Parallel XW inverters not sharing the load equally

thorbjornw
thorbjornw Registered Users Posts: 14 ✭✭
I have installed a 50kW PV system with 10 MPPT80-600 charge controllers and 3 XW+ 6848 Split-phase inverters in parallel (1 Matser and two slaves) and a 144 kWh battery bank. As there is a 60A limit on the internal transfer switch of the inverters, I have set it up with two external contactors so the system is either completely off-grid (when the battery voltage is over a certain minimum) or with all the load transferred to the grid (when battery voltage is low). To monitor the system we have a battery monitor and the XW combox.
After changing some faulty communications cables the system is now running with no error messaged and the combox gives a nice overview of it. But there is a serious problem with the sharing of the load between the three inverters. When we switch over the load (around 15 kW) from the grid to the inverters, around two thirds of the load falls on the Master while the last third is distributed more or less evenly on the two inverter slaves, bringing the load on the Master well above the 6.8 kW rating. The three inverters have inverting enabled, while charging, grid-support and sell are disabled (all the charging is done by the PV).
What we have tried to do is (i) check the AC sync cables (even if there is no sync error meesage), (ii) shift the role of master to one of the other inverters (but the problem is repeated with the new master), (iii) recheck the AC out wiring.

Comments

  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,730 ✭✭✭✭✭✭
    I did a similar size system down in Costa Rica and I will say it is much easier now so hang in there. Is this a home please and new equipment?
    Not much time for this but a few points for now. All similar devices must have different address numbers.
    Combox has alot of new features and should look like below, if not update it because there is firmware now that will prevent the old issues.
    Your external contractors have me concerned because you are in the logic now and not the XW+
    If you have not done firmware upgrades before, please state that as it does take time to get up to speed.
    Good Luck and be careful!




    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • thorbjornw
    thorbjornw Registered Users Posts: 14 ✭✭
    This is a hotel and our normal load is 8-12 kW with peaks of 17-18kW (and lows of 2-3 kW), so with our 3 times 6.8 kW inverters that should give us sufficient margin. All the equipment is brand new. The role of the contactors is to shift the whole load to the grid when the battery bank reaches a critically low level, and then back to solar when the sun is shining and the solar panels have charged the batteries a bit. The idea is to harvest sun energy by day and dispatch it by night as most of our consumption is by night. I designed it as this because of alerts by Schneider Electric that when you have 3 or more inverters, external contactors are necessary because of limitations of the XW inverters (some problem related to the by-pass relays not opening exactly at the same time so this wears down the internal contactors). It is true that using external contactors we lose a lot of options for loadsharing with the grid, but as the electricity company does not accept (or pay for) excess current (yet, that may change soon), this has up to now not been important.
    Anyway, what I have done now is (i) update all the firmware (not sure that really has achieved anything important), (2) rechecked the wiring once again (no error detected), (3) rechecked the output AC current of each inverter. The latter gave a possible clue: one of the inverters had a significantly lower output voltage. So I reset that one to factory defaults and configured the whole system once again with the XW configuring tool. The other thing I did was to disconnect the system completely from Grid input. Miraculously, the system is now working and sharing the load nicely. I had planned to calibrate the output current, but it turns out the XW configuring tool, I bought, does not have this option activated (I suppose it is limited to professional installers). But for the moment, I am happy. The load is shared with a variation of less than 10% deviation, as recommended by Schneider. So I cross my fingers.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,730 ✭✭✭✭✭✭
    That is good to cross fingers :)  Since it sounds like it may not stay stable,  consider that you are past the number of devices Xanbus can reliably handle. Typically that number is 20 devices. Each XW+ takes up 3. Each Mppt or other device is 1. In large systems like yours and even larger, the correct way to monitor, update, and use the system is to master slave the Combox. The config tool is OK but Combox will make it very easy with the pre-set charts and graphs and easy recovery of a corrupted device (lightning, power outage, or dumb things that we all sometimes do.)  Very powerful for a system like yours with custom logs. Drop me an e-mail, sounds like you did a good job. --Dave
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • thorbjornw
    thorbjornw Registered Users Posts: 14 ✭✭
    Well, according to the engineer from Arizona Wind and Sun who sold me the system, the number of devices I have is close to the Xanbus limit, but should just work. It does, but a bit slowly. I shall consider the combox master-slave solution, but as it means buying one more combox and set it up as slave, is not that cheap, so I will give it a chance to work as it is now. Regarding the system you installed in Costa Rica, with 3 XW 6848 inverters without external contactors, does it work well? According to some notes from Schneider I found on the net, the risk is that the internal contacts will wear out. The engineer from AW&S apparently did not know that, but after he investigated it, he agreed and proposed me an external contactor set-up, which is the one I have used (with some modifications). Actually, the only equipment not working well is the Battery Monitor. Seems it is not well integrated into the Xanbus network with some strange behaviour. I will look into it, when I have secured the functioning of my Nickel-Iron bank. It is not easy to make the initial charge of the bank as it should be (overcharging with 200 Amp a couple of times) as the MPPT80-600 CC have a lot of limitations (I guess to protect the owners of Lead-Acid batteries).
  • mike95490
    mike95490 Solar Expert Posts: 9,583 ✭✭✭✭✭
    50,000 x .8 = 40000w @ 64V  should give you about 600A of solar power at the combined outputs of all the controllers.

