ggunn
PE (Electrical), NABCEP certified
- Location
- Austin, TX, USA
- Occupation
- Consulting Electrical Engineer - Photovoltaic Systems
But fusing only the positive for either case?
What is going to be the industry standard when replacing inverters?
- Thread starter pv_n00b
- Start date
- Status
Carultch
Senior Member
- Location
- Massachusetts
Double-fusing is no longer necessary, like it has been for non-isolated systems. The positive side is the de-facto side that will likely get the only fusing, although if desired, you could fuse the negative side instead.
jaggedben
Senior Member
- Location
- Northern California
- Occupation
- Solar and Energy Storage Installer
I was getting confused between the fusing and the disconnecting requirements. Thanks all for clearing things up.
Just out of interest, where did you get that figure from?
Carultch
Senior Member
- Location
- Massachusetts
Today, the inverters usually come with a 10 year warranty, so it is a realistic figure.
jaggedben
Senior Member
- Location
- Northern California
- Occupation
- Solar and Energy Storage Installer
I think the range is greater than ten years. It does not figure that when manufacturers issue 10, 12 or 15 year warranties they expect that all of their products are going to fail before that period is up.
I have noticed that commercial range inverters have shorter warranty periods than residential though.
I have noticed that commercial range inverters have shorter warranty periods than residential though.
ggunn
PE (Electrical), NABCEP certified
- Location
- Austin, TX, USA
- Occupation
- Consulting Electrical Engineer - Photovoltaic Systems
Not all their products but a significant number of them. That's why inverter warranty extensions are fairly expensive; they figure into the cost the portion of the inverters that they anticipate will fail and have to be replaced.
jaggedben
Senior Member
- Location
- Northern California
- Occupation
- Solar and Energy Storage Installer
Understood, but let's back up. pv_noob said 7-10 years is 'typical' lifetime for inverters. Boeseker asked where that number came from. Most residential string inverters have a 10 or 12 year warranty. So while the number of those inverters that fail under warranty may be 'significant' (10%?) I certainly don't believe that 7-10 years is the 'typical' lifetime. I believe it's more like 12-15 years. On the other hand many commercial inverters now only have a 5 year standard warranty, so 7-10 years may be 'typical' of commercial inverters and maybe that's what pv_noob had in mind.
Just trying to shed light on what Boeseker asked about. Real world data is probably not applicable yet since most inverters in the world haven't been around 7-10 years yet.
It seems to be what installers are reporting and it lines up with standard warranty periods that in the past have been 10 years or less. Keep in mind that a warranty is basically an insurance policy. The cost of the warranty is related to how many inverters they estimate they will have to replace under warranty. You can buy an extended warranty for some period after the regular warranty expires, as with all insurance the manufacturer is betting that the inverter will not die and you are betting it will. It does not mean that if you buy a 10 year warranty extension on a 10 year warranty that they make a special inverter for you that is designed to last 20 years.
Last edited:
What will be the best way to handle the white conductor in the array when it is changed to an ungrounded system? The NEC allows single conductors that are not white in a grounded PV array to by marked at terminations as white. In that case the marking can just be removed and it's good. But if someone actually used white conductor there is no provision anywhere in the NEC I know of that would allow it to be marked as any other color. I get this because it's less of a hazard to someone to mistake a grounded conductor to be ungrounded than it is for someone to mistake an ungrounded conductor to be grounded.
The warrany period is what the manufacturer supports within the cost of the product.
Not an indication of life expectancy.
In my experience over a number of decades, failures are rare.
But what do I know.............
My posts keep being deleted so someone knows better - it would seem.
jaggedben
Senior Member
- Location
- Northern California
- Occupation
- Solar and Energy Storage Installer
This has always been a bit of a bugaboo for me because then you also have that sticker that says "WARNING: In case of a ground fault indication, normally grounded conductors maybe ungrounded and energized." So the white color never had a reliable meaning and it was always, and will always be, advisable to measure voltages before touching anything related to an array you are servicing. So you raise an interesting code compliancy point but for me personally I really don't care because I'll make no assumptions based on color.
I guess if repaired/refurbished/overhauled systems are going to be reinspected then an inspector can make you replace white conductors. I couldn't argue with that; I sort of think they shouldn't have been white to begin with and I've certainly been glad to see white disappear in recent years. However, if I'm simply replacing an inverter on an array and wiring that is in good enough condition to safely operate, I'm neither going to call for an inspection nor replace white conductors. I'll re-identify conductors with tape (+ = red, - = black) and strive to leave a note for the next guy.
