Does Your System Qualify as a "Separately Derived Source"?

[Ozark Tinkering]

Recently there's been a lot of talk around grounding and neutral issues on some of these AIOs coming from China without any UL testing. Here's a start to the conversation that I think any DIYer needs to pay very close attention to prior to embarking on a build that attempts to act as their separate source of power. It applies to ALL systems regardless of manufacturer. This is for US installations but likely has would apply anywhere based on best practice imho. Near the front of the NEC is Art. 100 and NEC Art. 100 defines a Separately Derived System as “An electrical source, other than a service, having no direct connection(s) to circuit conductors of any other electrical source other than those established by grounding and bonding connections.”
That describes most of us.

Grounding separately derived systems
Grounding
By Jack Smith
As I promised in my last column, this "Solid Ground" column deals with transformers and grounding. Let's start by addressing some definitions. Opinions differ on the "official" definition of "distribution" transformer. However, the U.S. Federal Register (Vol. 71, No. 81; page 24,995, issued April 27, 2006) defines a distribution transformer as one that meets all of the following criteria:

• Has an input voltage of 34.5 kV or less
• Has an output voltage of 600 V or less
• Is rated for operation at a frequency of 60 Hz
• Has a capacity either of 10 kVA to 2500 kVA for liquid-immersed transformers or 15 kVA to 2500 kVA for dry- type transformers.

This definition specifically excludes autotransformers, as well as drive (isolation), grounding, machine tool (control), and non-ventilated (dry-type) transformers, among a long list of others.
Although vague, this definition seems to imply that a distribution transformer is typically a utility-owned unit usually found at a substation or the utility side of a service. Many of us recall labeling these as power transformers. Apparently, the current widely accepted usage is that a distribution transformer serves a utility customer, while a power transformer serves an area.

Service entrance​

According to the National Electric Code (NEC), Article 100, the service entrance is the single point at which electrical power enters a facility. Service equipment usually includes circuit breakers, switches, fuses, and their accessories, and connects to the load end of the service conductors. The service equipment is the main control and electrical supply cutoff, but does not include metering equipment.
Service conductors - according to the NEC - originate at the utility service point and terminate on the line side of the service equipment. Conductors and equipment on the load side of the service equipment - such as secondary conductors from customer-owned transformers; conductors from generators, UPS systems, or photovoltaic (PV) systems; and conductors serving remote structures - are considered feeder conductors.
A separately derived system is "a premises wiring system whose power is derived from a source of electric energy or equipment other than a service. Such systems have no direct electrical connection, including a solidly connected grounded circuit conductor, to supply conductors originating in another system," according to the NEC. Examples of a separately derived system include transformers in which the supply, or primary, is isolated from the secondary except by magnetic coupling; generators (stand-alone or alternate power source) where the grounded conductor (neutral) is not solidly connected in the transfer switch; battery/inverter systems where the output is not interconnected; and off-grid PV systems.

Grounding and bonding​

Grounding means connecting something to the earth. Bonding means connecting objects together. A separately derived system must be grounded to earth at the source. All non-current-carrying metal parts and equipment must be bonded to the grounding point of the derived system. Bonding of metal equipment provides an effective ground-fault current path to ensure that the electrical system is safe from the potential for electric shock and fires. Proper grounding of separately derived systems stabilizes phase-to-ground voltage. NEC Article 250 in general and 250-26 in particular address the grounding requirements of a separately derived system.
Proper transformer grounding is critical. Making a grounding connection - typically to building steel, which is required to be bonded to all cold-water pipes - establishes a ground reference. Make proper bonding connections by exothermic weld, not clamps that can loosen over time. Ensure that the high-frequency impedance of the grounding electrode conductor is as low as possible. Wide, flat conductors have less inductive reactance at higher frequencies, and are preferred to round conductors for that reason. The distance between the neutral-ground (N-G) bond at the transformer and the grounding electrode should be as short as possible.
The neutral and ground must be connected to the transformer neutral bus. Making the N-G bond at the main panel is not advised in order to segregate normal return currents from ground currents. The transformer neutral bus is the only point on the system where the neutral and ground should be bonded.

