DC Grounding

D

dtdarrah7

Member
Location
San Antonio
Occupation
Electrical Engineer

Grounding & Bonding for a 24V DC Hydroponics Setup - Needed Input​


Hello all,

I am designing an electrical system for my Senior Design Project, which powers a hydroponics setup, and I’d appreciate some guidance on grounding and bonding.

System Overview:​

  • Power Source: 120V AC to 24V DC charger (Victron Blue Smart IP22, 24V 16A x2 in parallel) acting as a power supply and battery charger (includes backflow prevention).
  • DC Bus:Powered by the charger, distributing power to three circuits:
    1. 24V 100Ah LiFePO₄ battery (backup power source).
    2. 24V DC to 120V AC inverter (powers two sump pumps, controlled via hi-low sensors).
    3. 24V DC to 5V DC buck converter (powers a Raspberry Pi for monitoring).
  • Operation: Runs on building power under normal conditions; if power is lost, the system islands and supplies loads via the battery + inverter.

Grounding & Bonding Questions:​

  1. DC Panel (Metallic Enclosure): Should the chassis be bonded to the building grounding system?
  2. Battery Negative Terminal: Does it need to be bonded to the building ground, or should it remain floating?
  3. Inverter Grounding Terminal: Does this need to be bonded to the building grounding system?
  4. Hydroponics Water Tanks: Since the water has high electrolytic content, should I install grounding electrodes in the tanks?

My Thoughts So Far:​

  • I believe the DC panel chassis should be bonded to the building ground for touch potential safety.
  • I am reconsidering bonding the battery negative to building ground, as my DC panel is metallic, and I want to create a path back to the source for my OCP to work.
  • NEC compliance is a priority, but since my system is below 60V DC, I’m trying to determine the best approach for grounding while maintaining safe operation.
I would love to hear your input on this! Any NEC references or best practices would be greatly appreciated.

Thanks!
 
The DC panel chassis should be bonded; this can be done through the conduit or cable from the AC-to-DC charger. This is equipment grounding, not system grounding.

I would *not* bond the battery negative, at least not without further investigation. DC systems below 60V are not required to have system grounding. Read the AC-to-DC charger manual, it may contain specific instructions on this. There is a possibility that it already effectively grounds the system negative, or (less likely) that it would be dangerous to do so. Since it's not code required I wouldn't do it unless equipment documentation told me to.

I'm more concerned about grounding for your inverter. That system is supposed to be grounded but possibly that cannot be done safely without it being separately derived from the AC side of the charger. Is all this equipment designed to work together or are you cobbling it together yourself? There are potential saftey issues here depending on the internal design of the charger and inverter. If they are designed by a single manufacturer to work together, read the manuals.

I would not put grounding electrodes inside water tanks, you'll just get corrosion I think.
 
Great! This is excellent! As you said, I will bond the DC panel to the building ground as it is equipment grounding.

I agree. I was looking into high-impedance (I guess it's high-resistance in the case of DC) grounding, which can be used to monitor for SLG faults while reducing fault currents to a few mA (perhaps this strategy would need additional actuation/mitigation from the control circuitry to trip my OCP).

The manufacturer makes the inverter and chargers (Victron Energy). I have opted to install additional load monitoring and actuation from Victron (Smart Dongle for the inverter and the shunt sensor and "smart" disconnect for the battery - I am installing a manual emergency disconnect to bypass this automatic disconnect as well).

Bonding the inverter's chassis ground to the AC output neutral is an option, according to the manufacturer's specifications/manual. I have decided that, at a minimum, I should feed the AC output to a GFCI and then plug in the pump loads for an extra layer of protection.

Would this inverter be considered a separately derived source? I am having difficulty deciphering the code to answer this, especially since it is not feeding building loads, just the hydroponics pump loads.

I agree. I was also thinking about this, and having all of the formation protections in place eliminated playing a grounding electrode in the tanks.

Thank you very much for the information!
 
Last edited:
I think the inverter would count as an Separately Derived System (SDS), so the neutral side should be bonded to the "ground".

That said.....
If you haven't already, read NEC article 250 on grounding. Look at the definitions of grounding electrode, grounding electrode conductor (GEC), equipment grounding conductor (EGC), and the like. The GEC and EGC serve different purposes. Also, look at where fault currents can flow and how those might trip an over-current protective device (OCPD).
 
An inverter is not necessarily a separately derived system. But, if the manual says you can bond the AC output neutral to ground, then it probably is one. Really depends whether it has an isolation transformer in it, which may not be clear in the manual.

For what it's worth, from what I've seen (albeit on paper, not in person) Victron products are not exactly designed and documented with NEC compliance in mind.
 
Excellent, these insights are invaluable! As I understand, a Separately Derived System is a "power supply output, other than a service, having no direct connection(s) to circuit conductors of any other electrical source other than those established by grounding and bonding connections." I believe lumping the inverter with the Battery makes it a separately derived system - I am not sure, but I think there is galvanic isolation, at least in the inverter (I opened it up to look at the location of the bonding jumper and saw/felt the toroidal transformer in the unit - I can only assume that there is indeed isolation). Based on the definition of a separately derived system and NEC 250.30(A)(4), it seems that bonding my inverter to the building grounding system would make the building grounding system the inverter's grounding electrode. Thus, the wire bonding the inverter to the building grounding system would be the GEC. Does this sound correct?

So, an EGC would only exist if I were to feed a circuit, such as a GFCI outlet in a weatherproof metal enclosure, from my inverter. Does this sound correct?

If Victron products are not designed with NEC compliance in mind, I want to be sure I can do everything I can to bring the system utilizing the equipment up to code. Next, I need to confirm or deny whether the products are listed. I thought they were, but I need to see if I can find the UL sign or any other NRTL insignias on the chassis of the products.
 
jaggedben said:
An inverter is not necessarily a separately derived system. But, if the manual says you can bond the AC output neutral to ground, then it probably is one. Really depends whether it has an isolation transformer in it, which may not be clear in the manual.

For what it's worth, from what I've seen (albeit on paper, not in person) Victron products are not exactly designed and documented with NEC compliance in mind.
Click to expand...
I don't know that the DC side of a PV system qualifies as an SDS. The inverter does not supply power to the modules and for modern inverters there is no grounded conductor on the DC side to bond to the EGC.
 
Inverters that contain transformers that galvanically isolate the DC and AC sides can be considered separately derived systems per the NEC definition. However, article 690 has its own things to say about DC system grounding, which basically takes precedence over anything in 250 applying to an SDS. And over the years, most of what 690 had to say got removed as inverters with transformers became less common. Also as far as I know the code doesn't directly address DC grounding when PV and battery are in parallel: essentially, it relies on manufacturer instructions to avoid issues there. Code requirements for grounding of other DC systems are conflicted and not comprehensive.

It's also important to note that most PV inverters nowadays are 'transformerless' (a.k.a. non-isolated) and therefore cannot have a grounded DC conductor because that will create a dead short between AC and DC when the inverter operates. I don't know if that's the case with Victron, but also even when a DC conductor is functionally grounded that is done through the inverter, not a user installed solid connection to ground.. Again, it's very important to read manuals carefully or contact manufacturers in these situations where you don't know how the inverter works internally, before you go grounding anything willy-nilly.
 
Got it. Thank you all! The inverter's manual instructed the grounding of the inverter's AC output.

These comments have been instrumental to the way I am designing this system. Thank you again!
 
You must log in or register to reply here.
Top