Do modified-sine waves affect overcurrent protection?

Status
Not open for further replies.
George Stolz

George Stolz

Moderator
Staff member
Location
Windsor, CO NEC: 2023
Occupation
Hospital Master Electrician
The other night, I was working way too late on that house in the hills. I realized there was a freezer plug buried in the garage behind a bunch of material, and mistakenly thought it was the Shangri-La of my problem, which it wasn't.

Anyway, it being late, and me being tired and lazy, and a deadbeat, I figure, "Ah, to heck with it, I'll just kick the breaker from here and put that plug in."

Bzt. Bzzzt. Bzzzzzt.

The lights flickered a bit, but that breaker hung tough. Grumbling, I throw my gloves on, put the stupid plug in, and go about my business.

Today I get a call about the house. It has a gas range with electric ignitors. The cooktop portion works, the oven works, but when they tried to turn the griddle in the middle on...

Bzzzzzzzzzzzzzzzzzzzzzzzzzzz... Plop.

The inverter kicked off, complaining of an overload.

It sounds like a dead short. A dead short on a 20 amp circuit. A 20 amp circuit that's all of 15 feet long, from the breaker to the receptacle immediately above the panel.

The breaker never kicked.

Being intuitive at times, I drew a correlation between a 15 and a 20 amp circuit not kicking their respective breakers before bringing the whole system to a screeching halt.

It's a modified sine wave inverter. Do normal off the shelf circuit breakers not heat up enough, or magnetize enough, or whatever, on a modified sine wave?

What other factors come to mind?

I am going to contact the solar dude to get his input on this. I see a serious safety concern with OCPD's in a house not reacting to ground faults. :(
 
Re: Do modified-sine waves affect overcurrent protection?

Does the inverter have a "available fault current" rating? I would think that an inverter would be a pretty "soft" supply, and may not be able to deliver the amps neccesary to trip a 15A CB via the magnetic trip mechanism. Did it ever trip? Have you looked at a trip curve of your breaker, looking at the max current your inverter(s) could supply?

I'm not sure if a non-sinewave waveform would change much. After all, many breakers can do DC as well as 400Hz. Square D has a bulletin on rerating for DC, 400Hz, high altitude, and high ambient temp here.
The multipliers aren't more than 15%, I doubt this small change would produce the effects you are talking about.
 
Re: Do modified-sine waves affect overcurrent protection?

I'm thinking along the same line as mhulbert. It just may be that it's not supplying enough current or fault current to trip the breaker, because it's too small.
 
Re: Do modified-sine waves affect overcurrent protection?

Thanks for the responses, and the link, that was helpful.

I'm headed up there today, and I have an invitation to open it up, so I'll be looking for some numbers on it. I need to add a branch circuit to the generator from the inverter.

My question is then, if it is too small, how do I compensate for it? Or do I? :confused:
 
Re: Do modified-sine waves affect overcurrent protection?

Survey says: 43 Fault Current Amps.

My guess is, a 20 amp breaker would need something to the tune of 100 amps for an instantaneous trip.

So...do I worry? It seems pretty horrible for the whole house to shut down if any one circuit has a short, but I guess it could be considered "safe".

Thoughts?
 
Re: Do modified-sine waves affect overcurrent protection?

George,
I'm looking at a trip curve for a QO breaker ( i don't know what you are using, but it will be similiar for all 20A MCCB's). At 43A, you are lookin gat 10-30 seconds to clear the fault. This is definetly in the breaker's thermal region, not the magnetic trip portion. Check out the trip curve here.

While I would guess that your installation is "safe", I'd want to know what kind of overcurrent protection the inverter uses to protect itself. It seems that if it kept starting into a fault, it may burn itself up. Your system is not "coordinated", as the smallest neccesary OCPD is not tripping on a fault. This could be quite a PITA down the road, because if the inverter keeps shutting down on overload, how do you figure out which circuit is causing it? Also, how long does it take the inverter to shut down on a short? If you had a line to chassis short in a blender, for example, would the case be energized for any amount of time, producing a shock hazzard? If you have an overload on one circuit for an extended amount of time, the breaker should finally trip.

Some options for you may be:
-get another inverter and paralel it for higher fault current, if possible.(Would reduce trip time down to 1-5 seconds at a dead short...this may not be quick enough for the inverter) This would give them some redundancy as well.
-Use fuses after your breakers, make sure they will blow quickly at 43A
-Put GFI's in for the receptacles, to decrease ground fault hazzards, and possibly regain some coordination.

None of that is cheap...but it sounds like this is an expensive system to begin with! Maybe others will chime in with some good ideas.

mike

[ November 16, 2005, 11:33 AM: Message edited by: mhulbert ]
 
Re: Do modified-sine waves affect overcurrent protection?

