importance of earthing a rack battery and its cabinet ??

[webbbn]

It helps with static electricity too, imagine you walked out of your warm cosy house into the garage and touched the battery, because you like to see the display, but you are carrying 1000's of volts because of your nylon slippers. That zap you give the battery front could cause a problem to the delicate circuitry in the BMS.

What sensitive electronics do you have exposed on your BMS display? If a static sensitive pin is exposed to you touching it randomly, then, not only is it a bad design, grounding the device will likely make it worse, since static electricity wants to reach ground.

How many times have you shocked your cell phone while picking it up or interacting with it?

 

[Quattrohead]

You got me there, no static in Florida ?

 

[webbbn]

I've lived in Sothern AZ all my life, and never had a problem with the heat. I spent some time in Destin FL, the summer of the big oil spill, and remember having to duck into stores walking down the street because I just couldn't cool off! You can have that humidity!

 

[stienman]

Some of this doesn't make sense.

First, I find it hard to believe that anyone would ground the positive lead of the battery. It's very common for the negative terminal of DC powered devices to be connected to ground, which is going to be a problem.

Here are a few resources you might enjoy reading regarding positive ground electrical systems:

What is a charge controller with a "positive ground"?

Specifically, what is the difference between these two Renogy 10A charge controllers? They are both identical in appearance, but one is "positive ground".

diysolarforum.com

What is the purpose and difference between Positive and Negative Ground type SCC?

I do not understand this and I read how people return the SCC if it is a positive and not a negative. 1. What is the recommended setup, pos or neg? 2. What is the best or preferred one, pos or neg? 3. Which SCC are affected by this (requirement?) PWM or MPPT or Both? 4. How do I test/determine...

diysolarforum.com

Positive ground PWM charge controller question

Hi, I've got a question about positive ground PMW charge controllers. I'd like to install an inline fuse and a cut off switch between my 30 amp HQST charge controller and the PV panels (2 100 watt Newpowa compact panels) This charge controller is a positive ground. Normally, if this was a...

diysolarforum.com

Getting ready to order charge controller- Is there any difference between these two?

Hi everyone, I am getting ready to order the final items needed to do my solar install. I would like to know if there is any difference between these two charge controllers; both are made by the same company; one is a Will recommendation and the other is not; one is less expensive; and one...

diysolarforum.com

positive ground mppt

I ordered the positive ground mppt a while back before I even started setting anything up. I assumed it was all negative ground as that is/should be the standard but unfortunately i was wrong again. Missed the fine bold print lol I saw some manufacturers suggest grounding the positive from the...

diysolarforum.com

Positive Ground Cars

Positive ground electrical system in old cars

www.restore-an-old-car.com

What are the tradeoffs for positive vs. negative ground?

I just got a classic Ford 8N tractor running again. Someone trying to be helpful flipped the battery around, connecting negative to ground. This made the engine crank backwards! So I had spark,

mechanics.stackexchange.com

Positive ground simplified

www.bugeyeguys.com

Why Positive Ground in Telcom Power Systems?

Today, we use wires with excellent insulation and these moisture-induced corrosion arguments carry less weight. Back in the old days, however, telecom wiring had cloth-based insulation that would g…

www.mathscinotes.com

https://www.outbackpower.com/downloads/documents/appnotes/Positive_Grounding_of_OutBack_Devices_Rev2.pdf

“Negative” 48 Volt Power: What, Why and How | Newmar Powering the Network

I don't contend that it's common, but the fact that it exists and there are still products sold today that support this system type means we have to be aware of it when designing, upgrading, or maintaining a solar power system.

Just to give an example, if you powered a Pi from a buck converter connected to your batteries, and positive is connected to ground, then the "ground" on the pi would actually be -5V WRT ground. That's not going to go well if you connect a USB port to something that has the negative voltage terminal connected to ground.

Exactly. If you're working on a solar system then you need to be aware of whether it follows the normal convention or not. Typically the end user doesn't know or care, and typically both terminals are "floating" with respect to ground.

However, high side switching with a P type mosfet is slightly more costly than low side switching with an N type mosfet, and so very cost-efficient controllers (ie, cheap junk) may choose a positive ground.

Regardless of the system ground, isolation is generally a good idea and solves a number of other problems in addition to possible positive ground systems. If you want to power your raspberry pi from your battery bank, use an isolating buck converter. If you need to measure voltage or current, use something with galvanic isolation.

Second, I don't think the batteries should be connected to the metal case, so there should be no chance of the case getting energized by the batteries or inverter. The case should be electrically isolated from the rest of the system.

It's better if they're not connected to the enclosure, I agree. In general grounding of either leg of the power supply shouldn't be done, and if it is done it should occur at only one point in the system to avoid ground currents.

But in a negative or positive grounded system you have no choice. If you connect your equipment grounds together, and one piece of equipment connects negative or positive to ground, then the rest of your enclosures are connected to the power source.