    What voltage are you charging the batteries at , should be close to the XW limit of 64V right ?

    Powerfab top of pole PV mount | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
    || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
    || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

    solar: http://tinyurl.com/LMR-Solar
    gen: http://tinyurl.com/LMR-Lister ,

  • thorbjornw
    thorbjornw Registered Users Posts: 14 ✭✭
    We have just started the system up, and unfortunately we have cloudy and rainy weather, which is unusual here. But when the sun shines, the production is around 30-35 kW (around 500-550 Amp), but going up to 40-45 kW occasionally (around 700 Amp). We have wired and fused it for up to 1,000 Amp. I have set the bulk charging to 64V, which looks as the maximum, despite that the specification sheet for the MPPT80-600 says 67V. Now my battery bank is unusual, as it is Nickel-Iron, and I would like to charge it hard to get it initiated (66V) - unlike Lead-Acid the NI-FE needs tough charging. But it seems the charge voltage is limited to 64V. So to get my batteries working well, I may need to take 1 or 2 batteries out of each bank to cheat the charger. Strange.
  • mike95490
    mike95490 Solar Expert Posts: 9,583 ✭✭✭✭✭
    I too have a NiFe bank, and the XW gear can only supply 64V.  It redlines at 67v and faults out at 68V.   I use a morningstar and classic solar controller and can get to the 67V i need via them.  When it's rainy, I can charge, but only partially, via the XW inverter.  if you pull batteries from the string, then you start having trouble at the low voltages.
     A longer, slower commissioning charge will not hurt, but it may take several weeks to accomplish at lower voltages.   Set Absorb times for 3 or 4 hours, and have lots of distilled water on hand for refilling batteries.
    Powerfab top of pole PV mount | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
    || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
    || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

    solar: http://tinyurl.com/LMR-Solar
    gen: http://tinyurl.com/LMR-Lister ,

  • new2PV
    new2PV Solar Expert Posts: 305 ✭✭
    edited July 2017 #9
    Could you rewire the xanbus to divide the devices better? I think this is causing your BM not to work properly.
    XW6848 inverter with 2 X mppt 60 150 CC , with Canadian solar 260Watt panels 2 x 3.5 kw array
  • thorbjornw
    thorbjornw Registered Users Posts: 14 ✭✭
    Seeing the behaviour of the system, I reckon that taking one or two batteries out might solve the charging problems, but would get me very quickly below the 40V minimum required by the XW inverters to operate. So I agree, this would be a bad solution. For the moment, I will just let the system continue to work and see if the batteries firm up. The problem as it is now, is that they get too quickly down to the 44V that I have put as minimum for my external contactor to cut the system out and switch to the grid. I have put absorb time to 6 hours (but as they switch out of bulk around 11 am, and charging after 4 pm is minimal, it is in reality 5 hours).
    They do need a lot of water. I have installed a deionizer with a watering gun, which makes watering a bit easier. But still, 120 batteries require around 2-3 hours work each time for our maintenance guy.
    Regarding the battery monitor, it is probably because of too many devices on the network (3 inverters and 10 Charge Controllers). So I have ordered one more ComBox to set one up as master and the other as slave. I assume this will solve the issue.

  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,730 ✭✭✭✭✭✭
    The master/slave  config of combox  will solve the reliability issue with xanbus only. I had a much different system design when I did this. Schneider likes to use the contactors for big systems, I don't.    As I said before drop me an e-mail.
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • mike95490
    mike95490 Solar Expert Posts: 9,583 ✭✭✭✭✭
    Be sure your water DI system has 2 cartridges in it, so you can run one down to exhaustion, and then leapfrog the 5% used to the raw water inlet and always have a 100% cartrige at the outlet, one dose of bad water will ruin the bank,
    Powerfab top of pole PV mount | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
    || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
    || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