SolarPro
Senior Member
- Location
- Austin, TX
Exactly. Leaving a note in one of the cans is a great idea.
When my HVAC system breaks down, our mechanical contractor doesn't wait in line, pull a new permit and incur a bunch of unnecessary expenses. They just fix the problem and get the system up and running as soon as possible.
I'd expect our solar contractor to do the same. Customers just want to get their money maker back on line. The AHJ already knows that there is a PV system on site. Bob's your uncle.
When a new permit has to be pulled for repair work is not easily defined in my experience, and I don't agree that it's always going to be a no permit process. These are not HVAC systems. Frankly I think pulling a permit for an inverter replacement will come down to don't ask, don't tell for most people who don't want to complicate their lives but that's probably not the way it should be. They will just do the work and if someone reports it and the AHJ asks they will just say they were doing a simple repair and did not think a permit was required. Better to ask for forgiveness than permission right?
From my personal point of view a replacement that dropped in a substantially similar inverter, same isolation, same power rating, same AC current rating, same breaker rating, should be a no permit repair. Increasing the power rating, changing the array from grounded to ungrounded, or other major changes I would assume the building department might want to inspect it so that would be worth asking about. The utility might want to know too. If a dumb inverter is replaced with a new fangled smart inverter they might want to connect to it and make the distribution system better.
jaggedben
Senior Member
- Location
- Northern California
- Occupation
- Solar and Energy Storage Installer
To me the red parts are the key. The interconnection agreement says what you can and can't do without revising the agreement. If you increase the power rating more than allowed, or whatever, then you're going to be obligated to tell the utility. If you're obligated to tell the utility, then you ought to at least have an approved building permit in case the utility asks you about it, even if you opted not to tell them initially.
Electric-Light
Senior Member
hehehehe thanks for reminding about the heated discussion I've got into over the wording a long time ago.
http://forums.mikeholt.com/showthread.php?t=171411
Now with freshly redrawn diagram
The TOP models the "transformerless" or unisolated topology while the BOTTOM is the traditional isolated design. I used a battery charger for illustrative purpose. Both provides the same function and charge the battery. The battery sees the same charging effect as they're both capable of providing current effectively to terminals which is called differential mode current.
Going into differences..
Transformerless: The voltage to battery is regulated entierly through series SCRs and chokes. The common mode fault current and voltage is tied to the utility. You will see 14v charging between + and -, but the vehicle chassis reaches + 170v pk every line cycle. Semiconductor swiches like MOSFET transistors are not perfect insulators and they are not considered the same as physical gap insulation. They have leakage current and usual failure mode is bolted short across terminals. Available fault current is set by the PoCo's transformer and the wiring impedance. Breakers can not really limit peak let through current. Current limiting fuses do, but it depends on the customer using the correct fuses. The uncertainty of people using the right fuse is the primary reason for phase out of fuse panels in homes. Let's leave it at that. If you accidentally touch the fuel tank with a corded drill, fault current can go through the PCC->charger->car->drill chuck->EGC->utility transformer with the entire utility fault current available with a fair chance you'll blow a hole through the tank and catch the gas on fire. Conducted EMI created by solar system can also flow into the system.
Transformer galvanic isolation: Energy can only cross the utility to solar system boundary magnetically short of lightning strike on panels inducing enough voltage to arc across the two windings. I also feel that owing to the sky facing, exposed, large surface area of solar panels, it's exponentially more likely to accidentally have a contact with primary compared to the chance of 120v/240v phase conductor touching an overhead line. I'm currently looking into what impact such would have and if galvanic isolation makes any difference in the severity of such.
Carultch
Senior Member
- Location
- Massachusetts
Fault current cannot pass through a transformer? I'm confused.hehehehe thanks for reminding about the heated discussion I've got into over the wording a long time ago.
http://forums.mikeholt.com/showthread.php?t=171411
Now with freshly redrawn diagram
The TOP models the "transformerless" or unisolated topology while the BOTTOM is the traditional isolated design. I used a battery charger for illustrative purpose. Both provides the same function and charge the battery. The battery sees the same charging effect as they're both capable of providing current effectively to terminals which is called differential mode current.
Going into differences..