Looking for trouble​

Excessive ground current and ground loops can cause equipment malfunction, inaccurate instrument readings, and safety issues. Two sources for excessive ground current are illegal N-G bonds, which can show up in subpanels, receptacles, or equipment; and "isolated" ground rods. Subpanel N-G bonds create a parallel current path by allowing normal return current to return through the grounding conductor. This creates a situation where the equipment safety ground would become the only return path if the neutral ever becomes open. Dangerous voltages could develop if the return path is high impedance.
Separate "isolated" ground rods are notorious for creating two ground references at different potentials. This situation causes ground-loop current to circulate to attempt to equalize this difference in potential. This situation can cause intermittent system and equipment problems as well as a potential safety and equipment hazard.
Transformer ground inspection should be part of your maintenance routine. Here are some suggestions of things to look for when inspecting transformer grounds:

• Check the integrity of the N-G bond using a high-quality ground impedance tester - a high-impedance N-G bond can cause voltage fluctuations
• Check the integrity of the grounding conductor and its connection to building steel using a high-quality ground impedance tester - fault currents return to the source through these connections; their impedance should be as low as possible
• Measure for ground current on the grounding conductor using a high-quality clamp meter - anything above 1 A should be cause for suspicion.

Grounding is fundamental to the safe and reliable operation of your facility. However, understanding proper grounding concepts and NEC requirements can be difficult. If in doubt, don't guess. Contact your local authority having jurisdiction for assistance.

Grounding Separately Derived Systems

As I promised in my last column, this

www.fluke.com

 

[Ozark Tinkering]

NEC
250.20 (B) Alternating Current Systems of 50Volts to less than 1000 volts. Alternating-current systems of 50vollts to less than 1000 volts that supply premisses wiring and premises wiring systems shall be grounded under any of the following conditions:

(1) Where the system is 1-phase and can be grounded so that the maximum voltage to ground on the ungrounded conductors does not exceed 150 volts

(2) Where the system is 3phase, 4 wire, wye connected in which the neutral is used as a circuit conductor.

(3) Where the system is 3phase, 4-wire, delta connected in which the midpoint of one phase winding is used as a circuit conductor.

 

[Zil]

Could someone explain to me how the title of this post matches the content?

 

[Ozark Tinkering]

Could someone explain to me how the title of this post matches the content?

Sure. I'm attempting to furnish DIYers (mostly U.S. DIYers) here a place to sort out what is required by the National Electric Code when it comes to establishing their neutral grounding and equipment bonding in a safe and legal way. For those who aren't concerned with meeting the NEC because they will never be subject to inspection or don't reside in the U.S. the NEC was and has continues to develop based on proven facts and electrical theory to make electrical installations more safe.

If the reader doesn't understand what a Separately Derived system is and they also are installing generation equipment to make their own power or plan to do so, they need to understand NEC art.100. I should have included that in my OP and did go back and edit it to include;
NEC Art. 100 defines a Separately Derived System as “An electrical source, other than a service, having no direct connection(s) to circuit conductors of any other electrical source other than those established by grounding and bonding connections.”
That describes most of us.

Did that answer the question adequately? Does it apply to you?

 

[Mike 134]

Recently there's been a lot of talk around grounding and neutral issues on some of these AIOs coming from China without any UL testing. Here's a start to the conversation that I think any DIYer needs to pay very close attention to prior to embarking on a build that attempts to act as their separate source of power. It applies to ALL systems regardless of manufacturer. This is for US installations but likely has would apply anywhere based on best practice imho. Near the front of the NEC is Art. 100 and NEC Art. 100 defines a Separately Derived System as “An electrical source, other than a service, having no direct connection(s) to circuit conductors of any other electrical source other than those established by grounding and bonding connections.”
That describes most of us.

