Bonding Neutral to Ground in a house

lazza · October 2014

HI

I have a tricky dilemma in a new type of installation we have just put in.

As Spain's unfriendly solar laws make grid-tie systems highly unfavourable, one solution are hybrid inverters. Some new cheaper products have come onto the market which compensate somewhat for the added costs of batteries. Here is the inverter that has been installed: http://revosolar.com/solar-shop/en/inverterschargers/160-expert-inverter-charger-5000.html

These inverters you can program to only draw from the grid when there is no solar production and the batteries are low. So it switches daily or quite frequently between inverter supply and grid supply.

The problem that i've identified is this: The N wire from the grid is referenced to ground at the transformer station. So when the system pulls from the grid, it's a ground referenced system.

However, when the inverter is on (unlike the Victron inverters which have an option to connect or disconnect the N-Ground bond.).. I examined the voltages and the output is floating, that is to say that the N wire is not referenced to ground. This will have implications in the operation of RCD devices etc.

My question is: Can I reference the N wire within the house to the PE ground connection with no ill-effects? That way the system will be ground-referenced under both circumstances. But it does mean that the N wire will have more than one ground reference point when pulling from the grid- at the transfomer station and within the house. IS THIS A PROBLEM?

Hope my question makes sense

Cheers
Larry

inetdog · October 2014

Re: Bonding Neutral to Ground in a house

If the house wiring neutral is not ground referenced when it is being supplied by the inverter only, there must be a transfer switch / transfer relay inside the inverter. And that relay must be switching the neutral as well as the hot wire.

Some inverters, most commonly ones with MSW output, do not isolate their output from the battery input. The practical result is that the "neutral" wire is not only not tied to ground at the inverter output but is actually actively driven to a voltage which has a time varying relationship to ground. In this case, there is no way that you can ground the inverter output and either side of the battery bank without creating a short circuit that will destroy the inverter.

If your inverter output is truly floating (no current can flow to or from ground without an additional connection), then the right place to ground the neutral is inside the inverter on the inverter side of the transfer relay.

Bottom line, however, is that except for the niceties of the electrical code there will be no serious consequences from leaving the neutral floating. Your RCD devices should still work just fine if they are like the US GFCIs in construction.
And, if you add an additional ground to neutral bond on the house side of the inverter transfer switch you actually may have RCD problems as a result of that.

Some more detail on the RCD operation: When you are operating off the inverter, your loads are not ground referenced. If a fault occurs between any part of the load wiring and earth ground, the RCD will not trip. But the reason that is will not trip is that no fault current will be flowing and there will not be any safety hazard created by that fault.
Once you switch back to a grid connection, that equipment fault will draw current and the RCD will trip. As a practical matter, not a problem.

SolarPowered · October 2014

Re: Bonding Neutral to Ground in a house

Neutral and ground are both grounding faults, at system main, that's what allows over current protection to work for the most common of EGC~GEC. Neutral maintains a float for this reason, and indicated as the signal reference, ground (Bare Copper) is really......Just a GEC safety, unless noted EGC. Then again I don't know any type of code in spain or how Nuetral is recognized for anything other than a reference for that reason.

If its grid tied like in the case of micro inverters, and inverters requiring monitoring the neutral is actually used as a signal reference for the monitoring. I don't know what type hybrid inverter you are referencing because in the US UL/ETL may have different requirements than you have in spain, so I don't know how the hybrid in your case references unless its merely just a signal wire.

Typically for (some not all) hybrid systems we engineer 100amp pool and spa digital timmer transfer switches. The timmers will always dictate what is transferred from grid, or off grid , and this is all calculated on the battery system size, its the most basic of on and off timed switching without requiring a manual over ride.

Attachment not found.

I may also be misinterpreting the OP's question but from my knowledge the best circumstance is to maintain a distance between Nuetral and ground, if there is only one source for bussing, this will allow the neutral as a reference to be identified more clearly on a ohm/open/closed loop circuit.

Best example I can probably provide is from a sub panel using one ground buss, (or as seen in the picture, the neutral buss) confusing, isn't it. That's why separation is important, not only for key identifiers but most important, referencing. I believe in your case if you maintain separation you should be fine, but then again, I don't know Spains code.

Cariboocoot · October 2014

Re: Bonding Neutral to Ground in a house

In this case we are talking about the European 230 VAC 50 Hz standard, not the North American 240 VAC 60 Hz standard. In NA terms the Euro system has no neutral; it bonds one L to ground to diminish shock hazard and simplify over-current protection in the 120 VAC NA fashion. It is not always done (the Euro "standard" is amazingly "flexible").

The ground bond at the transformer is common to both standards, albeit using a different wire (N or L). There may be an additional N-G bond in the service panel; this will always be present in NA systems.

Larry's problem (if I understand correctly) is that when the transfer switch changes loads to the inverter it lifts both L1 and L2 from grid, thereby losing the N-G bond which is only at the power pole. This would remove all safety provided by the grounding. As such he wants to install an additional ground and bond in the service panel that would remain when the transfer switch changes, and he wonders if this will be a problem when using grid power. It should not be, provided care is taken that the same L side is grounded in both locations (i.e. L2 at pole, L2 at service panel).

To figure that out, install additional ground rod and measure Voltage potential between in and L1, L2 at the service panel. One will be full Voltage, the other should be near zero and that is the one you want bonded to ground.

Alternately bond one leg of the inverter before the transfer switch (which should operate on both lines) so that the bond occurs only when the inverter is active. This still must be on the 'correct' leg as per above and some additional arcing may take place at the switch (a lot if you get the grounded line wrong).

lazza · October 2014

Re: Bonding Neutral to Ground in a house

Thanks Cariboocoot-- that's the issue spot on.

