Use of "Pot METAL"  Electrical Outlet Boxes in Outdoor Masonry Walls
by Joe Mehaffey,  MSEE and Electrical Contractor
Revision 2, 21 September  2008

I once believed that the use of cast metal boxes in outdoor masonry walls would insure permanence.  UNfortunately, this has turned out to be false in some cases.  In August of 1997, I had  14 Thomas and Betts (Red Dot) cast metal outlet boxes installed in the retaining walls of a terrace garden.  Seven of the boxes were double duplex outlet boxes and five were single outlet boxes and two boxes were used to house outdoor light switches.  Unused holes in the boxes were sealed and a hole was drilled in the lower front of each box cover to insure that if any water did get into the box, it could drain out.  All went well for about ten  years.

In the latter part of 2007, the Ground Fault (GFCI) circuit breaker on the circuit began to trip occasionally when rains occurred.  By the summer of 2008, the breaker could not be reset even in dry weather.  An investigation revealed a DC resistance to ground of about 100K ohms with an ohmmeter on the black (hot) wire, but with 110vac applied to the black wire, the current reading of about 50ma was more than enough to trip the GFCI breaker.  

The covers to the boxes were removed and in the bottom of each box a variable amount of "grey-white dust" and larger metal particles were found.  Some boxes had a quarter inch of this "grey dust" on the bottom of the box.  The paint coating on the inside (and outside) of the boxes had completely disintegrated.   Some boxes had part of the bottom, and top eaten away, cracked, and loose ready to drop out when the cover was removed.  Virtually all boxes had the sides and top "bulging inward" as if a powerful hand had squeezed the sides of the box together.  The bending was from zero in some boxes to 1/4" in others.  I believe this was due to the corrosion of the box.  Much of the wall metal was so fragile that grabbing the wall with a pair of pliers and pushing toward the inside would easily break off a hunk.  Obviously, the corrosion damage was so severe that there was no option but to replace all the boxes.  

As a test, and starting before removing any receptacles,  we read an ohmmeter attached to the black (hot) wire as the receptacles/switches were removed from the boxes.  Most (about 70%) of the time, as a receptacle was cut free, the resistance shown by the ohmmeter increased.  When we finished,  the resistance had gone to over 2 megohms and the current measured on the black wire with 110vac applied was less than 1ma AC.  At about 30pf per foot and 400ft of wire, the .012uf  gives an AC impedance of about 220K ohms.  This would produce an AC current of 110/220000 =  0.5 ma due to the capacitance of the hot wire to ground.  This then is in reasonable  agreement with our measurements and indicates that the leakage fault is corrected.

Next, we examined the electrical outlet devices which were standard "spec grade" outlets from P&S.  Almost all were in fairly good condition considering they had been in a high dampness location for ten years.  However, we noted that some devices had a layer of grey/white dust, sometimes pretty thick, covering the portion of the outlet that was INSIDE the box.  We noticed that between the "hot" screw and grounded mounting  strap which runs vertically along the back of the device, many had signs of a blackened stripe.  On these we measured and found there was conductivity, but only when we moistened the "dust" and applied 110vac across the hot terminal and ground on the device.  Here I think we found the current path that was tripping our GFCI breaker.  

So.. It turned out to be the combination of a) the disintegrating pot metal box causing potentially conductive dust to fall on the outlets and stick and b) moisture to wet the dust, make it conductive and thus  generate a fault.  We replaced the outlets with new WR/CR rated receptacle devices.

I have used and seen used literally thousands of these cast metal boxes in masonry before without problems.  However, I have no  prior  personal history of what happens if these "pot metal" boxes are placed in a masonry wall that stays damp most of the time.  Most brickwork on homes and offices drys out pretty quickly after a rain.  This particular wall system apparently stays damp long after a rain occurs.   In this case, I observed that IN THE SAME WALLS,  identical boxes used for telephone jacks did NOT exhibit the same degree of deterioration as shown by the grounded outlet boxes.  My conclusion is that the electrical outlet boxes were tied to the electrical SAFETY GROUND and this allowed currents generated by the electrolysis of the dampened "pot metal" (a cheap Zinc Alloy) to flow  to the electrical ground and this electrolysis effect caused chemical reactions to destroy the integrity of the cast box.  Since the phone boxes had no such path to electrical ground, they were affected to a much smaller extent by corrosion.

To effect a (hopefully) permanent repair for this problem, the old Pot Metal boxes were cut out of the masonry wall with a hammer and chisel.  While this was not easy, the boxes had become extremely brittle and most boxes came out of the wall with about 30 minutes of careful hammering.  Some took an hour.    Much care had to be used to avoid damage to the wires which entered the box through a hole in the rear of the cast box.  The cast metal box was replaced by an outdoor PVC PLASTIC box made by Carlon.  I believe this will rather permanently fix the problems because PVC boxes a) are moisture resistant,  b) are not conductive to ground and have no path to ground for the safety ground wire and  c) do not corrode when exposed to moisture.  In addition, Hubbel Corrosion Resistant (CR) and Weather Resistant (WR) receptacle devices were used as replacements.  This combination should be stable in the presence of moisture and temperature cycling that is sure to occur in this application.

In addition, all of the seven Red Dot double outlet covers with weather resistant doors were defective.  None of the doors would fully close.  Many appeared bent in the hinge area, possibly through use.  Others had  the hinge+spring  corroded or contaminated with dirt and/or insect debris and the door would not close under spring pressure.  I selected mating Carlon double duplex covers but these do not seem especially robust.  Time will tell.   However, they ARE easy to replace if needed.

Summary and Recommendations:

Cast Zinc Alloy (Pot Metal) cast electrical  boxes failed in a high dampness masonry wall application in 10 years time.  The nature of the failure was determined and replacement PVC Plastic boxes and Weather and Corrosion Resistant electrical outlet devices were used to replace the prior defective boxes and devices.   This is an unusual failure but I would recommend that cast "pot metal" outlet boxes NOT be used in outdoor EMBEDDED INTO MASONRY walls or other construction which might remain wet for periods of time.  Definitely, I would not use such metal boxes in any application near salt water.

RETURN to: The GPS Information Website                              RETURN to: HOME BUILDING and REPAIR INFORMATION section