in intimate contact with the armor for its entire length.

ACL is lead covered AC; and ACV is a dry location, exposed only, industrial or commercial only, not smaller than #4 type of AC.

So, garden variety AC used in branch circuits had the drain wire, by Code, from whenever the local jurisdiction adopted the 1959 NEC.

Second. From the 1953 NEC:

1953 NEC
Article 250. Grounding
Grounding Conductors
2591. Material. The material for the grounding conductor shall be as follows:
. . .
b. For Conductor Enclosures and Equipment Only. The grounding conductor for equipment and for conduit and other metal raceways or enclosures for conductors, may be a conductor of copper or other corrosion-resistant material, stranded or solid, insulated or bare, a bus-bar or a rigid conduit, steel pipe, electrical metallic tubing or the armor of armored cable, except that under conditions favorable to corrosion a grounding conductor of copper or other corrosion-resistant material shall be used.

Last. From the 1937 NEC:

1937 NEC
Article 250--Grounding
2501. General. This article treats of protection in the use of electrical circuits and equipment by grounding. Insulation, isolation, and guarding are suitable alternatives under certain conditions.
. . .
Means of Grounding
2551. General. The path to ground from circuits, equipment, or conductor enclosures shall be permanent and continuous and shall have ample carrying capacity to conduct safely any currents liable to be imposed on it, and shall have impedance sufficiently low to limit the potential above ground, and to facilitate the operation of the overcurrent devices in the circuit.
. . .
2559. Fixed Equipment. Metal boxes, cabinets and fittings, or non-current-carrying metal parts of other fixed equipment, if metallically connected to grounded cable armor or metal raceway, are considered to be grounded by such connection.
2560. Portable Equipment. Non-current-carrying metal parts of portable equipment may be grounded in any one of the following ways:
. . .
b. By means of a grounding conductor run with the circuit conductors in wire assemblies, provided an approved multi-prong plug or equivalent is used, one prong for the purpose of connecting such grounding conductor to the grounded metal raceway or cable armor. This conductor may be uninsulated but if an individual covering is provided for this conductor, it shall be finished to show a green color.

 

[tryinghard]

...From the 1937 NEC:...2559. Fixed Equipment. Metal boxes, cabinets and fittings, or non-current-carrying metal parts of other fixed equipment, if metallically connected to grounded cable armor or metal raceway, are considered to be grounded by such connection.
2560. Portable Equipment. Non-current-carrying metal parts of portable equipment may be grounded in any one of the following ways:...b. By means of a grounding conductor run with the circuit conductors in wire assemblies, provided an approved multi-prong plug or equivalent is used, one prong for the purpose of connecting such grounding conductor to the grounded metal raceway or cable armor.

Al, this doesn't surprise me. If a particular installation no longer meets safe enough standards to new it should be the prerogative of the journeyman to change it even though originally it's compliant - but the choice to leave it could be applied as well if judged safe enough. Kinda like older cars that did not have seat belts legal yes but situationally could be upgraded to include them, and at least a good journeyman should determine this if electrical related.

 

[al hildenbrand]

. . . it should be the prerogative of the journeyman to change it even though originally it's compliant - but the choice to leave it could be applied as well if judged safe enough. . .

It's my opinion that the electrical professional can, and should, inform the occupancy owner of hazards that can be observed, but, in this case, the case of pre-1959 armored cable, making a blanket condemnation is unfounded without opening up boxes and walls and ceilings (if no attic) to examine the mechanical assembly.

Just because non-bond wire BX is present doesn't mean there was a lack of good workmanship at every stage of the installation up to the present moment.

As a present moment observer, I can infer some possibilities and I will so inform the owner. The owner decides.

In the absence of "obvious hazards" (a term of art used in my geographic area) I am not in a legally defensible position to tell the owner s/he must replace the BX.

 

[mivey]

Click HERE.

Thanks for the research. I was unsure if they had allowed BX for grounding in the past because they did not know of it's inadequacies as an EGC or if they just did not allow it. It appears it was allowed because they were still on a learning curve about grounding as well as the inadequacies of armored cable without a bonding strip.