But does it? your wires from the inverter to the grid connect directly to the service, not through a transfer switch, which isolates 2 different power supplies and can include the neutral. Today's inverters are "transformer less"

Grid connected solar systems using an inverter are NOT separately derived systems.

 

[Ozark Tinkering]

But does it? your wires from the inverter to the grid connect directly to the service, not through a transfer switch, which isolates 2 different power supplies and can include the neutral. Today's inverters are "transformer less"

Grid connected solar systems using an inverter are NOT separately derived systems.

Hi Mike, only you can answer the question in the title of this thread I posed and if you have a grid tied system (as I do,) then you answered correctly and need not concern yourself with Art. 100 of the NEC. Instead you are governed by Chapter 6 Article 690. If you are one of the many here that is building a solar system incorporating batteries and inverters and producing power in an "off-grid" manner then Article 100 of the NEC apply to you. https://allstarce.com/wp-content/uploads/2016/12/2017-NEC-Code-Ch-6-Article-690-1.pdf

 

[Ozark Tinkering]

Separately derived systems are power sources with no direct electrical connection between any of the circuit conductors and supply conductors. For example, no direct connection exists between the primary and secondary winding of a delta-wye transformer. Generators, converter windings, UPS systems, and solar photovoltaic systems are separately derived systems only when all circuit conductors — including the grounded (neutral) conductor — are independent of the supply conductors. If you wish to use a separately derived system, you must use a 3-pole transfer switch for a single-phase, 3-wire system, or a 4-pole transfer switch for a 3-phase, 4-wire system [Art. 100 definition and Sec. 250-20(d) FPN No.1].

 

[400bird]

Odd, that my UL listed Schneider inverter (and Solar Edge PV inverter) both had to be treated as separate power sources by the permitting department at my AHJ.
The PV disconnects only the 2 hots with a knife switch.
The Schneider disconnects the 2 hots with a breaker.

 

[Ozark Tinkering]

Odd, that my UL listed Schneider inverter (and Solar Edge PV inverter) both had to be treated as separate power sources by the permitting department at my AHJ.
The PV disconnects only the 2 hots with a knife switch.
The Schneider disconnects the 2 hots with a breaker.

Well this article might clear that up for you? A grid tied system has to comply with Art.690.12(1) through (5)

Rapid Shut Down​

In 2014, the National Electrical Code included a new requirement. 690.12(1) through (5) A short requirement, only six sentences.


PV system circuits installed on or in buildings shall include a rapid shutdown function that controls specific conductors in accordance with 690.12(1) through (5) as follows.

1. More than 5’ inside a building, or more than 10’ from a PV array
2. Controlled conductors shall be limited to not more than 30V and 240 volt-amperes within 10 seconds of rapid shutdown initiation.
3. Voltage and power shall be measured between any two conductors and between any conductor and ground.
4. The rapid shutdown initiation methods shall be labeled in accordance with 690.56(B).
5. Equipment that performs the rapid shutdown shall be listed and identified.
   But those six short sentences had a big change on PV system design. Gone were the days of simply wiring the solar panels, or modules, to the grid tied inverter or charge controller through a simple pass-thru or combiner box. Rapid ShutDown (RSD) has a noble cause. It is to protect the firefighters trying to put out a fire in your home or business. Even when they turned off the grid power to your house, and the inverter automatically shut itself off (in accordance with UL1741), the wires from the solar array all the way down into the inverter or controller were still live. In grid tied systems, they could have as much as 600VDC. Combine that voltage with a firefighter’s ax to vent the roof, and you have a disaster on your hands.
   
   

   NEC 690.12 Rapid Shutdown of PV Systems on Buildings - Inverters | altE

   Since its start in 2014, the NEC requirement 690.12 for rapid shutdown of PV inverters has undergone changes in 2017 and 2020. Stay up to date here.

   www.altestore.com

 

[Clifton]

Grid connected solar systems using an inverter are NOT separately derived systems.

