Surge Protection

[svetz]

So, did you add on of these guys to your installation?

The idea is they short-circuit any stray voltage so those voltages don't add to the wear and tear on your more expensive electronics. Obviously the one in the image is lot more sophisticated than the power strips, but the same basic idea. They even have LEDs so you know when they're working (versus consumed). I'm thinking about getting one (Florida lightning and all), but wanted some second and 3rd opinions.

I particularly like how easy they are to install, you just attach to L1/L2/N. They make them for AC and DC.

These suppressors have nothing to do with fuses or circuit breakers. Generally surges & spikes are too quick. 50,000V could even jump across a blown fuses.

Basically these devices are like the sacrificial anode in the hot water tank, they take the spike and wear out so your more expensive gear doesn't. They do wear out, that's why they have an LED, when the LED goes out it's time to replace it. I figure if they're wearing out they most being doing something good.

I may not live in the lightning capital of the world anymore, but we still get plenty.

So, basically this is like an insurance policy, spending $100 to help increase the longevity of your inverter. It also works on the whole house, so I won't have to replace the ones on the PCs/TVs/Router when they wear out.

 

[svetz]

It's not the single direct hit that's liable to fry your gear... it's all the constant small hits from any surge that chip away at it and it can be happening dozens of times a day.

Lightning is fun stuff... did you know it can generate antimatter? https://phys.org/news/2017-11-lightning-chance-antimatter.html
Discovered a few years ago by satellites!

Here's the NASA reference, they estimate it happens 500 times per day around the world.

 

[dougbert]

I had one on my array 1, but have moved things, and need to re-attach. Maybe I need a new one, but mine doesn't have a GOOD LED

 

[svetz]

Been doing some additional research...saw all these on the internet... so they must be true! (That's sarcasm, so take 'em with a grain of salt)

How much lightning do you get? Check out this map, TLDR: other than a few hot spots, probably only a concern for the southeast.

A spike is considered less than 3 nanoseconds, a surge is considered greater than 3 nanoseconds.

80% of surges occur from within the home due to motors in the home turning on/off; typically lower voltages than external sources.

The most common occurrence of an external surges is caused by the power company (e.g., capacitor bank switching, brownouts, blackouts), next is is lightning, and then finally most rare is if you live next door to Will when he's experimenting.

Anything containing a microprocessor is being constantly degraded by strong voltage fluctuations; essentially causing electronic rust. No surprise electrons running through silicon degrades them, but it would be nice to how higher voltages affect the degradation rate.

A good ground is very important since SPDs redirect to ground.

Nothing will stop destruction caused by lightning strike within 90 feet, but lightning 10 miles away can still cause spikes

There are different types that work on DC, AC, and different voltages (see www.youtube.com/watch?&v=-B64V_FZBFQ)

 

[svetz]

This is interesting (well, to me anyway)... the IET (UK authority) mandates surge protection in some instances (e.g., hospitals) and has a formula to calculate the risk level (CRL) to determine if an SPD must be installed elsewhere (ref). In terms of SPDs as insurance, it's sort of like an actuarial table.

If the CRL value is less than 1000 (or less than a 1 in 1000 chance) then SPD protection shall be installed. Similarly if the CRL value is 1000 or higher (or greater than a 1 in 1000 chance) then SPD protection is not required for the installation.

The CRL is found by the following formula:

CRL = fenv / (LP x Ng)

Where:

fenv is an environmental factor and the value of fenv shall be selected according to Table 443.1
LP is the risk assessment length in km
Ng is the lightning ground flash density (flashes per km² per year) relevant to the location of the power line and connected structure

The reference has information to calculate the variables. For example, LP is an amalgamation of electrical line characteristics dealing with the length of the line. From their calculation, underground lines only count half as much as overhead lines.

Example 1
Building in rural environment in Notts with power supplied by overhead lines of which 0.4 km is LV line and 0.6 km is HV line Ground flash density Ng for central Notts = 1 (from Figure 05 UK flash density map).