This piece you provided from the 1937 NEC just underscores the fact that they intended for it to be a good ground but they just did not realize it was not:

Means of Grounding
2551. General. The path to ground from circuits, equipment, or conductor enclosures shall be permanent and continuous and shall have ample carrying capacity to conduct safely any currents liable to be imposed on it, and shall have impedance sufficiently low to limit the potential above ground, and to facilitate the operation of the overcurrent devices in the circuit.

Thanks again for the info.

 

[mivey]

So the real question is whether old AC cable is better off being made non continuous somehow or is whatever grounding effectiveness it has worth insuring it is as good as it can get?

How about old AC cable should be put on a GFCI or maybe even an AFCI device.

By this I primarily mean making sure connectors are installed properly and are tight.

Won't really help.

Is enough impedance introduced to prevent overcurrent devices from operating, or will enough current flow to operate device just will take a little more time?

Most of these paths have been found to be woefully inadequate and do not even meet the UL specs that were in place at the time (some due to not understanding the performance difference in the DC resistance vs AC impedance). They looked good in the lab but were pitiful in the field. Some of the part of learning about the cable.

The process of learning is why we have a bonding strip now in AC cable and a length limit on unbonded flex.

 

[mivey]

Al, this doesn't surprise me. If a particular installation no longer meets safe enough standards to new it should be the prerogative of the journeyman to change it even though originally it's compliant - but the choice to leave it could be applied as well if judged safe enough. Kinda like older cars that did not have seat belts legal yes but situationally could be upgraded to include them, and at least a good journeyman should determine this if electrical related.

Find me the word "grandfathered" in the code book.

From 410.42 A & B:

"Exposed metal parts shall be connected to an equipment grounding conductor"
&
"Luminaires directly wired or attached to outlets supplied by a wiring method that does not provide a ready means for grounding attachment to an equipment grounding conductor."

We are required to connect to an EGC but the existing system does not have anything that can qualify as an EGC. Unless we take one of the alternatives, we have to provide a good return path for the fault current.

 

[mivey]

It's my opinion that the electrical professional can, and should, inform the occupancy owner of hazards that can be observed, but, in this case, the case of pre-1959 armored cable, making a blanket condemnation is unfounded without opening up boxes and walls and ceilings (if no attic) to examine the mechanical assembly.

Just because non-bond wire BX is present doesn't mean there was a lack of good workmanship at every stage of the installation up to the present moment.

Workmanship is beside the point. The non-bond wire BX has been found to be inadequate. The times that is does perform well is the exception, not the norm. It is not about the workmanship, it is about the product itself.

As a present moment observer, I can infer some possibilities and I will so inform the owner. The owner decides.

And I think they will not have enough information to make a safe decision. If our industry barely understands it, how will a homeowner be expected to?

In the absence of "obvious hazards" (a term of art used in my geographic area) I am not in a legally defensible position to tell the owner s/he must replace the BX.

Not necessarily replace but provide additional protection like GFCI or even AFCI protection. That would be my minimal recommendation.

As for the "standard" recommendation, I would have to look at the condition of the wire, if there was older K&T in place, if there were BX that had been exposed to moisture, if insulation had been added over the BX, what modifications had been done, etc.

If it were my family, I would replace it or give it a thorough look-see because the potential for other things like a fault between outlets would bother me too much. Believe it or not, I would feel safer with K&T. I think the unbonded BX is that poor of a product. FWIW, the bonded armored cable is adequate (barely), but not that great of a product as far as I'm concerned. I use the MC cable with an insulated ground conductor instead.

 

[tryinghard]

Find me the word "grandfathered" in the code book.

From 410.42 A & B:


We are required to connect to an EGC but the existing system does not have anything that can qualify as an EGC. Unless we take one of the alternatives, we have to provide a good return path for the fault current.

You don't see this statement in my quote either. There are some hazards that require all past brought up to current and some not, this cable is not. It's not “grandfathered” as much as it's simply allowed to remain as is, installed correctly at its time of course. Its covering is the effective fault path, and if the professional installer judges differently they should correct accordingly, just that simple.

I think most often this stuff about effective fault path is too much for a home owner to grasp, so I'd go there carefully and slowly if at all.

 

[al hildenbrand]

Again, from post # 26.

It's way to easy to overthink this.

Get a multimeter out and measure the Ohms from the ceiling box to the Grounding Electrode System.

 

[mivey]

Again, from post # 26.