This is the problem. It could be but if is not separately derive you need an inter connect agreement with the utility. A generator can be separately derived or not. Tie your output, neutral to their current carrying conductor or share the same ground, you need a utility agreement. If you can keep them separate while you generate, it is an sds.

 

[Ozark Tinkering]

I should reiterate this thread only covers installations in the U.S. under jurisdiction of the NEC. Other nations have their own rules and reasoning.

 

[pollenface]

Mine operates at 50hz so yeah ... nah...

 

[Ozark Tinkering]

Mine operates at 50hz so yeah ... nah...

You read and understood!!! LOL!

 

[Zil]

Yes, that is my understanding of Separately Derived Source. I study and work with mobile systems and most of those are off grid systems. I leave the S&B installations to those that specialize. But I do enjoy learning new stuff. Earth/Grounding and dc negatives are always controversial for some unexplained reason.

​

 

[fafrd]

There are several inverter-specific threads on the question of Autotransformer neutral connection, etc, but since my question is generic to any Autotransformer connected in parallel to the grid, I hope it’s not considered too big of a thread hijack to ask it here.

Several vendors including Growatt and Solis are now promoting 240VAC hybrid solutions with Autotransformers to deliver split-phase 120VAC output.

These solutions result in an Autotransformer that is always connected to L1, L2, and N when the grid is up and so the Autotransformer will be balancing whole-house load.

This means some currents flowing through the Autotransformer that don’t have to (all imbalance should be flowing through the transformer on the utility pole instead), meaning some heat being generated and energy being wasted, but I’m more focused on safety.

There is a video on one of the threads where an Autotransformer connected ‘in parallel’ with the utility transformer is considered a ‘current loop’ and a safety hazard.

So first, I’d like to question the claim that this forms a ‘current loop’: since the primary bond between neutral and ground within the mains is in between these two coils, it forms two parallel loops rather than a single loop (a figure 8). Whatever balance current flows on Neutral through the Autotransformer just reduces the amount of balance current that needs to flow through the utility transformer.

So I’m just questioning the claim this ‘Autotransformer in parallel with utility transformer’ configuration forms a true ‘loop’ and would appreciate clarification from anytime with appropriate expertise.

And secondly, I’m wondering whether there is anything in NEC that speaks to this: is it OK or against NEC to have an Autotransformer connected to L1, L2, N on the load-side of the primary N-G bond?

 

[fafrd]

Further to this question about safety of having an Autotransformer wired permanently to L1 L2 and N, I have some measurements to show proving that the Autotransformer balances a portion of full-house imbalance:

I’ll summarize the relevant measurements here based on the following code for the various modes and loads:

Normal: 240V inverter connected
Bypass: inverter bypassed
CL_ON: Critical Load breakers ON
CL_OFF: Critical Load breakers OFF
HD_ON: 1875W hairdryer ON (non-CL)
HD_OFF: 1875W hairdryer OFF

All measurements are of current through the Neutral wire from mains panel to Autotransformer:

Bypass CL_OFF HD_OFF: 0.0A
Normal CL_OFF HD_OFF: 0.0A

(So with no loads running, there is no imbalance current flowing through the Autotransformer, with or without any 240VAC current flowing into or out of the 240VAC inverter/charger).

Bypass CL_OFF HD_ON: 2.7A
Normal CL_OFF HD_ON: 2.4A

Proof that the Autotransformer is balancing ~306W or ~16% of imbalance caused by 1875W hairdryer running off of one leg (the other ~84% of the imbalance presumably being balanced by the utility pole transformer). The 1875W hairdryer was the only load running for these two measurements, from what I’ve understood.

The difference between Bypass and Normal modes could just me measurement accuracy or could be due to the increased Autotransformer loop resistance when flowing through the transfer switches within the inverter rather than through the bypass switch within the Autotransformer.

Bypass CL_ON HD_ON: 1.9A
Normal CL_ON HD_ON: 2.1A

The imbalance on the Critical Loads appears to be on the opposite leg than the leg the hairdryer was powered from, so connecting critical loads reduces the magnitude of whole-house imbalance.

