Victron Balancer + 48VDC Axpert 4KW

[Warren]

Hi all

 

I am planning to add to my current battery bank. Currently I have 4 x 12VDC in series to get required 48VDC for inverter. Batteries are 100AH. I would like to add another bank in the same configuration and then in parallel to up the amps, and yes I know its not good to add new batteries to an older bank. My current bank is not to old yet. Less then 50 cycles. So I have the following questions

 

• What effects will I have on charging the 2 series banks in parallel?
• Should I put the newer series bank infront of the old ones to allow them to start charging first and reduce the strain on the older ones?
• Can I use a Victron balancer with my axpert and if so how many do I need as I see they only support up to 24VDC?

Awaiting your expertiese and any advice, pros and cons.

 

[Guest]

What I have learned, it is not good to have batteries in separate banks if the cables are not the exact same length.

 

Here is a wiring diagram from off-grid specialists in (I think) Canada. It is 24v, but follow the same logic for 48v.

This way you ensure some batteries are not working harder than others, for that kills a bank over time resulting in the need to replace the whole bank, even though there are still good batteries in between.

 

[Chris Hobson]

• What effects will I have on charging the 2 series banks in parallel?

 

Done properly there should be no negative effects (apart from different aged strings).

 

 

 

• Should I put the newer series bank infront of the old ones to allow them to start charging first and reduce the strain on the older ones?

 

Connect your two strings via a bussbar. Ensure that the cabling for the one string is the same length as the cabling for the other string. This is for all cabling from one busbar to the other so if your 1st string has 1 long cable, 1 medium cable and 1 short cable  then you have to replicate this for the 2nd string. I just have a standard length and all my cables are the same length. The cables connecting either end of the string to the busbar must be the same length in both strings. The reason for this is that you want each string to have the same total resistance and therefore charge at the same rate. Uneven charging/discharging is the major cause of battericide (a wonderful term).

 

 

 

• Can I use a Victron balancer with my axpert and if so how many do I need as I see they only support up to 24VDC?

 

If you have 4 batteries you need 3 BBs  you can then link 1st string and 2nd string and still use 3BBs for the additional string. 

 

https://www.victronenergy.com/upload/documents/Datasheet-Battery-Balancer-EN.pdf

 

The Victron BB to my mind is pricey at just over R1k a unit so it is almost like buying a 3rd string. Victron makes very good kit and if money was no object my house would be blue (like Plonky's).

 

There is a Chinese manufactured battery balancer that I have installed on my 4 x 260Ah batteries and after 4 months it appears my batteries no longer go out of balance. I have been tossing some ideas around but unfortunately spent my Xmas break doing crisis management with stock drink water supplies failing and 40+ ^oC temperatures.

 

Jaco has asked for me to put up a review which I will tackle tomorrow.

 

Chris

[Mark]

Thanks Chris

 

Good luck with your water/stock issues... I hope rain comes soon...

 

Praying for a wet 2016 for the country!!

 

Cheers

Mark

[___]

Banks in parallel is more of an issue when charging, than it is during discharging. During discharge, each string naturally contributes according to its own capacity. You can even parallel up different capacities (not different chemical setups or technologies though) and it will distribute the work according to their relative capacities.

 

The trouble comes when you recharge. One string will reach 100% SoC before the other does. This leaves you with two options: Stop charging, in which case the lesser string will develop sulphation in the long term, or continue charging, in which case the better string is overcharged. If the strings are similar in capacity and vented lead-acid types, this bit of overcharging is usually not a deal breaker. You pay a small price in terms of charge efficiency and overall life of the batteries, but not too terrible.

 

See this paper: http://neuralfibre.com/paul/wp-content/uploads/2007/05/can-we-now-sin.pdf

[___]

Of course there's a bit of an inventor in me thinking that if you charge them separately and combine them in some manner at night, you'd get a lot of extra life out of an old bank. You won't replace it... You'll continue using it along with the new bank until it's well and truly dead :-)

[Guest]

Of course there's a bit of an inventor in me thinking that if you charge them separately and combine them in some manner at night, you'd get a lot of extra life out of an old bank. You won't replace it... You'll continue using it along with the new bank until it's well and truly dead :-)

Morningstar says you can have more than one controller on a bank.

 

So I am wondering, if you have 2 banks of batteries, have a controller on each bank. Would that not solve the problem?

[___]

Morningstar says you can have more than one controller on a bank.

 

So I am wondering, if you have 2 banks of batteries, have a controller on each bank. Would that not solve the problem?

The simplest way is to have two separate banks (same battery tech, that is don't mix AGM with vented or Silver Calcium with regular, etc). Each bank has its own solar panels and it's own charge controller. Once the sun is down, latch the two banks together so they discharge together. When the sun is back, separate them and charge separately.

