Parallel inverters AC wire length...

[Mark-]

Hello

From the drawing below, inverters in parallel. Do the all the AC wires, (black red and grey) need to be the same length to the buss bars?

 

[zanydroid]

Following to learn.

It's not necessary for Grid Tie inverters.

My guess: For stacked inverters, i guess it depends on if the communications between the stack is enough to compensate for imbalance, and what the manufacturers want you to do. Unlike with a battery a PSW/power supply has some brains in it capable of balancing across multiple sources.

 

[Mark-]

Following to learn.

I know the current sharing wires running between the inverters must all be the same length, I asked the manufacture. The inverters come with the required wiring but, because of my inverter placement the wires were not long enough.

The user manual is silent on the AC wires and I have not asked the manufacture.

 

[zanydroid]

Which inverter is this?

 

[n2aws]

AC wire length doesn't matter, as long as they are appropriately sized, and within a few 10's of feet.

I'll also go out on a limb and say, wire length from batteris to inverters don't have to be the same either on any system using a BMS.

The "same wire length" concept is due to internal resistance of batteries, and wanting to keep them balanced as much as possible from the physical connection. in a BMS world, the BMS deals with the internal resistance, and the voltage curve is so flat with LFP through the majority of the cycle, I don't believe that the whole "same size cables" theory applies any longer, for that chemistry.

 

[zanydroid]

in a BMS world, the BMS deals with the internal resistance,

How does the BMS address this? There are no power electronics in most BMS. Cheapest ones just have monitoring and FET disconnect. Your post implies that maybe those FETs are used to match impedance. Maybe (but I’m not sure how this can happen with non communicating BMS, and that won’t work with contactor based BMS)

Power supplies and inverters have power electronics controlled already by a feedback loop, current sharing cable etc is just another input that can be put in place.

 

[n2aws]

How does the BMS address this? There are no power electronics in most BMS. Cheapest ones just have monitoring and FET disconnect. Your post implies that maybe those FETs are used to match impedance. Maybe (but I’m not sure how this can happen with non communicating BMS, and that won’t work with contactor based BMS)

Power supplies and inverters have power electronics controlled already by a feedback loop, current sharing cable etc is just another input that can be put in place.

I missed a major detail in my post.. Good catch.

I should have said "a BMS with active balancing"
The BMS doesn't *actively* address it. But the internal resistance issue isn't a concern when the BMS is handling the internal state of cells.
If they have different internal resistances (which is the primary reason we used to care about cable lengths in parallel batteries), the BMS will actively balance and get those cells in sync, even with the internal resistance differences.

 

[timselectric]

Parallel AC outputs don't have to be exactly the same length. But it's good to stay close, to keep the loads balanced between the inverters.
It's more important for the battery connections. Because of voltage drop at the lower voltage. The balancing cable signal is based on the battery voltage reference. So it's important that all inverters see the same battery voltage.

 

[zanydroid]

I missed a major detail in my post.. Good catch.

I should have said "a BMS with active balancing"
The BMS doesn't *actively* address it. But the internal resistance issue isn't a concern when the BMS is handling the internal state of cells.
If they have different internal resistances (which is the primary reason we used to care about cable lengths in parallel batteries), the BMS will actively balance and get those cells in sync, even with the internal resistance differences.

That's still kind of confusing to me because the active balancing you describe operates at cell level within a single battery. While I thought we were discussing paralleling between batteries (IE outside balancing BMS). Usually people don't use balancer across batteries if paralleling, right?

 

[sunshine_eggo]

Link #5 in my signature.

Victron has specific requirements for single phase parallel inverters or multiple parallel inverters on the same phase of a multi-phase system.

Near identical AC and DC resistances are required for balanced load sharing.

 

[sunshine_eggo]

Looking back at the OP, basically, your L1 inverters should have the same length AC and DC cables.

The L2 inverters should have the same length AC and DC cables

AND you shouldn't go too thick on your AC cables.

The L1 and L2 cables may be different from one another.

 

[Mark-]

Parallel AC outputs don't have to be exactly the same length. But it's good to stay close, to keep the loads balanced between the inverters.
It's more important for the battery connections. Because of voltage drop at the lower voltage. The balancing cable signal is based on the battery voltage reference. So it's important that all inverters see the same battery voltage.

Thanks for the response.

 

[Mark-]

Link #5 in my signature.

Victron has specific requirements for single phase parallel inverters or multiple parallel inverters on the same phase of a multi-phase system.

Near identical AC and DC resistances are required for balanced load sharing.

Thanks for the response.

 

[timselectric]

Yup
Only parallel connections need to be the same.
Series connections do not.

 

[Mark-]

Looking back at the OP, basically, your L1 inverters should have the same length AC and DC cables.

The L2 inverters should have the same length AC and DC cables

AND you shouldn't go too thick on your AC cables.

The L1 and L2 cables may be different from one another.

All are L1/L2 inverters.

 

[sunshine_eggo]

All are L1/L2 inverters.

Sorry. stuck in Victron mode.

I suspect you could get away with:

All N the same length
All L1 the same length
All L2 the same length

However, the three lines could be different lengths.

Since they're all common on the DC side, all DC cabling for each inverter same length.

 

[Mark-]

Yup
Only parallel connections need to be the same.
Series connections do not.

Thanks for the response.

 

[Mark-]

Sorry. stuck in Victron mode.

I suspect you could get away with:

All N the same length
All L1 the same length
All L2 the same length

However, the three lines could be different lengths.

Since they're all common on the DC side, all DC cabling for each inverter same length.

No problem. Thanks for the response.

 

[SolarSIGuy]

Parallel AC outputs don't have to be exactly the same length. But it's good to stay close, to keep the loads balanced between the inverters.
It's more important for the battery connections. Because of voltage drop at the lower voltage. The balancing cable signal is based on the battery voltage reference. So it's important that all inverters see the same battery voltage.

How close do inverter L1 to L2 lengths (inputs and outputs) really need to be in physical length? With an estimate of 1ns delay per 6 inches of cable, and a 60Hz period, 1ns would seem to be in the noise. Could we be talking about the potential for feet of mismatched length? I wouldn't want to do that but just curious where the boundary is.

 

[timselectric]

How close do inverter L1 to L2 lengths (inputs and outputs) really need to be in physical length? With an estimate of 1ns delay per 6 inches of cable, and a 60Hz period, 1ns would seem to be in the noise. Could we be talking about the potential for feet of mismatched length? I wouldn't want to do that but just curious where the boundary is.

There's no actual boundary.
The greater the difference, the greater the imbalance.
The inverter with the shorter conductors, will carry the larger part of the load. Proportionate to the difference. And based on how much the conductors are loaded.
At 1 foot difference and 10% loaded, it might be barely noticeable.
But at 10 feet and 80% loaded, it should be very obvious.

 

[SolarSIGuy]

There's no actual boundary.
The greater the difference, the greater the imbalance.
The inverter with the shorter conductors, will carry the larger part of the load. Proportionate to the difference. And based on how much the conductors are loaded.
At 1 foot difference and 10% loaded, it might be barely noticeable.
But at 10 feet and 80% loaded, it should be very obvious.

Thank you timselectric! This makes perfect sense to me.