Is it necessary to shut down an inverter for IV curve tracing?

Status
Not open for further replies.

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
I need to correct myself. The 2017 code would require a disconnect at the combiner to open both poles and thus there's no possibility utility sourced current would be an issue.
 

ggunn

PE (Electrical), NABCEP certified
Location
Austin, TX, USA
Occupation
Consulting Electrical Engineer - Photovoltaic Systems
I think it should be fine as long as both sides of the string under test are air gapped from the utility power or the manufacturers printed instructions says the inverter's panel end has the same degree of insulation that is required as anything else while in the powered down state and that it is ok to perform work on it without airgapping it.

What I meant was that I think it's a bad idea to leave the inverter running and open the fuse holders one at a time to get a reading. Opening the fuse holders under load is not a good idea.
 

SolarPro

Senior Member
Location
Austin, TX
I was speaking figuratively. Isolation wise, the inverter is just a solid state relay and can never completely disconnect. For example, the light bulb connected behind a touch on/off switch is still considered hot even when turned off electronically.

Tell me about it. I had to replace two switches (new out of the box and rated for the application) on a new build because they let enough leakage current through in the OFF position that the LEDs never turned all the way off.
 

Electric-Light

Senior Member
I'd place a wager that it if you accidentally fault negative to ground the inverter will shut down too fast to trip overcurrent device on the AC side.

Fuses are quite expensive but can limit current in a way nothing else can. A semiconductor switch can be considered like a normal piece of wire once it commutates erratically or fails. Semiconductor switches can not hold back fault current when things go wrong.

When you jack up a car, you're pushing energy into something much larger and a blown hydraulic means falling car. This is why inherently fail resistant passive design like a jack stand is used when personnel safety is on the line that positively holds the weight mechanically instead of relying entirely on hydraulics. The safety reliability requirements are lower when failure can only cause equipment damage. Semiconductor switches are very fast can fail easily and it does so in a way that allows load to come crashing down.
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
I was speaking from experience, albeit a limited amount.

I don't think your hydraulic metaphor holds any ... fluid. :cool: I mean, I understand the concept of failing safe, but anything besides hard numbers on all the actual components involved is pointless to discuss, in my opinion. There's a lot more to a solar inverter than a couple of semi-conductors. And the inverters are designed to control current and respond to problems in a few cycles.
 

Electric-Light

Senior Member
I was speaking from experience, albeit a limited amount.

I don't think your hydraulic metaphor holds any ... fluid. :cool: I mean, I understand the concept of failing safe, but anything besides hard numbers on all the actual components involved is pointless to discuss, in my opinion. There's a lot more to a solar inverter than a couple of semi-conductors. And the inverters are designed to control current and respond to problems in a few cycles.

It absolutely does. Safety critical matters are often discussed in collective experience of many others, known characteristics of components in use and reasoning and predictable chain of events if it fails. It is known that hydraulic systems can creep and hoses can blow. It is known what will happen when the hose blows. Many highly experienced pilots can tell you he's never had to actually use an oxygen mask outside of training, therefore he can not speak of experience in using oxygen mask in real situations. Would you posit that oxygen masks are unnecessary in airplanes? How do you come up with "real numbers" when it is something new enough that it has yet to have major life depriving incidents?

Discussion of safety is usually the collective experience from a large pool, reasoning over seriousness of the outcome in "what if" situations.

Inverters are "designed" to control current but they usual failure mode is loss of control. \Limitron thermal electrical fuses melt and interrupts before sub half cycle and limit reaching peak current. Fuses can fail by opening when they shouldn't but they do not fail to stuck closed. It minimizes component explosion and arc energy but still not adequate to protect IGBTs. Upon restart with new fuses and cleared fault the IGBT could be damaged and shorted and leave one of the panel wires directly tied to one of the AC poles.

Diagram shows them as switches but they're transistors. They're designed to switch and control the load. When they fail, it is sudden and very commonly fail shorted. Often, electrical surge/mishaps can cause them to fail shorted. When it does, it is the equivalent of welded relay contacts and do not respond to controls. The red loop is the path for fault current if one of the leads short to ground. This can happen because the midpoint of (utility's) transformer is grounded. If transistor becomes stuck on without a short, the panel side becomes live with utility power. A lot of circuitry is on the side that controls the transistors which failed shorted transistors do not respond to.

