Surge Protection

[Hedges]

I'm using the larger Delta 600 series on AC and DC side, also a surge capacitor on AC. Here's the DC:

I started a thread with test results of the Delta lightning arrestors. They are not breaking down at the voltage I expect:

Delta LA 602 Lightning Arrestor "Hey! Is this thing on??"

I've been using Delta Lightning Arrestors on my system since they were spec'd by Real Goods as part of the package I bought from them. I now have a surge capacitor on the AC input as well, which is supposed to slow rising edge of voltage so it doesn't get as high before the Silicon Oxide...

diysolarforum.com

 

[svetz]

I started a thread with test results of the Delta lightning arrestors. They are not breaking down at the voltage I expect:

Delta LA 602 Lightning Arrestor "Hey! Is this thing on??"

I've been using Delta Lightning Arrestors on my system since they were spec'd by Real Goods as part of the package I bought from them. I now have a surge capacitor on the AC input as well, which is supposed to slow rising edge of voltage so it doesn't get as high before the Silicon Oxide...

diysolarforum.com

Midnight has some videos where they show competitors failing. Not sure how common it is, might be a manufacturing defect, might just be junk sold on the market to unsuspecting people.

 

[Hedges]

A friend who is smarter than I just looked at my setup and was appalled that I didn't have surge protection. We do get a lot of lightning during Monsoon season. Midnite looks like the way to go, but one thing I don't understand is how I would connect it to the combiner box? Does anyone have an install they could show?

Your bigger concern is probably the AC line (assuming you're grid tied), because power lines are higher and go further away. That would go somewhere in your main breaker panel or anywhere up to the inverter itself. Connection is something like white (or green) wire to neutral (or ground), black & red wires to L1 and L2. (assuming neutral already connect to ground on your property, no need for separate protection on it, but your utility connection may vary.)

DC side is similar, white or green to ground, black and/or red to the two PV lines.
Even without a direct hit, nearby lightning can induce voltages, which is what these devices protect against.

It looks like the Midnight arrestors do have three varistors in a delta configuration, which clamps voltage between PV lines and each line to ground. That is what I think is required for some systems, and was lacking in other arrestors I looked at.

Another nice feature of the Midnight unit is lights to show it is connected (no blown fuse isolating it) and protection is active (not blown internally.)

Instruction manual has diagrams:

http://www.midnitesolar.com/pdfs/SPD_Installation_Manual.pdf

 

[Ampster]

It is a complicated subject from lightning rods to direct the strike to ground to the induced currents created by nearby strikes. I would guess the combiner box is not the most critical part to protect. @Hedges summed it up well as i was getting another cup of Java.

 

[Ampster]

It depends on where you expect the strike to occur? Most of us assume it will hit our roof where our panels are. Hedges reminded me that if you have overhead electrical wires a strike there could induce currents in the lines coming into our home if we are on the grid. In that case the combiner box is trivial compared to the electronics in your home including your inverter. It all depends on where you are standing. Just dont be standing at the top of a treeless hill during a lightning storm.

 

[Hedges]

I read somewhere that you should have one at the first possible point in the system, which would be the combiner box for me. That makes sense to me-why not stop the surge as early as possible to protect things downstream. I don't know, I guess I have a lot more reading to do.

Yes, clamping early is good, and a path to to earth out at the PV array is good. That leaves a longer wire run back to your equipment; resistance and inductance will attenuate the remaining surge so additional devices in the equipment protects it.

You could install another commercial arrestor (or just a varistor) at your equipment. Some of my Sunny Boys have DC and/or AC varistors in their disconnect switch. I plan to add more back in the combiner (where I presently have Delta "arrestors", which I don't think are good as anything except a spark gap for lightning; could get them to fire all the way up to 5 kV)

To me, buying the Midnight unit doesn't feel like DIY, so I was selecting varistors and parts to make status LEDs. But then I found a new old-stock unit on eBay and bought it for the AC side.

