48V system charge settings?

ChrisOlson · November 2012

Re: 48V system charge settings?

Artboard wrote: »

Absorb is set for 58V for :2:00 hrs.
Float is 55V

We have 4000-series batteries as well. When we initially had problems with low SG, Surrette told me to increase the absorb voltage to 31, minimum 3 hours time, with temp compensated controllers. They told me the lead-calcium grids in the 4000-series need higher charging voltage.

After three months of deficit charging (a brand new bank) with the same settings you posted, it took a month of running EQ's once a week at 32 volts, and absorbing at every opportunity at the new higher voltage setting to finally get them up to 1.265. They've been fine ever since, and that was 2 years ago.

We bought a new 48 volt inverter now, so double that for 48 volts. Might want to call Surrette tech support to get the full scoop on charging. If you've been deficit charging yours for three years you might have some problems. Be interesting to hear what Surrette tells you to do about it.

If I had it to do over I wouldn't buy Surrette batteries again - I'd get Trojans. I don't really like the gawd awful voltages that the Surrettes require.
--
Chris

Cariboocoot · November 2012

Re: 48V system charge settings?

ChrisOlson wrote: »

If I had it to do over I wouldn't buy Surrette batteries again - I'd get Trojans. I don't really like the gawd awful voltages that the Surrettes require.
--
Chris

I agree. That and the issues with quality/service they've had lately make them not worth the price in my opinion.

ChrisOlson · November 2012

Re: 48V system charge settings?

Cariboocoot wrote: »

I agree. That and the issues with quality/service they've had lately make them not worth the price in my opinion.

What's really disappointing for us is the charging efficiency is so low with them because of the high voltage. They're in the low 80's% for amp-hours in vs amp-hours out. Trojans are way better than that.
--
Chris

BB. · November 2012

Re: 48V system charge settings?

And it can be difficult to find AC inverters (especially the less expensive ones) that will take that high of DC input voltage and not warn/fault.

-Bill

ChrisOlson · November 2012

Re: 48V system charge settings?

Exactly. I have our new SW Plus 5548 set as high as it will go - 64 volts. And I have the temp compensation limit set in the Classics to 63.5 so we got .5 volt of headroom.
--
Chris

Artboard · November 2012

Re: 48V system charge settings?

Hi Cariboocoot,

I've spent the last two weekends working on the system. Disassembled the battery banks, cleaned all corroded connections, and connected just one battery bank to one 48V bank (times two, since I have two) and EQd them.

Changes I made in the 6048 config panel were

1. Increased max voltage to 70V (to avoid turning off during EQ).

2. Decreased Battery capacity from 1600 AH to 800 AH, since I only had two banks going at the same time.

First time things worked OK, because the batteries all came back to 1.265/cell (except 1 cell stuck at 1.185). When I switched the 6048s to EQ the second two banks of batteries, the second bank off one of the 6048s decided not to EQ at all. In fact, the SGs of that bank dropped significantly!

After reading the advice on http://www.smartgauge.co.uk/batt_con.html regarding how to wire up 4 batteries, and your advice about just connecting each 6048 to a single 48V battery bank to do a rescue EQ, what do you think about me just connecting two 48V battery banks in parallel, connecting each 6048 from diagonal corners of each bank, and doing this 2x, since I have 2 - 6048s? I have uploaded a diagram illustrating my two theories, and would appreciate advice on each.

The two inverters aren't connected to the same battery points? Well one will have a different Voltage available than the other. Guess what that will do.

Take the battery bank apart. Connect one 48 Volt string to one XW and another to the other one. Straight forward one string each, all with clean, tight connections. Fire up the generator and charge them fully. Then try EQ on just those strings. Use the gen for the whole sequence if you have to: you're trying to save the batteries and the fuel cost is not important. Keep trying an EQ cycle until you see no further improvement in SG. At that point, no matter what the reading is, you're done. If the cells don't come up to around 1.265 and aren't consistent, the batteries are done too. (Do not power any loads from the inverters while doing this.)