Environmental factor fenv = 85 (for rural environment – see Table 2)

Risk assessment length LP
LP = 2 LPAL + LPCL + 0.4 LPAH + 0.2 LPCH
LP = (2 × 0.4) + (0.4 × 0.6)
LP = 1.04

Where:

• LPAL is the length (km) of low-voltage overhead line = 0.4
• LPAH is the length (km) of high-voltage overhead line = 0.6
• LPCL is the length (km) of low-voltage underground cable = 0
• LPCH is the length (km) of high-voltage underground cable = 0


Calculated Risk Level (CRL)
CRL = fenv / (LP × Ng)
CRL = 85 / (1.04 × 1)
CRL = 81.7

In this case, SPD protection shall be installed as the CRL value is less than 1000.

 

[svetz]

Ulmo posted this on the old forums, too good to not repost:

 

[gnubie]

It's not the single direct hit that's liable to fry your gear... it's all the constant small hits from any surge that chip away at it and it can be happening dozens of times a day.

That goes for the surge protector too. MOVs silently degrade with each low energy event. Their response time increases and current handling ability decreases.

 

[StevAR]

Fwiw, during lightning my 12vdc car- backup camera monitor gets solid white static during a nearby strike.

I usually unplug and/or turn off stuff when that happens! Nothing has died yet, but it's a worry.

It's disturbing!

 

[Ulmo]

Ulmo posted this on the old forums, too good to not repost:

Thanks!

Good info but the only way I can watch his videos is at 1.5 - 2.0 speed...

Yup! That section isn't too thought-intensive, so it's good to speed it up. But some of his discussions can use about as much thought as he gives you time for when he's talking slow.

Here's my original post which includes a description of what I already did:

My most recent reference to this came from Jack Rickard's video at youtu.be/-TCsr_mJfvI?t=1654 starting at timepoint 27:34 in the video (and I see Will Prowse already watched that video a few months ago because he commented there). Jack Rickard talks about it for 1 hour, 1 minute, and 1 second starting at 27:34. So, I ordered two Littelfuse V271BA60's, and a few days later on around September 13, 2019 I installed two Littelfuse V271BA60's in my home breaker panel like Jack showed he did for his V321BA60 in his video (which I couldn't find because he bought them all; I figure it's ok at the lower rating since the lightning strike map for California shows a lot fewer strikes). Product page at www.littelfuse.com/products/varistors/industrial-high-energy/ba.aspx and link there for its data sheet; they have buttons where you can check stock and order at all the supply companies they sell those to.

I'm thinking I'll install something like that or the unit pictured in the first message above on every installation I manage from now on. I already installed a whole house surge protector on the last house panel I replaced at another site I managed in around 2012 or so. So, it seems I'm already in practice of installing these whenever I remember. Thanks to Jack, I remembered again, and learned how, all in one video!

I'm curious if I will ever notice the difference with my varistors installed here. We have 2 Tesla PowerWall 2's and a 5,880Wp solar system, and a number of medium-priced computers. I figure the $200 cost for two Varistors was trivial compared to the expense of dealing with ruined equipment due to spikes.

I think I want to add a MidNite surge protector on another part of my system if I expand it; I already feel like I ought to put one on the PG&E meter panel as it is. Right now, the LittelFuse is on my backed up loads panel in the house which has the old original ground line since it used to be the main service panel and only panel, but now the separate main service panel is bonded to it (a ground line connecting the two) and the new main service panel is where the ground connects to neutral (per code), so that's funky, but oh well; kind of typical in upgrades, really. But anyway, I might want one or two more surge protectors: one more on a new DIY solar + battery build that I somehow hook into the existing home, and one more on the PG&E utility side right at the meter, since sometimes that load panel gets disconnected from the inside backed up panel during outages (by the existing Tesla Powerwall Gateway) or electrical work (I'll turn off the breaker connecting PG&E). Currently, there is a car charger on the outside main service panel not backed up by batteries and without a dedicated surge protector, so if my batteries went into backup mode, the car could be zapped by a surge and there wouldn't be any surge protector connected. Since I don't have an EV to charge right now, and we don't get much lightning or utility outages, I can wait until something like that changes until I install the one or two additional protectors.

I feel like the utilities are going to play anti-solar games and other games like turning off the grid randomly (such as with the recent letter from PG&E talking about turning off the grid during times of dry air), and that would be enough of a reason to add one more surge protector on the utility side as I said. I probably have a year before they play that game around here.

 

[svetz]

So, this is probably an EE theory type question... but Midnight makes surge protects for the DC side. I'm not sure I see the value with direct current.