    solar: http://tinyurl.com/LMR-Solar
    gen: http://tinyurl.com/LMR-Lister ,

  • thorbjornw
    thorbjornw Registered Users Posts: 14 ✭✭
    I have two spare cartridges. There is a light indicator on that turns red when it is not working (I think it measures the resistivity of the water). Should I be without deionized water, I guess I would have to buy distilled water (or buy a water distiller).
  • thorbjornw
    thorbjornw Registered Users Posts: 14 ✭✭
    Schneider tells me that to get past the 64V charging limit, would require new firmware, and firmware updates are decided by the engineering department. They advice med to get more people to join and write to them asking for increasing the 64V charging limit. It would be nice to know if I am the only one interested in this. Seems some serious lobbying is needed. The latest firmware update for my Midnite Classic includes the option of Nickel-Cadmium batteires, for which the ceiling for the charger is 66V. So if Midnite can do it, it shouldn't be that difficult for Schneider either.
  • westbranch
    westbranch Solar Expert Posts: 5,183 ✭✭✭✭
    The bigger they are the more they head for the cheapest product and lowest common nominator ie the value that most people want 64...  the power of diminishing marginal returns for $$ inivested, not being a leader of the pack or on the bleeding edge...!
     
    KID #51B  4s 140W to 24V 900Ah C&D AGM
    CL#29032 FW 2126/ 2073/ 2133 175A E-Panel WBjr, 3 x 4s 140W to 24V 900Ah C&D AGM 
    Cotek ST1500W 24V Inverter,OmniCharge 3024,
    2 x Cisco WRT54GL i/c DD-WRT Rtr & Bridge,
    Eu3/2/1000i Gens, 1680W & E-Panel/WBjr to come, CL #647 asleep
    West Chilcotin, BC, Canada
  • Estragon
    Estragon Registered Users Posts: 4,496 ✭✭✭✭✭
    Total speculation on my part, but my guess is Midnite engineering is considerably more "customer facing" than Schneider, just as a function of size.

    More than once, I've gone through the circles of hell that's support, with escalations, bad music etc., with an issue I understood pretty well, but the support people were pretty much deer in headlights on. I don't blame the support people (I used to run a business with tech support and engineering functions), as the issue was pretty technical. Someone in engineering would understand the issue in a 2 minute conversation, but I was told point blank that "engineering isn't a customer facing function". I get that you don't want expensive engineers dealing with routine support issues, not to mention having them get really pissed at having to deal with know-nothings being critical of their work instead of actually working on engineering problems, but ideally there should be a way to get real world issues to engineering without corporate filtering.

    In a perfect world, big companies wouldn't be prone to these information silos, but we don't live in a perfect world.
    Off-grid.  
    Main daytime system ~4kw panels into 2xMNClassic150 370ah 48v bank 2xOutback 3548 inverter 120v + 240v autotransformer
    Night system ~1kw panels into 1xMNClassic150 700ah 12v bank morningstar 300w inverter
  • mike95490
    mike95490 Solar Expert Posts: 9,583 ✭✭✭✭✭
    Schneider tells me that to get past the 64V charging limit, would require new firmware, and firmware updates are decided by the engineering department. They advice med to get more people to join and write to them asking for increasing the 64V charging limit. It would be nice to know if I am the only one interested in this. Seems some serious lobbying is needed. The latest firmware update for my Midnite Classic includes the option of Nickel-Cadmium batteires, for which the ceiling for the charger is 66V. So if Midnite can do it, it shouldn't be that difficult for Schneider either.
    Do you have contact info for passing +64v requests to Schneider ?
    Powerfab top of pole PV mount | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
    || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
    || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

    solar: http://tinyurl.com/LMR-Solar
    gen: http://tinyurl.com/LMR-Lister ,

  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,730 ✭✭✭✭✭✭
    It never hurts to ask! The problem is that when working voltages are picked in the design phase, the components and their "headroom"
    are baked in. Not easy to just widen up when the future is going the other way (lower voltage). The system I am running for them is all at 57.5volts.

     
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • thorbjornw
    thorbjornw Registered Users Posts: 14 ✭✭
    Yes. This is the link to put in requests:
    https://solar.schneider-electric.com/feedback/
    I have already written them. They say they will answer in 24-48 hours. So I cross my fingers. It would be very good if all others who have the need for higher in voltage could write them so they can see, we are some that have the issue.
    I am fully aware the Li-ion batteries don't need high voltage, and that is probably the future. But we are not really there yet, if it is for daily cycling. For back up it is fine. I have bought two Lifepo 100 Ah batteries with BMS from China for back-up (around 360 USD/kWh), and they work fine. But I wouldn't try to cycle them daily.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,730 ✭✭✭✭✭✭
    The link is for down the road, next build type requests.
    You need something now and I would try to get a firmware specific request in. They may do it if you will accept a reduction in the 5 year warranty. 

    There are plenty of people cycling multiple times per day with the LG (me for one) and for years in Austrailia/New Zealand and here on the forum with homebrew Prismatic cells.
    All you have to do is look at the efficiency and while your application is not there yet off the shelf, there are some very nice Li-ions in the store here from SimpliPhi.  All one needs to do is look at the chart below. 



    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • thorbjornw
    thorbjornw Registered Users Posts: 14 ✭✭
    I am aware that Ni-Fe have a low round-trip efficiency. But as I can produce energy quite cheaply (around 5.5 cents of a USD per kWh for the whole system and considerably less than that for marginal increments), I have just put in some extra solar panels. The weak point in a system that wants to compete with the grid is the storage. I looked at both SimpliPhi, LGChem, BYD B-box etc. but price per kWh was still too high to make it worth while. I even looked at Lithium Titanate Oxide (LTO), which are very promising with a very long life span. Again, too expensive.
     I got the Ni-Fe batteries directly from the factory in China for 330 USD/kWh. So if they work, it is a good investment, as electricity prices here in Nicaragua are high (21 cents of a USD per kWh).
     My other concern was decommissioning - there is no guarantee I can recycle the batteries here in Nicaragua and NMC has Manganese and Cobolt, which I understand should not just go to the municipal dump. I am sure the buyers of scrap metal here will be happy to take in the Ni-Fe batteries when there time comes. They are accustomed to both iron and nickel.
    But again, I agree the future is with Li-Ion.
    Do you btw. have a link for putting in a special request to Schneider for a more urgent solution? The 5-year guarantee is not much worth for me down here anyway, as I will not be sending the equipment back to the US in case of failure, so I don't bother to forgo that.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,730 ✭✭✭✭✭✭
    edited July 2017 #22
    I was just commenting that there are quite a few ways to go now Offgrid, certainly not in a really large system like yours.
    I did offer to send my contacts in Engineering an e-mail but you never said to?  I certainly will not put their e-mail addresses here on the forum.

    Our prices for electricity are close to 20 cents per KWH down in the Central Valley.  Grid storage does not make sense there financially even with the Cal SDIP incentive. it may down the road as California can do some crazy things. I am quite happy off their grid  B)

    What firmware revision is on your mppts?
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • thorbjornw
    thorbjornw Registered Users Posts: 14 ✭✭
    I have updated it to V01.02.00-0003 :):)
  • thorbjornw
    thorbjornw Registered Users Posts: 14 ✭✭
    Finally some good news. Schneider has made an update of the software permitting charging at 67V - this is not a public release yet. The version is V01.05.00-0001.
    I have just upgraded my 10 controllers. It gave me some difficulties for reasons I don't quite understand, as they wouldn't charge after the upgrade with no faults or warnings. However, I reconfigured the system to default values and once they were up and running I changed the parameters. So now they are working. I shall now try to give my batteries the initial charge with 67V - after they have been working for 4 months and only reaching around 50-60% of their capacity in the daily charge-discharge cycles. I hope it will improve the performance of the batteries. B)
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,730 ✭✭✭✭✭✭
    Good! They told me about you getting the firmware. I have copy also for another client. They may or may not release this for the reasons we discussed in the thread here. Hope this works out for you!
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • thorbjornw
    thorbjornw Registered Users Posts: 14 ✭✭
    Unfortunately I had to downgrade again. The chargers failed consistently to start in the morning again without resetting and reconfiguring the whole system once more. The downgrade solved immediately the problem - the ten 10 CCs started up one by one as the downgrade proceeded, without any need for reconfiguring or even reboot, and now they work as before. But limited to 64V as before. If you have better luck, please let me know.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,730 ✭✭✭✭✭✭
    Tell them please. They probably did not test them with so many devices. I know we talked about running up against the limit a combox can control. I am not saying that is the problem but... What a pain!  At least you and work on your suntan there!
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • thorbjornw
    thorbjornw Registered Users Posts: 14 ✭✭
    Quite sure it has nothing to do with the number of units, as I am well below the max for the Xanbus network. I also tried with only one CC connected to no avail.  I could get the system up an running with the new firmware for the CCs, but only after reconfiguring it resetting to factory default and after that changing the parameters for the custom battery settings. The CCs still curtailed themselves at around 63.4V (despite both bulk and absorption being set to 67V). Even when set to 64V the CCs would not start charging in the morning, just resting in operation mode with slow blinking, To get them to work a new reset to factory defaults would be necessary. So finally I downgraded the firmware again and they are working normally again.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,730 ✭✭✭✭✭✭
    edited January 2018 #29
    As i said, tell them! The system I saw worked fine. Work on your tan. We loved that place and the mistake we made out cruising was not to stay longer. Can you still anchor nearby?

    I can bug Engineering again if you want?
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net