George,
it sounds like you are in a bad place...not sure if the system is safe as-is, and a little late in the game to go re-designing anything. I think we are at the ragged edge of the code on this...rarely is there not enough fault current available. Ite reminds me of what happens when you ground a light pole by a ground rod only...not wnough fault due to the resistance of the ground. In your case, there is little resistance, just not enough current to trip you breaker.

I looked at some of the fuse curves, and it looks like they are about the same as a breaker in the 15-20A sizes. If you fused your circuits at 5-10A, you'd be OK, check out the fuses here,

I just remembered that when I worked in entertainment we used "airpax" circuit breakers in a lot of gear...I looked at their website for what I used to use, and I think this is it. The curves look a little more favorable for your situation. Of course, building these into a solution for your setup is a whole other problem!

As for a ground fault short...why not "test" things out, except instead of using you as a guinea pig, use your meter to see how long it takes for your inverter to shut down.

Have you looked at the "PV systems and the NEC" paper published by John Wiles at New Mexico State? Actually, you may want to get in contact with him on this whole problem you are having, I hear he is very knowledgable. You can get the paper and a bunch of other PV stuff here.


Wish I could be more helpful, hopefully the above links will give you some ideas!
mike
 
Re: Do modified-sine waves affect overcurrent protection?

It sounds like you may need an overcurrent relay rather than circuit breakers or fuses.

In for a penny;in for a pound.

You might also try getting in touch with the distributed Generation group at Sandia National Laboratory. They have been working on coordination for sometime now.

http://www.eere.energy.gov/de/
 
Re: Do modified-sine waves affect overcurrent protection?

I sent an e-mail to the solar dude around January 15th.
I have yet to hear back from him, despite a follow-up.

So, what should I do? Am I making a mountain out of a molehill?

Here's a copy of the e-mail I sent.
(Solar installer),

As I promised to (the H/O), I researched the issue we had with the "bolted fault" in the range causing the inverter to go down before the particular circuit breaker tripped. If you're unfamiliar with the event, I can elaborate. Essentially, a defective appliance had a dead short to ground, causing the inverter to lock down self-defensively. It would stay off for a period, and then attempt to come back on, and repeat the cycle as long as the breakers held. The circuit breaker never tripped, because there was insufficient fault current to trip it. You attached a label to the inverter, it said the max current output is 43 amps. According to the chart I looked at, it would take a normal circuit breaker anywhere from 8 to 35 seconds to trip with pure-sine wave, traditional AC power. The inverter shuts down before the breaker can react.

The chart I looked at can be downloaded here:
(The link in the previous post.)

Unfortunately, the only solid answer I can come up with is to increase the inverter size to compensate.

I'd like to mention it's not my intent to throw a cloud of suspicion over your installation, and scare everyone. (The H/O) is aware of the issue, and not life-alteringly concerned about it. For perspective, it was the middle of November when this issue came up, and I'm just now coming to my conclusions after taking some time to research and think on it. I'm primarily concerned about the people using the equipment, and secondarily the welfare of the equipment itself. If the circuit breakers do not instantaneously trip when there is a short circuit or a ground fault, then the entire grounding/bonding system is compromised, and I see a life-safety issue in that. 43 amps at the wrong place at the wrong time can kill someone, regardless of whether the inverter cuts the current off after several seconds.

That's the reason I'm writing this book.

In a perfect world, faults are uncommon, and the breakers installed right now should still protect against overloading, but I'm still concerned that if problems arise in the future, that troubleshooting bad circuits will be difficult. Given their remote location, it would be much better for the problem circuit to trip before the entire system goes down.

I was hoping you would be able to answer whether a short circuit (or ground fault) occuring, and causing the inverter to cycle on and off will damage the inverter, or if it's designed with this scenario in mind. In my opinion, it seems to be that the inverter would be damaged if it's left in that state (say, if the fault occurs when no one is home for the day).

On a much smaller, inexpensive scale, I once plugged my hole-hawg drill into my 300W inverter, on a lark, to see what would happen. (It was a cheap, Walmart inverter.) It was billed as having built-in protection against overcurrent damage, but the high load of the hole-hawg smoked the circuitry. It'd be a shame for the expensive (much better crafted) inverters to suffer the same fate from a short circuit.

In my admittedly uneducated opinion, I could envision the inverter taking that abuse and protecting itself for minutes, or even an hour or more, but for an extended 12-hour working day, I'd be concerned.