Cariboocoot · November 2012

Re: 48V system charge settings?

See this bit:

First time things worked OK, because the batteries all came back to 1.265/cell (except 1 cell stuck at 1.185). When I switched the 6048s to EQ the second two banks of batteries, the second bank off one of the 6048s decided not to EQ at all. In fact, the SGs of that bank dropped significantly!

That's why I suggested one string at a time. It also works better if the SG's are close in the cells of any re-assembled string. In other words, if there's one cell whose SG is far off from the others that battery/string will be the most difficult to bring up to normal. That one cell "stuck" at 1.185 is a problem; that battery is probably toast.

I do not understand why one string wouldn't allow the XW to go to EQ. How bad off was it? You do know you have to fully charge them, check the water level, and then EQ. It's a tedious process. The batteries should not be able to prevent the XW from going to EQ. The SG was going down probably because the inverter was running from the batteries instead of pushing power to them. Something must have gone wrong with the configuration that time.

I would not recommend trying to parallel one "recovered" battery string with an "unrecovered" one and running EQ. This higher Voltage process is hard on batteries and shouldn't be done any more than necessary.

Note: on your wiring diagram "Option 1" is preferable to "Option 2". In fact I wouldn't use #2 at all. When you connect two inverters to a common battery bank of any size they should be connected to the same points.

Artboard · November 2012

Re: 48V system charge settings?

Ok, in Option 1, should the 6048s still be configured as master/slave, with a 1600AH capacity, or should they be split up and run independently, i.e each 6048 has an 800AH bank capacity?

Cariboocoot · November 2012

Re: 48V system charge settings?

Artboard wrote: »

Ok, in Option 1, should the 6048s still be configured as master/slave, with a 1600AH capacity, or should they be split up and run independently, i.e each 6048 has an 800AH bank capacity?

They have to be Master/Slave if they are "stacked"; the AC outputs combined. If the AC is separate they can be considered two independent units.
This may be what affected the EQ cycle that didn't work: in a M/S config the Master unit would see its battery bank as fine and not allow EQ, keeping the second one from functioning in that mode. I'm not sure just how much communication/control goes on with the Xanbus but it is logical that could happen.

Artboard · November 2012

Re: 48V system charge settings?

That makes sense per what seemed to happen. AC outputs are combined, and originally both battery banks were EQ-ing solidly. When the Master topped off, it seemed to shut the slave down. Master continued to throw current into its bank, but the slave was cut off for some reason. I'll read up on Xanbus comms.

Now if I rewire as Option 1, I risk the same effect, i.e. the first 800AH bank gets topped off and the other dies on the vine.

When I left last weekend, I wired them as Option 2. When I arrived this weekend, the batteries seemed to be in pretty good shape, considering where they were the previous weekend. Note - during the week, there is almost no power used, and the batteries should be topping off with the new Absorb and Float settings. Although I understand the resistance discontinuity based on the wiring drawings in http://www.smartgauge.co.uk/batt_con.html, I still don't have the correct wires to optimize the resistance.

Are you sure the Option 1 is still the right way to go for this next week as I salvage the system with a reconfiguration/redesign?

Cariboocoot wrote: »

They have to be Master/Slave if they are "stacked"; the AC outputs combined. If the AC is separate they can be considered two independent units.
This may be what affected the EQ cycle that didn't work: in a M/S config the Master unit would see its battery bank as fine and not allow EQ, keeping the second one from functioning in that mode. I'm not sure just how much communication/control goes on with the Xanbus but it is logical that could happen.

Cariboocoot · November 2012

Re: 48V system charge settings?

It would be even more tedious, but you could just swap out whichever bank is on the Master inverter to make sure it gets charged & EQ'ed.

When the AC is coupled the DC has to be. If the AC is not coupled the DC can be but doesn't need to be. Can't run any other option than Master/Slave if AC is coupled. Essentially it turns the two units into one big 12 kW inverter which loads along providing power from the Master until load demands tell it to fire up the Slave unit.