Too much capacitance for my HyPot (reactive current charging capacitor at 60 Hz exceeds 1 mA threshold), so I used the diode and capacitor from a microwave oven, gradually charged it up adjusting HyPot voltage until it reached 600 VDC. Then I threw a switch dumping that into the arrestor. All three phases clamped at 200V.

This unit has indicator lights, and two banks of varistors - after the lower voltage ones blow a second set is waiting in the wings. Lights show which are still good, and dry contacts make status available for remote monitoring. I haven't opened the cans, don't know how many kA or J it can handle, but I ain't worried anymore (on the AC side.)

 

[Hedges]

My panels are on a ground mount, thick steel frame attached to 3" steel pipe that is set in concrete at all 4 corners. No electrical lines anywhere near me. So you don't think it's worth it to put one on the combiner box for $100?

Certainly worth it.
Being off grid eliminated the most likely source, now you can address remaining risks.
I was thinking for a moment the status LEDs could affect ground-fault detector, but their DC units only have an LED between PV + and -, not to ground, so no problem.

 

[Hedges]

I get what you mean about DIY, but heck, I can't build an inverter or a solar panel, and you are several light years ahead of me in tech knowledge (3 phase solar??), so I'll get the prefab ones. Thanks for the advice.

Thanks, my day jobs have spanned microprocessors to missiles to mass spectrometers, so I used what I know to DIY and to question what vendors offer.

The Midnight units look good, and since they make the inverters and charge controllers that are to be protected, they don't miss failure modes or sell useless products like some competitors.

You can still buy additional varistors from Digikey and add them at your charge controllers. Only a couple dollars.

I don't know exactly what voltage Midnight clamps at. I tested some varistors from SMA disconnect switch and was surprised to find they fire at twice the 600V max input. I know they have to be enough above for process variations and temperature, but I'm going to add some that trip just above 600V. I know my strings are 480 Voc under ambient conditions, always below 600V in coldest weather.

 

[Hedges]

Sound good. At this time I'm only testing with 1 mA, which finds the knee of the curve. Or discharging a 1 uF cap, into whatever resistance/inductance the wiring provides. And I included a resistor in series, in case this kind of cap isn't meant to take that abuse.

I've got a 10 kV HV pulser that I haven't tried yet. It has hundreds of transistor in parallel to switch capacitor bank to output. That together with HyPot or power supply from a deceased microwave oven should do some interesting things.

Training on lockout/tagout and stored energy that I received at past jobs has prepared me well. Or maybe not. If there is radio silence then you'll know.

 

[Sunny Days]

How good are the din rail types compared to these?

 

[Hedges]

How good are the din rail types compared to these?

All comes down to specs, assuming the product actually meets specs.

Different technologies would vary in speed and V/I curves, but main thing would be peak current handling, total energy (Joules), and voltage they trigger at and clamp to.

Also how you wire them up. Ideally very short stubs of wire to the protector (or no stub, daisy chain from source to protector to load.) Also a longer wire between source and protector, also after the protector leading to load might help. Voltage would be dropped across the higher impedance of the long wire.

 

[efficientPV]

I was hit by lightning and it wasn't that bad actually. I had most stuff repaired in a few hours. I got some nice hocky puck size MOV I've never got around to installing and bins of other new MOV. My experience didn't do much to scare me. It is the STP effect, how do you know if it is actually working. Most of these devices don't actually protect equipment but prevent fires.

That said, UL has a different standard for protection devices at the meter than at the outlet some distance away. Inductance matters. If you look at the data sheet for MOV, it is rated at less than a quarter inch lead length. I've seen people coil up extra wire from a surge protector. Good luck with that. Ground path needs to be as short as possible. I knew a communications installer and he would tie five knots in each power line for added protection. Certainly several turns around an iron pipe before a surge protector will increase its effectivness.

 

[bruce_loco]

For the UK, is using UL1449 devices is probably not useful.
Could an inspector pass since the UK standard is IEC61643-1/11?