To be sure we're on the same page, some terms:
Short circuit: An unintentional contact of near zero resistance from line-to-line or line-to-neutral.
Ground fault: An unintentional contact of near zero resistance from line-to-ground. This is either through the bare conductor of the circuit, or other paths (people).
GFCI: Disconnects a circuit if the current to ground reaches a lethal shock level to people (4ma).

I see some other solutions:

1. GFCI protection of the main breaker of the panelboard would protect the inverter against ground faults. It would have tripped in the range incident, where the switch was touching the frame of the range. Problem is, it would knock out the whole house until the fault was tracked down. It's a lousy option.

2. GFCI protection of each branch circuit. This is probably the best route, aside from increasing inverter current. The troublesome circuit would trip, and the other circuits and the inverter would be uneffected. This wouldn't provide any protection for short circuits, but at least half the battle would be won.

Problems are, GFCI breakers are at least $20+ a piece, which ain't cheap.

So, what are your thoughts?

I appreciate your attention to this,
George
Click to expand...

[ February 05, 2006, 03:59 PM: Message edited by: georgestolz ]
 
Re: Do modified-sine waves affect overcurrent protection?

George,
That letter looks spot-on with what kind of problems you may have if the current situation is left as is. Perhaps all of this is over the solar installer's head and he has no clue what to do? It seems that as the electrician, you should provided the solution, since you fully understand the problem. Also, you will be the one the ho calls when something goes wrong.


have you looked at an overcurrent relay on the inverter? Also, does the inverter have any kind of "trouble" set of closed contacts that could trigger a contacter?

A quick google brought up this OC relay. This would be a lot cheaper than GFI protection, and it also has UV/OV, var trip times, etc, so you could dial it in to match the inverter(s). Pair this with a contacter and an enclosure, and you're good to go for a few hundred in materials. You still don't have diverstiy(unless you put one of these on each ciruit...), but you'll save the inverter and you can troubleshoot by switching all of your breakers off and reapplying power till you have a trip.

mike
 
Re: Do modified-sine waves affect overcurrent protection?

Pierre, George's source of supply only provides 43 amps of fault current, even with an imposable zero impedance ground fault path the breaker will be slow to open during a bolted fault.
 
Re: Do modified-sine waves affect overcurrent protection?

Originally posted by mhulbert:
Perhaps all of this is over the solar installer's head and he has no clue what to do?
Click to expand...
I think it's that, with maybe just a pinch of lawyer-fear mixed in. All the work is to code, so there's no real footing for any lawsuit should another event occur, IMO.

This guy is a nice guy, but he's solely a solar installer.

It seems that as the electrician, you should provided the solution, since you fully understand the problem.
Click to expand...
That's a good point.

Also, does the inverter have any kind of "trouble" set of closed contacts that could trigger a contacter?
Click to expand...
I can find out. :)

A quick google brought up this OC relay.
Click to expand...
That looks awesome, I'll download the specs tonight. Thank you so much. :)
 
Re: Do modified-sine waves affect overcurrent protection?

I am not sure how a manufacturer could get a listing if there is only 43 amps of fault current available, and 20 amp overcurrent devices are permitted to be installed - unless the overcurrent devices are specifically manufacturered and tested for this type of system....but I am somewhat ignorant of this type of system and the engineering/design aspects that are in the standards followed during the manufacturing process.
 
Re: Do modified-sine waves affect overcurrent protection?

George,have you checked the true RMS output of the inverter under load?
Will line powered OCR's accept a modified sine wave?
 
Re: Do modified-sine waves affect overcurrent protection?

Originally posted by pierre:
...but I am somewhat ignorant of this type of system and the engineering/design aspects that are in the standards followed during the manufacturing process.
Click to expand...
Me too! :D

As the homeowner told me, "We're not re-inventing the wheel here. This is done all the time." I just wonder if some of the important questions have gone unasked, in deference to the "coolness" of this emerging technology. Who knows?

Check out this link to the company's home page.

The spec's for the smallest inverter is shown as 56 amps. :confused:
 
Re: Do modified-sine waves affect overcurrent protection?

GFCI devices may not work properly with modified sine wave power from an inverter. Xantrex has an Application Note on their website showing the results from testing various models of GFCI receptacles with their inverters.
 
Re: Do modified-sine waves affect overcurrent protection?

Steve, thanks for the reply. The string of "fail" columns on the Cooper row is unsettling, as that's the brand I used in the home. :(

I guess I'll be spending more time on this this weekend. :roll:
 
Re: Do modified-sine waves affect overcurrent protection?

The 43 amps is from a pair of inverters. They have all the "redundancy" they'd care for, and can afford this late in the game, I'm sure.
Click to expand...
George,if you mean you have parallel inverters, they can be out of phase with each other.
Mark
 
Status
Not open for further replies.
Top