If you do connect, say the negative terminal of the battery to the case, and ground the case, you've now made a second electrical connection between your battery and the inverter, which could bypass your fuses/breakers. That's not a major problem if all your fuses, switches, breakers, and relays are on the positive side, but are you sure that's the case in the inverter? Low-side MOSFET switches are very common, especially in high current circuits.

Correct. A better design doesn't ground either side of the battery.

It's also good practice to fuse both positive and negative terminals with a linked breaker, though many systems still only fuse one side.

 

[webbbn]

Here are a few resources you might enjoy reading regarding positive ground electrical systems:

It's also good practice to fuse both positive and negative terminals with a linked breaker, though many systems still only fuse one side.

Thank you for that. There seems to be some confusion in those references about positive ground and low-side switching, which are two different things, but, of course, both are problematic.

This conversation has convinced me to use CAN bus for communicating with my BMSs and to use double breakers.

FWIW, I've been quixotic campaign against low-side switching for years, and I've made many high-current high-side switches using N-channel MOSFETs. It really costs very little in the overall cost of a device, unless it's a very low-cost device.

That was in pursuit of another quixotic quest to convince high-power drone pilots that it's not a good idea to have a drone power up when they're bent over it, plugging in a battery.

 

[rmharrington]

The batteries are bolted to the rack, with 4 bolts each.
This should be good enough for bonding everything together (you can verify this with a meter).
The bond wire going from the cabinet door to the cabinet, takes care the door. You can run a ground wire from the cabinet side connection of this wire, to your grounding system.
Wire size should be the same as for the inverter.
You can connect it to any part of the grounding system, as long as the path back to the N/G bond is the same size or larger.

Edit for clarity:
Once something is connected to the grounding system, it becomes part of the grounding system. Which allows it to be a connection point for other equipment grounding.

Hi, Tim. I appreciate the competent answer, but have a followup question. You didn't mention the marked ground bolts, one on each battery. Seems these are designed for a purpose. Are you saying I should ignore them, and ground directly to the cabinet bolt that bonds the door?

Thanks

 

[timselectric]

Hi, Tim. I appreciate the competent answer, but have a followup question. You didn't mention the marked ground bolts, one on each battery. Seems these are designed for a purpose. Are you saying I should ignore them, and ground directly to the cabinet bolt that bonds the door?

Thanks

That cabinet bolt, covers everything in the cabinet.
If the batteries were stacked on a wooden shelf. Then the ground screw on each battery would need to be used.

 

[solarHandyman]

As I have bus bars capable of 700 amps, every metal box and racking is bonded to the grounding scheme for my AC loads and solar panels frames. I was worried about static shock or worse. No negative bond added to ground on the DC components. A lot of battery cells are positively grounded.

I bought 2 AWG ground wire, drove an eight foot rod in the yard, bored a hole in the foundation with my hammer drill, dug a trench, and bonded the metal racking of the battery cabinet to the rest of the grounding system for the house with a 2 AWG split bolt. I have two rods on the north side of the house near the electric meter, and a rod at each freestanding solar array, all bonded to each other and the main panel tied to the grid. Every sub panels is bonded to the grounding system. The solar panels on my roof are also bonded to the grounding system, 5 rods in all, plus the cold water pipe, which for me is cooper because my house is old.

 

[timselectric]

As I have bus bars capable of 700 amps, every metal box and racking is bonded to the grounding scheme for my AC loads and solar panels frames. I was worried about static shock or worse. No negative bond added to ground on the DC components. A lot of battery cells are positively grounded.

I bought 2 AWG ground wire, drove an eight foot rod in the yard, bored a hole in the foundation with my hammer drill, dug a trench, and bonded the metal racking of the battery cabinet to the rest of the grounding system for the house with a 2 AWG split bolt. I have two rods on the north side of the house near the electric meter, and a rod at each freestanding solar array, all bonded to each other and the main panel tied to the grid. Every sub panels is bonded to the grounding system. The solar panels on my roof are also bonded to the grounding system, 5 rods in all, plus the cold water pipe, which for me is cooper because my house is old.

The only concern for you is a nearby lightning strike.
With all of those auxiliary ground rods.
Not sure if you are in an area that sees lightning.

 

[Zwy]

I understand how the ground/fault system works with a ac circuit on an ac breaker, but how does attaching the cabinet of dc batteries with no breaker help prevent you from getting shocked if there was a short and the case became “hot.”

If the rack or any metal is grounded (bonded) with an EGC, and there was a fault that made the metal electrically hot, the breaker that is supplying the power will trip. This is a ground fault and why neutral and ground are bonded at the service panel which is source. Current will travel back on the EGC at maximum amps and trip the breaker.

Any metal object connected to your system can be energized with AC power. While it may seem unlikely a DC battery case could be electrically charged with AC power, it still can happen. That is the reason an EGC is bonded to the battery case or rack.

i believe you but i dont understand. in an AC circuit, the fault was cause high amperage to travel through the breaker and flip the breaker, which prevents the electrocution. but in the case of a battery case becoming hot, i dont really even understand what would happen. can you explain it please?