Vic · November 2012

Re: 48V system charge settings?

ChrisOlson wrote: »

We have 4000-series batteries as well. When we initially had problems with low SG, Surrette told me to increase the absorb voltage to 31, minimum 3 hours time, with temp compensated controllers. They told me the lead-calcium grids in the 4000-series need higher charging voltage.

After three months of deficit charging (a brand new bank) with the same settings you posted, it took a month of running EQ's once a week at 32 volts, and absorbing at every opportunity at the new higher voltage setting to finally get them up to 1.265. They've been fine ever since, and that was 2 years ago.

If I had it to do over I wouldn't buy Surrette batteries again - I'd get Trojans. I don't really like the gawd awful voltages that the Surrettes require.
--
Chris

Have been meaning to respond to these comments about the Surrette 4000 series batteries.

From all that I know from Surrette, the two main series batteries -- the 4000 and 5000 -- are definitions of the nature of the battery case. The 4000s are single containment polypropylene case , and the 5000 is dual conatinment - polypropylene Jar and polyethylene outer case and replaceable cells, usually bolted together. The must be other differences in construction, as the warranties are different.

The spec sheets for any of the specific battery models that I've looked at do NOT spec the exact chemistry, nor do either of the Battery Manuals that have found on their site.

The C bank here are S530s, and are definitely Lead/Antimony batts. These 4000 series batteries charge fine with the standard, much lower Absorption voltages (than required by your 4000 series batts), as speced in the Solar Batt Manual, and have evidenced no stratification issues.

In another post, you mentioned that you are running T12-250s or a similar number. This does not appear to be a current model, but some of the TXX batteries still listed appear to be Marine batteires. Have seen no references to any of the Surrette batteries being Lead Calcium, although had wondered if their AGMs were.

So, not being one wanting to argue with you Chris, however, IMHO, many/most of the other 4000 series batteries are almost certainly NOT Pb/Ca, although it is difficult to prove from the data on the Surrette site.

I just am a bit defensive about Surrette. I have had very good service from Surrette Battery, and their products. And some members here who have had some problems with Surrette batteries have also had extraordinary warranty service from Surrette IMHO.

Anyway, just opinions, Vic

ChrisOlson · November 2012

Re: 48V system charge settings?

I don't know about any of that either, Vic, other than what they told me to do. This is their spec sheet for the batteries we got:
http://rollsbattery.com/public/specsheets/T12250.pdf

They also make a T12-250M marine battery with a black case. Ours are red case. After we got the SG up on them using the higher charge voltages they have tremendous capacity. Starting with the bank from 70% SOC my wife can load our new inverter to full rated load with her high-powered kitchen appliances and those T12's will hold the voltage at the inverter studs at 48.0 or better for two hours with no incoming power. At the end of those two hours it's down to 50% SOC and those batteries still hold the voltage at 48.0 or better under load.

Never seen another battery that can do that.

I will add that their charging efficiency seems pretty good up to about 80% SOC. But from 80% to 100% SOC it is quite low - probably about 80-83% efficient, or thereabouts. I think Trojans are better than that.

I will also add that with the bank now almost two years old they show absolutely no signs of reduced capacity from new. I don't have any sort of amp-hour meter to measure power out. But I know what I put in them with the Classics (that now show amp-hours with the newest firmware). I have been meaning to get a meter that measures amp-hours out (like a TriMetric maybe), but have never gotten around to it.
--
Chris

Vic · November 2012

Re: 48V system charge settings?

OK Chris,

Thanks for the Data Sheet on the T12 250s ... I did not search the Surrette site for the part number, just poked around. Had wondered if it might have been a Marine Starting batt, as the only Flooded Lead/Calcium batts that had seen before were for the Auto market. The DS does note that the standard terminal is "Automotive", with Flags optional ... altho with only 9 plates, they would not be a traditional starter ... dunno.