HLRXMnD.jpg


Not many people experience "stuck on" electro-mechanical relay in motion sensors but when they do the lights become permanently left-on. This usually happens when it closes into a fault. SSR will often fail when the load shorts out. The branch breaker trips. You reset the breaker to find relay is no longer controllable and permanently stuck-on. An inverter will have multiple switches arranged in a way that if they switch on at the wrong time, they will close two phases into each other which can erupt a fire ball inside the equipment box and trip the line fuses/breaker. (Hence usual warning on electrical equipment to not energize without cover!!!!).

8kdJgKK.jpg

NC/NO relays are self-checking. It is physically not capable of making on both sides simultaneously unlike solid state faulting.
1K1/1K2. ANY one relay sticking can not allow power to flow. The wiring is also routed through multiple series connected physically melting thermal limiter that backs up self resetting limiter so dangerous over heating is prevented. In addition the main burner must go through two valves in series so one stuck open can not prevent the flow to get stuck open. Why all this? Because, gas stuck-on in overheat conditions is well understood to lead to something where people get seriously injured or killed. The fail danger vulnerable parts in all inverters are transistors. They're not given redundancy because it increases complexity and costs efficiency. Inverter transistors are not life safety critical rated. This means it is not meant to be used as the only thing separating electricity and personnel. This is why you must service them with mechanical disconnect opened.
 
Last edited:

iwire

Moderator
Staff member
Location
Massachusetts
This is why inherently fail resistant passive design like a jack stand is used when personnel safety is on the line that positively holds the weight mechanically instead of relying entirely on hydraulics.

There are many applications where life safety relies entirely on hydraulic circuits.

You should not speak in absolutes. The dangers involved vs the costs or even the ability to mitigate those dangers need to be examined for each application. There is no 'always' solution.

The same applies to inverters, before we say additional safety requirements need to be added the need for them needs to be demonstrated. And by demonstrated I mean a lot more than a forum post. I mean show a real life event, not an imagined event.
 

Electric-Light

Senior Member
There are many applications where life safety relies entirely on hydraulic circuits.

You should not speak in absolutes. The dangers involved vs the costs or even the ability to mitigate those dangers need to be examined for each application. There is no 'always' solution.

The bottom of piston blew out. Elevator plummeted and crashed into the bottom. It doesn't happen often. Doubt they're going to wait for it to happen again before taking additional measures. http://www.citynews.ca/2007/01/25/citynews-rewind-the-elevator-plunge/

The same applies to inverters, before we say additional safety requirements need to be added the need for them needs to be demonstrated. And by demonstrated I mean a lot more than a forum post. I mean show a real life event, not an imagined event.

Things like air gapping utility power and wheel chocking is already in place. I am refuting the point of view that "it's fine" to work on panel side of non-isolated system and rely on inverter transistors to prevent utility sourced fault current from reaching the work area.
 

iwire

Moderator
Staff member
Location
Massachusetts
The bottom of piston blew out. Elevator plummeted and crashed into the bottom. It doesn't happen often. Doubt they're going to wait for it to happen again before taking additional measures. http://www.citynews.ca/2007/01/25/citynews-rewind-the-elevator-plunge/

Thank you for proving my points.

1) Hydraulic systems are often relied on for life safety

2) Changes to codes are typically made as a result of a failure, not the prediction of any possible failure.










Things like air gapping utility power and wheel chocking is already in place. I am refuting the point of view that "it's fine" to work on panel side of non-isolated system and rely on inverter transistors to prevent utility sourced fault current from reaching the work area.

Yes, you have put forth your opinion without showing an actual failure.
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
Diagram shows them as switches but they're transistors. They're designed to switch and control the load. When they fail, it is sudden and very commonly fail shorted. Often, electrical surge/mishaps can cause them to fail shorted. ....

What data are you speaking from that is specific to PV inverters?

...I am refuting the point of view that "it's fine" to work on panel side of non-isolated system and rely on inverter transistors to prevent utility sourced fault current from reaching the work area.