I dont however quite understand the following:

Alternately bond one leg of the inverter before the transfer switch (which should operate on both lines) so that the bond occurs only when the inverter is active. This still must be on the 'correct' leg as per above and some additional arcing may take place at the switch (a lot if you get the grounded line wrong).

If I bond the Neutral before the switch (grid-side), it wont have effect in inverter mode. If I bond the Neutral after (on the house side), then it will operate when both the grid and inverter are working (same L1-L2 output cables from the inverter for both modes). I cant think of a way it could only bond in inverter mode, unless the inverter has a separate switch for this bond (as the Victron inverters have) ??

SolarPowered · October 2014

Re: Bonding Neutral to Ground in a house

lazza wrote: »

Thanks Cariboocoot-- that's the issue spot on.

I dont however quite understand the following:

If I bond the Neutral before the switch (grid-side), it wont have effect in inverter mode. If I bond the Neutral after (on the house side), then it will operate when both the grid and inverter are working (same L1-L2 output cables from the inverter for both modes). I cant think of a way it could only bond in inverter mode, unless the inverter has a separate switch for this bond (as the Victron inverters have) ??

I believe what coot is telling you is to find the pole that isn't switched.

According to the diagram one of the pole is dedicated one is not.

Attachment not found.

inetdog · October 2014

Re: Bonding Neutral to Ground in a house

SolarPowered wrote: »

I believe what coot is telling you is to find the pole that isn't switched.

According to the diagram one of the pole is dedicated one is not.

Attachment not found.

According the the OP, his particular inverter (not Victron) does not have this feature.

lazza · November 2014

Re: Bonding Neutral to Ground in a house

Yes, exactly this inverter doesnt have that feature.

From the little I understand - adding another Neutral to Ground bond at the breaker panel of the house in question wont create a problem. The systems here in Spain are TT for domestic customers and i'll simply be adding another Neutral to Ground bond (to the Neutral-Ground bond at transformer), when the system is connected to the grid.

However, as they say, a little knowledge is a dangerous thing

lazza · November 2014

Re: Bonding Neutral to Ground in a house

I just tried creating the additional Neutral to Ground bond in my system here with the same type of inverter. It functioned fine in inverter mode, but when switching to the grid, the RCD tripped..... Why would that be??, if I'm simply adding a Ground bond to the Neutral wire in a TT system (turning it into a kind of MEN system)?

BB. · November 2014

Re: Bonding Neutral to Ground in a house

It depends on where the ground bond(s) are and the RCD is installed.

Basically, if you have a Ground bond (between Neutral and green wire), then an RCD, and then another Ground Bond--You now have shared current flow in both the neutral and green wire.

And RCD (GFI) works by measuring the current flow in the Hot and Neutral wire. If the current in the Hot wire does not equal the return current in the Neutral wire (within ~0.005 amps or 5 milliamps), then the RCD will trip.

If you have a bond, bond, then RCD, then load (with no ground bond), there is no current "leaking" through the green wire.

If you have floating AC (genset or floating inverter output--typically TSW-true sine wave), then RCD, and then one or more ground bonds, again no RCD trip.

If you have a genset+ground bond, then RCD, then load (no bond). No trip.

If you have a genset+ground bond, then RCD, then bond, then load, you will have a trip.

-Bill

lazza · November 2014

Re: Bonding Neutral to Ground in a house

Ah of course, that makes sense. So I'd have to make the bond in the service panel before any RCDs.

In my house if I measure the AC voltage from Neutral to Ground, it gives me a reading of around 4Vac (should really be 0Vac). Is this a problem? In this circumstance, if I make a additional bond from Neutral to Ground will it cause currents to flow through ground wire?

Also, the regulation states that the resistance between Neutral-Ground should be no more than 5 Ohms .. can i measure this with a normal multimeter?

Thanks
Larry

BB. · November 2014

Re: Bonding Neutral to Ground in a house

4 VAC between neutral and return is not a lot or unusual (in my limited measurements). And yes, it is more than enough differential voltage to cause current to flow and trip an RCD if you ground bond past the RCD (GFI).

I have seen 60+ VAC between two ground rods driven into soil ~60 feet (20 meters) apart. This was in a large industrial pumping installation (large pumps used to circulate salt water in a dolphin tank). It was enough to give me a shock.

If you find large voltages between different grounds--There is a possibility that you have a hot to earth ground current leak somehow.

On measuring Neutral to Green wire bonding resistance, you cannot do it with your DMM set to Ohms if there is a voltage differential--The meter puts a little DC current into the wiring and measures the voltage drop. If you have voltage already there, it will confuse the meter (at least that is my guess).

You can take a filament lamp and connect it between neutral and green wire. Measure the AC current through the wiring (lamp will not, or at least should not) light and measure the AC voltage across the lamp. R=V/I will give you an answer (assuming you measure non-zero voltage across the lamp).

Also, take your 10-100 Watt filament lamp and connect it between Hot and Neutral, and Hot and green wire ground (before the RCD) and make sure the wiring carries the current. You can even measure the lamp voltage:

100 Watt lamp ~ 0.23 amps
measure 233 volts from hot to ground without connection
measure 230 volts from hot to ground with lamp connected and on
3 volt drop with load
R=change in voltage/change in current = 3 volts / 0.23 amps = 13 ohms (failure in this case)

An electrician should also have a ground checker that can be connected to a wall outlet to verify earth/safety ground wiring (again, without an RCD in the circuit).

-Bill

Bonding Neutral to Ground in a house

Comments

Categories