Again, from post #52:

It is not about the workmanship, it is about the product itself.

 

[kwired]

There is no doubt in my mind that the old cable without the bonding conductor is a poor grounding conductor, I may not have realized how poor.

Now after the last ten or so posts lets get back to the issue in the OP.

What would you do if you were asked to install the ceiling fan as described in the OP and found it was supplied by old AC cable without the bonding strip?

Depending on the house it could be expensive job. What if you find most of the house is wired with the stuff?

If you turn down the job because they don't want to spend money to do it right they will find someone else to do it.

If they decide not to put in the fan at all the AC cable is still there anyway with the same hazards.

 

[mivey]

There is no doubt in my mind that the old cable without the bonding conductor is a poor grounding conductor, I may not have realized how poor.

It may be worse than we realized.

I found an article written in 2009 by John Sleights titled "An Evaluation of Old Armored Cables in Building Wiring Systems".

For those interested, it is certainly worth the $34 download price here: http://www.springerlink.com/index/n8j7x34t705g5201.pdf

Abstract:

Armored cable has been used in building wiring systems since the early 1900s and remains an accepted wiring method today. The original product developed in the early 1900s and that which is commercially available today have significant differences in construction and performance. Likewise the demand and expectation of building wiring systems has changed. This paper explores some of those differences as well as concerns and hazards presented by the continued use of older constructions of armored cable. Testing was performed on samples of aged cables removed from service to determine impedance and temperature rise of the armor when conducting AC current. The results were compared to the DC resistance of the armor as well as the applicable Underwriters Laboratories standard. The performance of cable samples was also tested in construction configurations likely to be encountered in buildings. Results indicate that ground fault currents conducted via the armor of older cable constructions can result in fire and shock hazards where the impedance and circuit characteristics limit the fault current to below the trip level of the circuit overcurrent protective device. Temperature rise of the armor can easily exceed the rating of the conductor insulation contained therein and in some cases present a fire hazard to combustible material in contact with the cable.

There are 41 pages of discussion and a good read. Even back in the 1905 NEC, it was the intent that the armor be the equipment grounding conductor and that it was intended for this to trip the breaker during a ground fault before things overheated and to keep the voltage of the armor & bonded items at a safe level. The paper shows that this cable is likely to get hot spots, damage insulation, and even burn adjacent combustibles at fault levels below the properly rated protection device trip point.

UL seemed to be playing catch-up in trying to come up with the proper specs for this cable. Even so, the cable in the field was not living up to the published specs. For #14, they started with a tentative 1.5 ohm/100ft in 1929, but did not implement a restriction until 1941. In 1948 they said that any AC with a bonding wire was good to go without investigation. Then in 1957 they made the bonding wire mandatory and increased the size. They also significantly reduced the previous ohm requirement (the #14 dropped from 1.5 ohms/100ft to 0.75 ohms/100ft).

It was noted that some patents in the learning years indicated that they were already experiencing problems with the cable. A 1927 patent application was awarded in 1931 for including a bare grounding conductor that addressed shock hazards due to defective armor grounding (mechanical connections and high spiral resistance problems). A 1928 patent was awarded in 1930 for a bond wire in contact with the inner surface of the armor that addressed problems with the high resistance which created heat and fire hazards.

The tests in the 1920's did not consider aged conduit or insulation effects. A 1950 test by UL determined that the allowable temperature rise (54?F) would be exceeded for cable in insulation and operated at rated current. The UL seemed to be behind the curve the whole way.

In 2008 the Fire Protection Research Foundation tested some aged wiring from across the US at the UL labs and found the cables with the bonding conductor were close to UL specs but the ones without the bond were way out (2.83 ohms/100ft on average for #14 with a range of 2.07-3.53 ohms/100ft).

Sleights' paper covered the results of a project where extensive testing of 48 different cable samples was made under a wide variety of conditions. The samples were subjected to fault currents, various loads, dielectric tests, open air tests, in-circuit tests, and in-wall tests. They measured items like resistance, impedance, temperature rise, and hot spots. They also looked at the effects on insulation and the materials near the cable.

Some of the findings should make you think twice about accepting unbonded armored cable as an adequate EGC or even safe in its existing condition, grandfathered code or no grandfathered code. Here are some (paraphrased):

The DC resistances exceeded those allowed by the UL.