The thing is, that's a strawman. No one has contended that it is safe to rely on inverter transistors. The OP asked whether 'opening the disconnect at the combiner is enough'. It is. As I explained above (notwithstanding one misstatement which I corrected in a subsequent post), there is not a situation in which a code compliant disconnect, when opened, would leave conductors inside the combiner connected to a utility side line-conductor through a non-isolated inverter.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
What data are you speaking from that is specific to PV inverters?



The thing is, that's a strawman. No one has contended that it is safe to rely on inverter transistors. The OP asked whether 'opening the disconnect at the combiner is enough'. It is. As I explained above (notwithstanding one misstatement which I corrected in a subsequent post), there is not a situation in which a code compliant disconnect, when opened, would leave conductors inside the combiner connected to a utility side line-conductor through a non-isolated inverter.
It is only a partial straw man.
The OP says that the disconnect only opens the + lead between panels and inverter input.
If the array is floating (- lead not solidly grounded) the code requires opening both poles at the disconnect.
If one side is still connected to the inverter and is not grounded, then a failure in the inverter could in fact energize the array.

mobile
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
If the negative is grounded (which, in older codes does not require solidly grounded) then it doesn't represent a fault current risk to ground. If the grounding has somehow been lifted then the inverter won't be operating. Moreover it's an impossibility to have an inverter with a grounded conductor that's connected to a utility ungrounded conductor through a non-isolated inverter. (That's a short).

Ergo, if the disconnect is or was code compliant, but doesn't float both sides of the array, then it doesn't represent the risk that is being discussed.

I'd say proper procedure is to open the disconnect and then measure voltage to ground. No voltage ... no risk.
 

Electric-Light

Senior Member
Moreover it's an impossibility to have an inverter with a grounded conductor that's connected to a utility ungrounded conductor through a non-isolated inverter. (That's a short).

Dk3nbBH.png

If something like this is what he has in mind and he's thinking about continuing to operate the rest of the system while lifting a leg only on the string being tested. it might not affect the measurement, but it likely will be a OSHA VIOLATION. Using any part of the system to open or close a circuit under load that is not intended for switching load is not allowed. Every part of the panel side is hot and liable to short out the utility power to the ground on unisolated (may also be called transformerless) systems. It's much easier to draw an arc with DC and a small arc can easily seed a large fault. A camera strobe has the capacitor voltage across the lamp. A pulse transformer is used to seed an arc which propagates into a flash. Arc blast works just like that.

An electrical shock may trigger the inverter to shut down via RCD. A hard fault to grounded object might open the fuse and blow up some transistors as the energy flow from utility to point of fault as I indicated in grey arrow.
 
Last edited:

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
You should reread what I've already written. You're demonstrating lack of knowledge of inverter types and the corresponding code history. Your diagram does not represent a code compliant installation under any code cycle. The OP is most likely working on an isolated inverter.
 
Last edited:

Electric-Light

Senior Member
You should reread what I've already written. You're demonstrating lack of knowledge of inverter types and the corresponding code history. Your diagram does not represent a code compliant installation under any code cycle. The OP is most likely working on an isolated inverter.

I think we're on the same page that the exact configuration used by the thread starter is murky, because he didn't clarify it. There's also no disagreement that utility sourced fault current is not possible on the string end if both wires to string under test is opened by appropriate means.
He noted that an engineer suggested possibility of utility sourced fault current which means it's unclear if the system is galvanically isolated.

My diagram is simplified so I am not surprised if you saw omissions.


Seems to me it would be hard to calculate an arc flash value if it's posited to travel through electronics whose exact composition is likely a trade secret. No?

How much available fault current there is I don't know how to guess although since the inverter is built to control current I would guess not much more than its operating current.
These posts demonstrate your lack of knowledge about limitations imposed by currently utilized technology in components of unisolated inverters.
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
My diagram is simplified so I am not surprised if you saw omissions.

Omitting either a transformer or a code required open switch pole is oversimplification to the point where the diagram is no longer useful.

These posts demonstrate your lack of knowledge about limitations imposed by currently utilized technology in components of unisolated inverters.

Then please enlighten me with some detailed data or calculations. Also, I asked above if you were speaking from data or experience that is specific to PV inverters, and you've yet to respond to that.
 
Status
Not open for further replies.
Top