That's the simplest way. Of course there are lots of little inefficiencies and things. Eg, bank1 is full, bank2 is not, but I cannot reallocate solar panels from the one to the other to get bank2 up to full charge a little quicker, so more wastage etc.

One possibility would be a special charge controller that divides charge between multiple banks. It will be expensive, but probably cheaper than two separate ones (some components can be shared, you'll only need a buck converter for each bank). Then you interleave the PWM signals to the banks and vary the duty cycle so as to distribute the power to the two banks. If one bank is full before the other, it will go to float (low duty cycle) and you can channel the solar energy to the other one. When the sun goes down, latch the banks together as per previous idea.

Third idea, but this is seriously inefficient. When bank1 is full, use a second dc-dc converter to bleed off energy from this bank into the other one. At night, bleed energy back into bank1 as required. That is, bank2 operates similar to one of these USB "power bank" things that's really just a lithium ion battery (so you charge the phone's internal battery using an external battery - not very efficient). This will be rather terrible, as you will suffer a 15%+ loss on each transfer, so you could easily end up with only a 50% round trip efficiency :-)

With present tech, I think only the first option will actually work.

[Guest]

Each bank has its own solar panels and it's own charge controller. Once the sun is down, latch the two banks together so they discharge together. When the sun is back, separate them and charge separately.

That's the simplest way. Of course there are lots of little inefficiencies and things.

 

Like this: Charging with 1 Controller.pdf

 

Or this: One array.pdf

 

And this: Multiple Strings.pdf

 

And this: 2 Chargers.pdf

 

Seems to me the industry is starting to acknowledge, as a result of costs, we need more options.

[___]

Like this: Charging with 1 Controller.pdf

 

Or this: One array.pdf

 

What I had in mind is the second one.

 

The first one is interesting. Some thoughts:

1. Using schotky diodes is obviously because of the lower voltage drop they have. You have to be careful though, schottky diodes have fairly low reverse voltage breakdown, so check the datasheet and make sure :-) Schottky diodes also have more reverse leakage than conventional silicon diodes, but I don't think this would be a problem in this application.

 

2. I wonder how you power the charge controller at night. If there is sunshine, some energy will bleed through R1 and power the charge controller early in the morning (while the PV voltage is still too low to bias the diodes), so it can start up, but at night your charge controller will shut down. Maybe the morningstar can deal with that, but it won't work with all charge controllers. The other possibility is that there is enough reverse leakage through the diodes to power the charge controller, but I doubt it. I suspect this configuration can only be used on the lower-end relay or pwm controllers, not on the fancier mppt ones.

 

3. This is not going to get around charging dissimilar battery banks (different sizes or ages). If battery1 in this diagram gets to the point where absorption is complete, but battery2 is a little behind, you're going to have to hold to the absorption voltage and end up overcharging battery1. This is okay if they are all vented batteries and the capacities are not too different, but it's still second prize :-)

 

The idea I have in the back of my head -- and I have no idea whether it is cost-effective -- is essentially two charge controllers in one, with one microprocessor control unit driving both banks. So thinking out loud here and just explaining it out, please ignore if this is old news:

 

1. The basic building block of an MPPT charge controller is a buck converter. A buck converter is a big old inductor/coil, a switch (usually a MOSFET), a reverse-biased diode, and a smoothing capacitor on the output. When you turn the switch on, you charge the coil with magnetic energy. When you turn the switch off, the magnetic field collapses and generates a reverse-EMF (the voltage is reversed) and this biases the reverse-biased diode forwards and sends the energy to the smoothing capacitor. By carefully controlling the mark-space ration of your switch, you can pump energy from input to output at the desired rate.

 

2. The rate at which you pump the energy is seen as an impedance by the solar panels. What the MPPT charge controller does is adjust this impedance so the panel runs at a voltage where it makes the peak power.

 

3. Now take your normal MPPT charge controller, but wire two buck converters to it. Pump half the energy through the one, and half the energy through the other. If the sum of the on-times divided by the sum of the off-times of your switch is the same as for the conventional single-buck controller, the impedance seen by the panels should be the same. That is to say: We only need to implement the MPPT bit once, and we simply split our on-time calculation over the buck converters that we drive.

 

4. Interleaving. You don't drive the buck converters with the same signal. You drive them 180 degrees out of phase (for the example case of two banks).

 

5. This is similar to having two charge controllers, except that you can allocate 100% of the charge to either bank, and you only need to do the MPPT bit once. So it is very similar to the first PDF above, except I have buck converters where that one has diodes :-)

[Guest]

I see no problem in the controller being off at night. Few milliamps saved.

Or am I missing something?