Since he critical loads imbalance is reducing while-house imbalance, increasing the resistance of the Autotransformer loop by going through the inverter’s transfer switches rather than the though the bypass switch decreases the effectiveness of the offset from critical loads and results in increased current through the neutral wire (or is just measurement error).

Normal CL_ON HD_OFF: 1.15A (initial)
Normal CL_ON GS_OFF: 0.52A (final)

One of the critical loads was a fridge, so that load was likely on for one of these measurements and off for the other.

So I believe these measurements and especially the measurement with critical loads switched off and non-critical 1875W hairdryer on (whether normal or bypass) prove this Autotransformer is always balancing a portion of the whole-house imbalance in parallel with the transformer on the utility pole.

Sending an amp or two or ~15% of whole-house imbalance through an Autotransformer needlessly doesn’t seem like a great idea, but is it dangerous (or against NEC code)?

Would appreciate any opinions from the experienced voices in this thread as to whether these measurements represent a safety concern or not…

 

[Ozark Tinkering]

There are several inverter-specific threads on the question of Autotransformer neutral connection, etc, but since my question is generic to any Autotransformer connected in parallel to the grid, I hope it’s not considered too big of a thread hijack to ask it here.

Several vendors including Growatt and Solis are now promoting 240VAC hybrid solutions with Autotransformers to deliver split-phase 120VAC output.

These solutions result in an Autotransformer that is always connected to L1, L2, and N when the grid is up and so the Autotransformer will be balancing whole-house load.

This means some currents flowing through the Autotransformer that don’t have to (all imbalance should be flowing through the transformer on the utility pole instead), meaning some heat being generated and energy being wasted, but I’m more focused on safety.

There is a video on one of the threads where an Autotransformer connected ‘in parallel’ with the utility transformer is considered a ‘current loop’ and a safety hazard.

So first, I’d like to question the claim that this forms a ‘current loop’: since the primary bond between neutral and ground within the mains is in between these two coils, it forms two parallel loops rather than a single loop (a figure 8). Whatever balance current flows on Neutral through the Autotransformer just reduces the amount of balance current that needs to flow through the utility transformer.

So I’m just questioning the claim this ‘Autotransformer in parallel with utility transformer’ configuration forms a true ‘loop’ and would appreciate clarification from anytime with appropriate expertise.

And secondly, I’m wondering whether there is anything in NEC that speaks to this: is it OK or against NEC to have an Autotransformer connected to L1, L2, N on the load-side of the primary N-G bond?

On the nec question after due consideration and a little searching, I've concluded that Augie the retired inspector from Tennessee had the best answer;
"NEC rules do not normally apply to anything beyond your wiring terminating at the machine.
The transformer and interior wiring are covered by a different set of requirements.
Hopefully the machine has a U.L. or other NRTL label which would provide some assurance that the internal wiring is to some standard.
I'm shooting from the hip, but to me it sounds like a specialty transformer with windings designed for that specific machine."

So the nec doesn't answer you specifically that I found. It comes down to; Is the device approved for grid-connected or does it qualify as a separately derived system? The answer sets your guardrails.

 

[Ozark Tinkering]

Further to this question about safety of having an Autotransformer wired permanently to L1 L2 and N, I have some measurements to show proving that the Autotransformer balances a portion of full-house imbalance:

I’ll summarize the relevant measurements here based on the following code for the various modes and loads:

Normal: 240V inverter connected
Bypass: inverter bypassed
CL_ON: Critical Load breakers ON
CL_OFF: Critical Load breakers OFF
HD_ON: 1875W hairdryer ON (non-CL)
HD_OFF: 1875W hairdryer OFF

All measurements are of current through the Neutral wire from mains panel to Autotransformer:

Bypass CL_OFF HD_OFF: 0.0A
Normal CL_OFF HD_OFF: 0.0A

(So with no loads running, there is no imbalance current flowing through the Autotransformer, with or without any 240VAC current flowing into or out of the 240VAC inverter/charger).