The DC resistance was not a good predictor of the AC impedance.

The cable temperature rise is not uniform and has hot spots.

In most cases, currents below the properly-sized breaker trip point causes temperatures to exceed the insulation rating, even when operated at room temperatures.

Under some conditions the cable can ignite surrounding combustible materials, even when protected at the cable's rated current.

Elevated armor voltages during faults can create shock and arcing hazard conditions because of the protection device not reaching its operating point.

Here is their conclusion:

The decision to continue use of old armored cable in building wiring systems should carefully consider the foregoing. Although not tested as a part of this project, if advanced branch circuit protection, such as ground fault circuit interruption (GFCI) or arc fault circuit interruption (AFCI) devices are employed to protect the entire circuit, a catastrophic failure may be averted. The failure of the cable to be serviceable will not. Degradation of armor due to corrosion can occur in concealed areas and remain undetected. Where missing due to degradation,installation or alteration, the lack of a fault current return path may allow metallic sections of circuits to remain at elevated potential for indeterminate periods.

They also make this statement about fire investigations:

Fire cause investigations at buildings where old armored cable was still in use need to consider it as a possible cause where it is found in the area of fire origin. Whether or not the particular cable was connected to an energized circuit or merely abandoned in place must be determined first. Unfortunately, arcing faults precipitated by fire attack often occur on energized cables and are not likely distinguishable from those initiated by cable failure as viewed with post-fire damage. The mechanism that precipitates the arcing is the same in both cases, thermal degradation of the conductor insulation inside the armor. The reader is also referred to Chapter 14 of Ref. [24] for further discussion regarding the difficulty of resolving cause versus victim arcing.

 

[mivey]

Now after the last ten or so posts lets get back to the issue in the OP.

What would you do if you were asked to install the ceiling fan as described in the OP and found it was supplied by old AC cable without the bonding strip?

Depending on the house it could be expensive job. What if you find most of the house is wired with the stuff?

If you turn down the job because they don't want to spend money to do it right they will find someone else to do it.

If they decide not to put in the fan at all the AC cable is still there anyway with the same hazards.

I think I would put in the fan and install ground-fault protection on the circuit. I would also take a look at the wiring to assess the condition to see if further recommendations would be in order.

The whole house may need protection installed but that would be beyond the scope of the fan job. I would certainly make the owner aware of the potential problem.

If they turn down the job, that is their business. I can't force anybody to do anything. I have done my due diligence and will make the proper note on the bill for the fan job.

If they don't want the fan job completed, I will charge them for time spent and move on.

In general, I think my response in #52 is still how I feel except that after reading the article I would want both ground-fault and arc-fault protection for this type wiring in my own home.

Not necessarily replace but provide additional protection like GFCI or even AFCI protection. That would be my minimal recommendation.

As for the "standard" recommendation, I would have to look at the condition of the wire, if there was older K&T in place, if there were BX that had been exposed to moisture, if insulation had been added over the BX, what modifications had been done, etc.

If it were my family, I would replace it or give it a thorough look-see because the potential for other things like a fault between outlets would bother me too much. Believe it or not, I would feel safer with K&T. I think the unbonded BX is that poor of a product. FWIW, the bonded armored cable is adequate (barely), but not that great of a product as far as I'm concerned. I use the MC cable with an insulated ground conductor instead.

 

[mivey]

FWIW, the article I mentioned also showed that some of the cable with the bonding strip did not meet the UL spec either. From what I read, it was close though. Probably close enough that I would not lose too much sleep over using the cable if it appeared to be in good condition. If it were my home, I would probably test the circuits just for peace of mind.

Another reason to use MC cable with a separate ground wire instead of the armored cable. Then again, I also pull a ground wire in conduit. I guess I have just seen too much old conduit that has been disturbed.

As for conduit-only grounds, I did read about the stream of sparks that fly out from conduit fittings during a fault (threaded included). It may have been in the article mentioned or in a different article where they were testing conduit but I don't feel like looking right now. Sparklers in conduit is just not my cup of tea but I don't think even the ground wire would eliminate all of that because the conduit still is a good portion of the ground path. The ground should help reduce the sparks though.

 

[M. D.]

Thanks Mivey,. for all the info.