Wires on the roof do act as antenna's providing an AC signal on top of the DC...but they're pretty short all in all so shouldn't be much impact. Surges from high current draws typically cause voltage to drop, so there's not a spike on a frequency as with DC.

I can see the MPPT clicking on/off and the inverter's rapid demands causing the surge events; but, since they are the cause, surge protection might not help. So is DC surge protection primarily for if you have heavy DC inductive loads (e.g., a solar AC)? Or perhaps this is the reason for putting a supercapacitor in front of the battery, using it's rapid response to filter aberrant AC and handle any voltage dips/peaks?

 

[gnubie]

Conductors have inductance. If there is a sudden significant drop in current you'll get an inductive spike. The faster the rate of change in current, the higher the spike. That's not the problem here.

Think more in terms of static charges accumulating on things and going zap zap zap. FETs do not appreciate over voltage and it takes 2/3 of bugger all current to break a FET if the voltage goes too high. Of course there are gross events such as lightning too.

 

[svetz]

Conductors have inductance. If there is a sudden significant drop in current you'll get an inductive spike. The faster the rate of change in current, the higher the spike. That's not the problem here.

Think more in terms of static charges accumulating on things and going zap zap zap. FETs do not appreciate over voltage and it takes 2/3 of bugger all current to break a FET if the voltage goes too high. Of course there are gross events such as lightning too.

Except for the roof wires, generally the DCs wire are short... so you recommend these without reservation at the combiner box then? A bit confused...FETs would be the inverter, no?

 

[gnubie]

All conductors have inductance.

Personally if I was out to protect my system from any sort of over-voltage event I'd be placing a protection device just before the charge controller / inverter / or what ever the panels are feeding and making sure it and the grounding system could handle the anticipated current. You have to consider the inductance of the ground wire too. Right angle turns bumps the impedance of the ground wire up quite a bit and that's the last thing you want if you are trying to cope with a lot of current as all you'll end up with is high voltages appearing where you don't want them to and a dead what ever it is you are trying to protect.

 

[Capt Bill]

I just started searching threads here for in on surge protection. I did post this in one other thread titled: Inverter surge compensation:

Wonder if this might be good idea at 120 ac Circuit Breaker Box (on AC out side of Inverter?

Delta CA302-R 2-Pole 125/250V Single Phase Surge Capacitor
handles surges on both legs of 240 split phase too.

Robot Check

and for Lightning surges that can come into breaker boxes from the grid supply lines
Delta LA302-R 2-Pole 300/120-240V Single Phase Lightning Arrestor

Amazon.com

When looking at options years ago, I found the Delta line to be reasonably priced, I installed both of these on a 100 amps service panel in Costa Rica where lightning is a norm in rainy season, along with grid off and on surges, and other surge like changes early in the mornings. One time during a lightning storm, inside near my breaker panel, I heard a loud crackle just after I heard the boom of a nearby lightning strike (that can sound like canons going off when striking nearby). I am pretty sure that was the Delta Lighten Arrestor sending a huge incoming lightning-generated spike that came to my breaker box via the grid lines in, and sent that spike to ground (instead of through all the stuff I had plugged into the grid.

My main Calif. home is not in a lightning prone area, but I do wonder if what people use on the Solar IN lines when living in a Lightning prone area (thinking of Costa Rica buddies). I plan to install Delta CA302-R Single Phase Surge Capacitor on my 120vac Breaker Box from my Inverter, with thoughts it may add some extra ability.

What is your Take ?

 

[svetz]

After watching the Midnight SPD comparison it made me pay more attention to the SPD specs and want independent laboratory testing.

...My main Calif. home is not in a lightning prone area...

From the dirty power thread I learned that while lightning can cause instantaneous failure, they're fairly rare. Whereas your utility (and to some extent internal devices and neighbors) are constantly causing smaller surges that are wearing away at your electronics.

 

[Steve_S]

It is interesting that some don't use such protections even if in a low lightning strike zone. I designed, and assembled multi-million dollar server rooms for Military & Government and let me tell you about protecting everything from anything possible... triple redundancy on everything and for darned good reason too... it is ALWAYS the unexpected that get's you in the end. Lightning, surges and brownouts on grid connections, you name it had to be planned & accounted for. Funny thing is, most of it is common sense stuff too while the military / gov systems can be a tad extreme (part of the critical risk factor though) it all makes good sense for everyone that depends on their equipment.