Glad that you have gotten on top of the battery recharge.

In my opinion, FLAs are quite efficient recharging to about 80-85% SOC, and the long grind through Absorption is where the efficiency drops significantly. Breaking down the water in the electrolyte into Hydrogen and Oxygen in the process of completing the charge is not efficient. Believe that this is why SLAs, particularly AGMs are more efficient in recharge. But, am no expert.

Also good news that you have found the NIB SW+ 5548. This is the a fine inverter, and surges better than almost any. I would not mind finding one as a spare. The prospect of brand S as replacements for my 5548s is not a comforting thought.

Good luck, Thanks, Vic

ChrisOlson · November 2012

Re: 48V system charge settings?

Vic wrote: »

Thanks for the Data Sheet on the T12 250s ... I did not search the Surrette site for the part number, just poked around. Had wondered if it might have been a Marine Starting batt, as the only Flooded Lead/Calcium batts that had seen before were for the Auto market. The DS does note that the standard terminal is "Automotive", with Flags optional ... altho with only 9 plates, they would not be a traditional starter ... dunno.

Our batteries have lead blade terminals.

I could have gotten 24 Trojan L-16's for about $1,600 less money. The dealer where I got them had a pallet of brand new L-16's sitting right there on the floor in the warehouse when I went to buy a new bank. But he told me the extra money for the Surrettes would be recovered because these T12's would out-cycle the Trojans because they don't have "soft" plates like a Trojan has. That remains to be seen, and he had to order the T12's for us. The shipping weight was almost 4,000 lbs and it took two weeks for them to come.

So I guess, so far, other than the high charging voltage I've been pretty happy with them. Although they are a pain to service because there's 144 cells to check water in.
--
Chris

Artboard · January 2013

Re: 48V system charge settings?

Since I started this discussion, I figured I'd update everyone as what I did with your advice. First, thanks to all of you for taking time to write your thoughts and experience.

1. I bought 2 longer wires and rewired the 4 banks as Method 4 in http://www.smartgauge.co.uk/batt_con.html.

2. I cleaned all terminals thoroughly and reconnected with electrical grease.

3. I made sure the cap vent holes all point away from the terminals.

4. Reset all bulk, absorb, and float voltages to the higher levels.

5. Equalized 2x-3x with Generator.

6. Picked up a Freas specific gravity tester. My autoparts SpGr tester was reading lower than the Freas.

Batteries are fully charged (with the exception of one cell in one battery that will not get higher than 1.200). System supported family of 5 between Dec 30 and Jan 9, with short solar days. Brought electric hot tub back to temp with Generator. Solar system sustained hot tub temps and cycling no problem.

Will pull the one battery from the bank, temporarily reduce the system capacity by one bank, and try to recover that one cell.

Water consumption is nominal now.

Any other thoughts? Pretty happy now.

Artboard · June 2013

Re: 48V system charge settings?

Cariboocoot wrote: »

Battery bank: four parallel strings of 400 Amp hour batteries; total 1200 Amp hours. If those four strings are not connected correctly current sharing will be uneven on both charge and discharge. See the Smart Gauge diagrams here: http://www.smartgauge.co.uk/batt_con.html

Also this bank wants 120 Amps peak current @ 48 Volts, which needs about 7.5 kW of array. You've got 6480 Watts. It should work, but is a little bit low. Probably maximum current is around 100 Amps.

Are you sure those charge controllers aren't 80 Amps each? Limiting the total charge to 80 Amps would be just a 6% peak charge rate. Add some loads while charging and you will have deficit charging periods.

At 58 Volts the Absorb setting is a tad low. The time of 2 hours is also suspect, depending on how deeply you discharge the battery. If you can measure the current to the battery during Absorb with no loads do so. If it remains above 3% of the battery capacity (36 Amps) the Absorb time is too short.

