Fuses on both DC + and -

K_K
K_K Registered Users Posts: 4
Hi All, Can anyone tell me exactly why NEC requires fuses on both -ve and +ve when using a transformerless inverter?

I know that in other countries (e.g UK) this is not a requirement, so was wondering why it is here.

Comments

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Most people when they work with electricity, they are dealing with a limited power source (i.e., one power source with, at least, a breaker or fuse in the circuit). And, usually one leg of the power source is "grounded".

    When you have multiple power sources (batteries, solar panels, power supplies, etc.) with paralleled outputs--Then the chances of failures which cause unexpected amounts of current to flow in unexpected paths is much higher.

    More or less, a ground referenced power system only puts over current devices in the "hot" lead(s). Since the return leads (or neutral in North American AC power systems 120/240 VAC split phase) is never above ground (a few volts within zero volts) of ground--There is no over current flow potential if a wire gets shorted to ground.

    With floating systems, there is a real issue with ground faults. For example, say you have fuses/breakers only in the + lead of a floating DC battery powered system. As long as the system is floating, a single ground fault does not cause any short circuit current to flow. However, say you have a large Lead Acid 12 volt battery bank (floating) in your RV. And you get a ground fault in the positive lead of  the 12 volt bus of the battery bank (high current available). Now the system is positive ground because of the + lead ground fault. The whole system still works OK and no obvious problems.

    Now you get a short in a 12 volt ceiling fixture between the negative lead and ground. The negative lead from the battery is now "hot" (remember positive ground system because of the + battery bus to ground fault), and if you trace the current flow--The negative lead does not have any circuit breakers/fuses in the "negative" wiring--That means that you have 100's of amps possible from the + to ground connection, and 18 awg return lead from the lamp and now fuse--You will not overheat/fuse the return lead from the lamp and possibly start a fire.

    That is why is is customary to either have a system ground referenced with fuses in the positive leads, or floating systems with (required in North America 120/240 VAC systems) with ganged circuit breakers (or cross tripping fuses) so that both sides of the power line have circuit protection and with ganged trip breakers, if one leg trips, the other breaker is tripped also so that all power is removed from the load circuit (both breakers open if one is tripped).

    Similar issues with multiple floating series/parallel solar strings. Each solar panel has (typically) a 12 AWG cable which is good for ~15 to 20 amps continuous current (NEC) on a Imp~10 amp rated panel.

    You put 5 parallel strings together (Imp~50 amps) and only fuses in the + lead of each string--For a floating power system, a short on the DC + bus (50 amps) and a subsequent fault on one of the panel - leads with 12 AWG could get back fed with ~40 amps from the other 4 parallel connected strings. Unless you put a fuse/breaker in the negative lead to each array.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • K_K
    K_K Registered Users Posts: 4
    Thanks Bill for the detailed answer.

    What you are saying makes sense to me. I just cant figure out why this doesn't apply in other countries. For example in the UK the IET code of practice for grid connected solar systems (available for free online if you're interested) specifically states that "For systems without any functional earth connection: The overcurrent protection devices are only required in one of the active conductors"  

    Any idea why this is acceptable there? would they not have the same issue you mentioned in your last paragraph?
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    In places that use "floating power"--Such as ships which use floating AC, have a system that detects if any of the leads have been grounded and raises an alarm. In factories that use(d) floating power supplies to power equipment before it has been highpot tested would once a month test the isolation transformers.

    One of the statements of safety regulation went along the lines that the specs. were designed to only protect against a single fault (like a short circuit, not both a short circuit and a failed closed circuit breaker). But they also had the requirement that an initial and subsequent failures should fail safely.

    So--An engineered solution for a floating system should have a detection if a floating circuit leg got grounded. And to a degree, we have ground fault detection and arc fault detection (and shutdown).

    Or--You have a non-ground referenced power system and it is "safe" with a single fault. So--If you assume you do not have to protect against two faults at the same time, then there is no reason to have breakers/fuses on both outputs.

    Just to give you an idea of how much I hate poorly designed floating circuits--Here is a rant I wrote a few years ago about floating circuits and solar powered systems safety--It was ignored by the "safety experts", but I did get some phone calls and emails from folks in the engineering community wondering why it was ignored (just people searching the web found the paper here):

    http://forum.solar-electric.com/discussion/9345/system-grounding
    http://kb1uas.com/mnsforum/index.php?topic=142.0
    http://www.midnitesolar.com/pdfs/DC-GFP-Draft3-5.pdf

    There is a lot of link rot in these old posts... But the basics is there.

    -Bill

    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • K_K
    K_K Registered Users Posts: 4
    Thanks a lot Bill. This point had been bugging me, but I think I've got it now. That's quite the rant :smile: