Which connections to earth/neutral are to be considered for a three-level inverter bridge?

Good morning all

I refer to the block Three-Level Bridge in library Simscape / Electrical / Specialized Power Systems / Fundamental Blocks / Power Electronics

The inverter has terminals +/-/N at DC side and A/B/C at AC side.

In usual applications there will be an earth at the AC side. Due to safety reasons I believe that the DC side should also be connected to earth. My question is how to do it without causing any faults nor disturbances in the system e.g: imbalancies, recirculating currents etc...

In principle I would consider the following:

• (-DC) terminal connected to a ground node i.e block Ground in library Simscape / Electrical / Specialized Power Systems / Fundamental Blocks / Elements
• N terminal: connected to a neutral node i.e: block Neutral in library Simscape / Electrical / Specialized Power Systems / Fundamental Blocks / Elements

Would you kindly review and comment. In particular, should the N terminal be connected to a neutral node or to the same ground node as with (-DC) terminal?

Thanks in advance, have a good day!

https://es.mathworks.com/help/physmod/sps/powersys/ref/threelevelbridge.html

Answer 1

[Joel Van Sickel]

Hello Juan,

The simple answer is to split your DC capcitor into 2, and tap this for where you connect your neutral. Don't add a ground to your converter.

See: https://www.mathworks.com/help/physmod/sps/examples/ac-dc-three-level-pwm-converter.html

Longer answer: The easiest way to handle the neutral is to split your DC link capacitor into 2 capcitors in series, of equal value, and hook the neutral up here. If you really wanted to ground the inverter, this could also be where you ground it, but I'd caution against it. While there is safety of one sort in grounding both sides of a system, there can also be dangers in that these systems are actually designed based on line voltage differentials. If you ground an inverter's negative rail, and connect it to a 3 phase system with grounded Y, the system will not work as intended, and actually creates the opportunity (gurantee actually) of unintended currents. Grounding schemes need to be careful considered for each and every application. If you intend to ground both sidess of a system, your topology and implementation need to be designed specifically for this and you need to account for the potential types of fault schenarios and what happens if the system is inbalanced. This involves more than just what you can recreate in simulation. Typically, in these systems, isolation (the grid) or lack of grounding (electric vehicles, but with very good insulation) will be used on at least one side to enable the inverter to apply the maximum line to line voltage and prevent other potential problems. A lack of grounding can also help with certain faults. Unfortunately, I can't do grounding issues justice in a forum post, but hopefully this gives you some direction.

As an example, make a simulation with a basic 3 phase inverter going into a 3 phase load that is Y grounded. Do the simulation with the inverter floating, with the inverger grounded on negative rail, and the simulation with the ground between a split DC capacitor, and see the differences. Also try a grounded inverter to a floating 3 phase load and a floating inverter to a floating 3 phase load (hook up ground through a 1 MEG resistor to negative rail). Insert a fault on the load for both instinces and see how they respond differently.

Here is an interesting reference that goes over some issues for grid tied inverters:

http://documents.dps.ny.gov/public/Common/ViewDoc.aspx?DocRefId=%7B21F6467D-8766-4705-ABBE-B13DC0005407%7D

Regards,

Joel