You need to fully charge and EQ those batteries manually until you see no more improvement in cell SG. Otherwise what you are doing is turning expensive batteries into scrap metal. 62 Volts should be sufficient for EQ, and leaving off that half Volt may keep the XW from dropping the gen. Normally when EQ is needed and you're short on sun the thing to do is Bulk up with the gen ASAP in the day and let the solar deal with the EQ.

Couple of suggestions you probably won't like:
1). Consider replacing those charge controllers. If they really do limit your panels to only 80 Amps total this is indeed a waste of panel.
2). Re-evaluate your load requirements. Do you actually need 28 kW of stored power? That is quite large. Considering your short solar day you might benefit from a higher panel to battery capacity ratio. Something along the lines of 800 Amp hours instead of 1200 may make all the difference (providing you can get full charge current).

I fixed the wiring per the smartgauge reference.

The Apollo charge controllers are arranged in a master with two slaves arrangement. I think each is capable of 80A in a MPPT config.

For now, I have disconnected one bank of batteries, so instead obeying to charge 4 banks of 48V for 1600 AHr, I am now charging 4 banks for 48V to get 1200 AHr.

I am contemplating adding more panels. How many more should I add to charge my 1600 AHr worth of batteries, and how might I configure them into my existing breakers and charge controllers? Will I need another charge controller as a new slave, too?

Appreciate all ideas.

Cariboocoot · June 2013

Re: 48V system charge settings?

In "rule-of-thumb" mode:

1600 Amp hours @ 48 Volts would have 160 Amps of charge current peak. That would require two 80 Amp controllers or three 60 Amp controllers and a total of 9,974 Watts of panel divided between them. That's a lot of panel.

Artboard · June 2013

Re: 48V system charge settings?

Correction: I am now charging 3 banks of 48V for 1200 AHr. One bank of 8 batteries sits idle. How long is wise to let them sit idle?

Artboard · June 2013

Re: 48V system charge settings?

So another 24 more Kyocera 180W panels? Can I stay with my current three 80W charge controllers?

NorthGuy · June 2013

Re: 48V system charge settings?

Artboard wrote: »

So another 24 more Kyocera 180W panels? Can I stay with my current three 80W charge controllers?

I'm sure people could say more if you had your configuration - panels, controllers etc. in your signature.

~250W panels are usually less expensive and require less brackets to install than ~180W.

YehoshuaAgapao · June 2013

Re: 48V system charge settings?

NorthGuy wrote: »

I'm sure people could say more if you had your configuration - panels, controllers etc. in your signature.

~250W panels are usually less expensive and require less brackets to install than ~180W.

150V MPPT charge controllers favor 60-cell (very common) panels in strings of 3 or 96-cell (ebay/china direct only) panels in strings of 2. 80-cell in strings of 2 will work if it don't get too hot (friday will be 118 in PHX, Saturday 117). 72-cell in strings of 3 will work if it never gets cold (like Indonesia).

It is my guess your 185W panels are 48-cell and in strings of 4.

More PV, even a little too much PV isn't bad. Panels are cheap and getting cheaper. Batteries are expensive and getting more expensive. Inverters and charge controllers appear steady in pricing change.

You can use different brand/size solar panel as long as they each set of brand/size are confined to their own charge controller. Not sure about effect of shared negatives between multiple charge controllers and how that will impact voltage mismatch. Probably be wise to keep the negative separate. Even my panels with different mounting angles and positions are MPPT-ing at different voltages (particularly flat panels vs the patio; even some difference between patio & roof due to temp difference)

YehoshuaAgapao · June 2013

Re: 48V system charge settings?

Artboard wrote: »

So another 24 more Kyocera 180W panels? Can I stay with my current three 80W charge controllers?

Assuming an optimal string size (4 for 48-cell, 3 for 60-cell, 2 for 96-cell), 80A charge controllers can take 5 strings each (45V). 60A is 4 strings (48V). 48-cell panels are around 180W, 60-cell around 240W, 96-cell around 360W, 72-cell around 270W, 80-cell around 300W for polycrystalline.

48V system charge settings?

Comments

Categories