Transformerless: The voltage to battery is regulated entierly through series SCRs and chokes. The common mode fault current and voltage is tied to the utility. You will see 14v charging between + and -, but the vehicle chassis reaches + 170v pk every line cycle. Semiconductor swiches like MOSFET transistors are not perfect insulators and they are not considered the same as physical gap insulation. They have leakage current and usual failure mode is bolted short across terminals. Available fault current is set by the PoCo's transformer and the wiring impedance. Breakers can not really limit peak let through current. Current limiting fuses do, but it depends on the customer using the correct fuses. The uncertainty of people using the right fuse is the primary reason for phase out of fuse panels in homes. Let's leave it at that. If you accidentally touch the fuel tank with a corded drill, fault current can go through the PCC->charger->car->drill chuck->EGC->utility transformer with the entire utility fault current available with a fair chance you'll blow a hole through the tank and catch the gas on fire. Conducted EMI created by solar system can also flow into the system.
Transformer galvanic isolation: Energy can only cross the utility to solar system boundary magnetically short of lightning strike on panels inducing enough voltage to arc across the two windings. I also feel that owing to the sky facing, exposed, large surface area of solar panels, it's exponentially more likely to accidentally have a contact with primary compared to the chance of 120v/240v phase conductor touching an overhead line. I'm currently looking into what impact such would have and if galvanic isolation makes any difference in the severity of such.
I know the current doesn't directly flow across the coil core, but the rapid spikes in current can still manifest via magnetism an induced current on the opposite side.
- Location
- Placerville, CA, USA
- Occupation
- Retired PV System Designer
The magnitude of the fault current induced on the secondary side of the transformer is limited by core saturation and so is not as high a multiple of the normal current as it is on the primary side.
So, yes, there can be fault current, just not as great as it could be in a transformerless installation.
mobile
Electric-Light
Senior Member
Not allowing direct flow limits some faults. Real components also act very differently from simplified ideal model.
A synchronous motor that mechanically turns a synchronous alternator by a non conductive shaft are not electrically connected. The output of the alternator is not electrically linked to the motor's power so you can touch any one lead to any part of the power source and it wouldn't do anything. There has to be a current flow across the alternator leads to induce any interaction.
If you place a load across the alternator, the current causes a load on the rotor magnetically which induces braking torque on the motor turning the alternator. The motor resists slowing by drawing more power. If the entire mechanical and electrical linkages are infinitely stiff, a bolted short at the output would stall the rotor and lock up the alternators at the power plant as if you threw a wedge in it. We obviously know that can't happen in actuality.
I know transformer doesn't have moving parts but I think this gives a better understand of galvanic isolation and the limiting effects of real life sized components.
electrofelon
Senior Member
- Location
- Cherry Valley NY, Seattle, WA
- Occupation
- Electrician
hehehehe thanks for reminding about the heated discussion I've got into over the wording a long time ago.
http://forums.mikeholt.com/showthread.php?t=171411
Now with freshly redrawn diagram
The TOP models the "transformerless" or unisolated topology while the BOTTOM is the traditional isolated design. I used a battery charger for illustrative purpose. Both provides the same function and charge the battery. The battery sees the same charging effect as they're both capable of providing current effectively to terminals which is called differential mode current.
Going into differences..
Transformerless: The voltage to battery is regulated entierly through series SCRs and chokes. The common mode fault current and voltage is tied to the utility. You will see 14v charging between + and -, but the vehicle chassis reaches + 170v pk every line cycle. Semiconductor swiches like MOSFET transistors are not perfect insulators and they are not considered the same as physical gap insulation. They have leakage current and usual failure mode is bolted short across terminals. Available fault current is set by the PoCo's transformer and the wiring impedance. Breakers can not really limit peak let through current. Current limiting fuses do, but it depends on the customer using the correct fuses. The uncertainty of people using the right fuse is the primary reason for phase out of fuse panels in homes. Let's leave it at that. If you accidentally touch the fuel tank with a corded drill, fault current can go through the PCC->charger->car->drill chuck->EGC->utility transformer with the entire utility fault current available with a fair chance you'll blow a hole through the tank and catch the gas on fire. Conducted EMI created by solar system can also flow into the system.
Transformer galvanic isolation: Energy can only cross the utility to solar system boundary magnetically short of lightning strike on panels inducing enough voltage to arc across the two windings. I also feel that owing to the sky facing, exposed, large surface area of solar panels, it's exponentially more likely to accidentally have a contact with primary compared to the chance of 120v/240v phase conductor touching an overhead line. I'm currently looking into what impact such would have and if galvanic isolation makes any difference in the severity of such.
I understand the concept of galvanic isolation and that intuitively it could be seen as safer to have isolation between two parts of an electrical system, but are you saying that transformerless inverters/Pv systems are less safe? If so do you have any evidence of this? Have there been cases of these high energy faults that would not have happened with a transformer based inverter?
- Status