I used the Midnite Protectors. Easy as pie to install but a tad on the pricey side, there are some other very good high quality suppressors out there that are well priced BUT there is also some shlock krap too, so best advice do not skimp and buy High Quality suppressors from reputable vendors ! I saw some cheapo knock offs a couple of years ago being flogged as the real deal but at deep discount, which could have killed you deeply too. (they came out of India).

 

[Hedges]

I just started searching threads here for in on surge protection.

What is your Take ?

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

http://www.deltala.com/download/LA602DC_CUT_SHEET_2.pdf

But I see a problem with it. My application (SMA Sunny Boy 10000TLUS-12) has DC inputs of 600V or less. Both (+) and (-) leads are ungrounded; it is a transformerless inverter. If used for 208 3-phase rather than 240 split phase, the DC wires would carry an AC signal, because they connect to the AC hots through inductors and active components.

I would like to use the lightning arrestor to clamp the two DC leads to be no more than 600V apart, also each to be no more than +/- 600V from ground. It hasn't been designed that way, just each wire +/- 600V from ground. I think it would allow them to be 1200V apart which would kill the inverter. So I've wired it with the common lead to one of the DC wires instead. I figure limiting DC input is more important than limiting voltage to ground - hopefully that could take 1200V. Using two of these DC suppressors per inverter would clamp all the voltages. I think three would cover my two inverters.

 

[Ampster]

I replied no because I use a different type of whole house surge suppression.

 

[Capt Bill]

Wow 6

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

http://www.deltala.com/download/LA602DC_CUT_SHEET_2.pdf

But I see a problem with it. My application (SMA Sunny Boy 10000TLUS-12) has DC inputs of 600V or less. Both (+) and (-) leads are ungrounded; it is a transformerless inverter. If used for 208 3-phase rather than 240 split phase, the DC wires would carry an AC signal, because they connect to the AC hots through inductors and active components.

I would like to use the lightning arrestor to clamp the two DC leads to be no more than 600V apart, also each to be no more than +/- 600V from ground. It hasn't been designed that way, just each wire +/- 600V from ground. I think it would allow them to be 1200V apart which would kill the inverter. So I've wired it with the common lead to one of the DC wires instead. I figure limiting DC input is more important than limiting voltage to ground - hopefully that could take 1200V. Using two of these DC suppressors per inverter would clamp all the voltages. I think three would cover my two inverters.

Wow to 600 v dc input; You are in big gun territory compared to my projects; and so is 3 phase. Wondering: Are you in an area prone to cycles of lightning? Is that the main reason why you want to put on this extra protection? ... just a question to draw ya out more for the group study :+)

 

[Hedges]

Wow 6

Wow to 600 v dc input; You are in big gun territory compared to my projects; and so is 3 phase. Wondering: Are you in an area prone to cycles of lightning? Is that the main reason why you want to put on this extra protection? ... just a question to draw ya out more for the group study :+)

I'm not presently using 3-phase (except for a pool pump; I put in a 3-phase motor and variable frequency drive.)

Something just under 600Voc is common for grid-tied string inverters. With no batteries involved, just connect a bunch of PV panels in series and have a bridge of four high voltage low current FETs. Some inverters take a wide range of PV voltages, and some require minimum voltage to be greater than peak of the AC sine wave (so minimum DC voltage is given on a graph vs. AC.) The 10kW transformerless inverters I'm using now have a narrow range, so 8, 36V panels in series is the only quantity that fits well across temperature.

I'm in an area of California without a lot of lightning, and utility hookup is from an underground transformer so I doubt I'll get much even if lightning hits a power line. My first system using SWR-2500U was designed by Real Goods and had both AC and DC protection, so I kept doing that as I expanded. My latest SB 10000TLUS-12 inverters have small protection devices built in, and I put more outside. It is cheap protection, and I now have a bank of Sunny Island (my picture) to take care of.

If anything, once I have the Sunny Boy set up as UPS for the whole house, that'll provide brownout protection for other equipment. At this time I manually transfer.

As I mentioned in my post, I don't think the DC arrestors have what's needed for transformerless inverters. I want three elements connected in delta (triangle) configuration. All of Delta's products are two or three elements connected in a "Y". I wrote to them and got no response. I think I need to buy one more of their LA602DC to make my own triangle. That way three PV leads (+), (-), GND will all be clamped to max 600V from each other.