Bypass CL_OFF HD_ON: 2.7A
Normal CL_OFF HD_ON: 2.4A

Proof that the Autotransformer is balancing ~306W or ~16% of imbalance caused by 1875W hairdryer running off of one leg (the other ~84% of the imbalance presumably being balanced by the utility pole transformer). The 1875W hairdryer was the only load running for these two measurements, from what I’ve understood.

The difference between Bypass and Normal modes could just me measurement accuracy or could be due to the increased Autotransformer loop resistance when flowing through the transfer switches within the inverter rather than through the bypass switch within the Autotransformer.

Bypass CL_ON HD_ON: 1.9A
Normal CL_ON HD_ON: 2.1A

The imbalance on the Critical Loads appears to be on the opposite leg than the leg the hairdryer was powered from, so connecting critical loads reduces the magnitude of whole-house imbalance.

Since he critical loads imbalance is reducing while-house imbalance, increasing the resistance of the Autotransformer loop by going through the inverter’s transfer switches rather than the though the bypass switch decreases the effectiveness of the offset from critical loads and results in increased current through the neutral wire (or is just measurement error).

Normal CL_ON HD_OFF: 1.15A (initial)
Normal CL_ON GS_OFF: 0.52A (final)

One of the critical loads was a fridge, so that load was likely on for one of these measurements and off for the other.

So I believe these measurements and especially the measurement with critical loads switched off and non-critical 1875W hairdryer on (whether normal or bypass) prove this Autotransformer is always balancing a portion of the whole-house imbalance in parallel with the transformer on the utility pole.

Sending an amp or two or ~15% of whole-house imbalance through an Autotransformer needlessly doesn’t seem like a great idea, but is it dangerous (or against NEC code)?

Would appreciate any opinions from the experienced voices in this thread as to whether these measurements represent a safety concern or not…

No offense but this thread isn't the place for this question. It's only focus is on helping readers determine which section of the National Electric Code applies to the system they are building. Some people are building off-grid systems and others are building grid connected systems and each has its own code requirements. This thread will attempt to provide good information for the user to make an informed decision.

 

[fafrd]

On the nec question after due consideration and a little searching, I've concluded that Augie the retired inspector from Tennessee had the best answer;
"NEC rules do not normally apply to anything beyond your wiring terminating at the machine.
The transformer and interior wiring are covered by a different set of requirements.
Hopefully the machine has a U.L. or other NRTL label which would provide some assurance that the internal wiring is to some standard.
I'm shooting from the hip, but to me it sounds like a specialty transformer with windings designed for that specific machine."

So the nec doesn't answer you specifically that I found. It comes down to; Is the device approved for grid-connected or does it qualify as a separately derived system? The answer sets your guardrails.

These inverters and Autotransformers are UL listed.

I don’t have any safety concerns when they are operating together as a ‘Seperately Derived System’ since that will only be the case when the grid is down and the Autotransformer is the only coil balancing the load (in an iff-grid configuration).

My concern is when this is operating in a grid-tie situation.

Those test results prove it essentially is an Autotransformer tied to L1 and L2 with center tap bonded to neutral.

Concerns raised by a video in the Growatt thread were:

-any ‘mismatch’ between winding of the Autotransformer and winding of the utility pole transformer - essentially if both coils did not center tap at an identical 50% point, that would cause current to flow as both coils ‘fought’ to balance neutral to different voltages).

- an overall concern about the two coils in parallel forming a ‘loop’ (though I question that since the primary ground to neutral bond is on between the two coils, forming more of a figure 8 than a loop). Still, of two seperate coils connecting L1 to L2 with a center tap to N causes a configuration where uncontrolled / dangerous currents can flow, I would have expected NEC code to have something to say about that configuration…

Any comments appreciated.

 

[Ozark Tinkering]

Any comments appreciated.

Again off topic.