It is not different. Any metal object could become electrically charged. Let's setup an example. Inverter malfunction sends AC power down the negative cable. This negative cable runs to the battery and the BMS is attached to the metal case. It is quite possible the AC power could travel thru the BMS to the case.

 

[zanydroid]

I understand the AC fault protection, but what is the right way to think about a 48V battery to chassis fault? What about a 120V battery to chassis fault? Seems like some kind of battery ground fault detection is needed but what is guaranteeing that with ungrounded batteries?

 

[Jim in Florida]

I might be simple minded here, but why not use EMT conduit to protect your dc wires from your rack of batteries to your inverter. Gives bonding by default. Or were the 600 pounds battery rack something you wanted to mover around from time to time?

 

[zanydroid]

I might be simple minded here, but why not use EMT conduit to protect your dc wires from your rack of batteries to your inverter. Gives bonding by default. Or were the 600 pounds battery rack something you wanted to mover around from time to time?

Pretty sure that or wireway is required by code (I think non metallic is ok too).

The PowerPro wireway you can buy with it + 18kpv does this

 

[Jim in Florida]

Pretty sure that or wireway is required by code (I think non metallic is ok too).

The PowerPro wireway you can buy with it + 18kpv does this

With the title of the thread in mind using non-metallic would skip the “earthing” step (unless an additional equipment ground conductor is used) while metallic conduit would provide bonding to earth assuming your inverter is properly grounded.

 

[Zwy]

I understand the AC fault protection, but what is the right way to think about a 48V battery to chassis fault?

There should be a fuse on the positive side of the battery. That is to protect against a battery to chassis fault.

I have 4 batteries in the cabinet, each one has a 175A T type fuse. If there was a short from battery positive to negative, the fuse will blow. I do not have a wire from battery negative to the cabinet.

What about a 120V battery to chassis fault? Seems like some kind of battery ground fault detection is needed but what is guaranteeing that with ungrounded batteries?

I think you are confusing AC power and the requirement for bonding enclosures to the equipment grounding system with the battery negative. Do not run a wire from battery negative to the chassis/enclosure. There is no need. Only run an EGC to a metal enclosure that could possibly become energized.

Much of the confusion is due to NEC referring to bonding as grounding. Perhaps in the future the NEC will correct this and only refer to bonding. Bonding is not the same as grounding.

 

[zanydroid]

I think you are confusing AC power and the requirement for bonding enclosures to the equipment grounding system with the battery negative. Do not run a wire from battery negative to the chassis/enclosure. There is no need. Only run an EGC to a metal enclosure that could possibly become energized.

I’m not making a statement either way about where the chassis will be bonded to.

What will guarantee that a battery DC+ or DC- fault to chassis will be cleared or detected?

Does the battery system voltage make a difference on what the code minimum mitigation for this is?

 

[ImAnIdiotPleaseBePatient]

If the rack or any metal is grounded (bonded) with an EGC, and there was a fault that made the metal electrically hot, the breaker that is supplying the power will trip. This is a ground fault and why neutral and ground are bonded at the service panel which is source. Current will travel back on the EGC at maximum amps and trip the breaker.

Any metal object connected to your system can be energized with AC power. While it may seem unlikely a DC battery case could be electrically charged with AC power, it still can happen. That is the reason an EGC is bonded to the battery case or rack.


It is not different. Any metal object could become electrically charged. Let's setup an example. Inverter malfunction sends AC power down the negative cable. This negative cable runs to the battery and the BMS is attached to the metal case. It is quite possible the AC power could travel thru the BMS to the case.

Gotcha. I was thinking it was for DC power leaking from one of the batteries, and I just have no idea what would even happen when DC current ends up on you ground system. AC power leakage makes more sense from a breaker that would trip.

 

[Zwy]

I’m not making a statement either way about where the chassis will be bonded to.

What will guarantee that a battery DC+ or DC- fault to chassis will be cleared or detected?

If the chassis was connected to the DC-, and a DC+ fault occurred, the fuse on the DC+ will blow.

If there isn't any connection from DC- to the chassis, nothing occurs. The battery is a low voltage DC storage device. Due to it being DC, current will try to make it the opposing pole but if there is not any connection to the opposing pole, no current will flow.

In your mind, you are thinking DC current is the same as AC current. Go to your car and touch the + and - terminals on the battery at the same time. Did you die? Even on a nominal 48V battery, will you die if you touch both terminals at the same time?

Does the battery system voltage make a difference on what the code minimum mitigation for this is?

What makes the difference is a battery is DC.

 

[timselectric]

If I remember correctly.
80v is the threshold for being deadly to humans.
That's what it takes to stop the heart.
Safety standards are less strict below this.

 

[zanydroid]

What makes the difference is a battery is DC.

Hmm, not voltage? DC is safer to a higher voltage than AC.

80v is the threshold for being deadly to humans.
That's what it takes to stop the heart.

What about available current?

Also the threshold for safe voltage level in rapid shutdown was set to 30V (and 240VA in some code iterations). Fixed 48V battery is way way above that.