Xantrex System Failures

glasstone2 · March 2013

I have a problem with a Xantrex installation, and I would appreciate some tips, hints, and any information from users here. I'm new to this Forum, but I have looked here from time to time over the previous year for a variety of information; so I'm aware that there is a good amount of technical expertise available.

An installer company from Austin, TX has been installing my off-grid Xantrex System for over six months, but I'm still on the grid with a White Elephant of an installation. At this point, I have four Xantrex XW6048 Inverters that should be able to supply up to 24kW for peaks. They are connected to 36 Schuco PS08 panels (240W each), and 2 48V banks of Trojan L16RE batteries. My typical usage is low running 1kW to 2kW average over the day, but the HVAC, freezer, water well, and especially a clothes dryer can cut in getting me near to 24kW for relatively short periods.

The installer intended to turn the off-gird system over to me over a month ago, but when they brought up the 4 inverters, 3 were supplying AC power to the house loads and the 4th was charging the batteries from the AC output of the other three--Perpetual Motion!

When the installer showed up 2 weeks later confident of turning the system over to me, he put the Inverters in what he called "search" mode. The load test stated up promisingly. The master inverter started supplying the load (a dryer and water heater); as we increased the load, two more inverters started taking up load including the inverter that had been charging the batteries in the previous test. We added an oven to bring the load up to about 6-7kW which should hardly tax the 24kW system--that's almost just 1 Inverter's worth. The 4th Inverter showed signs of life and started to come up. Immediately, the red fault lights on all four inverters began flashing. This continued for 10 or 20 seconds until the 4th inverter said it was supplying some of the load, and immediately the whole system crashed showing an F48 fault.

The installer's call to Tech Support at Xantrex was interesting. The Tech had him check the firmware: only one inverter had the current firmware. That is the first hurdle that has to be rectified. My installer had to order the Schneider XW Configuration Tool, as he was unaware of its existence (I knew about it). So now the System is sitting idle for an indeterminate time again while the installer figures out what to do.

My concern is the Under Voltage fault (F48 ). It might be an artifact of the 4 inverters running different firmware, but I'm concerned that I have a deeper problem. I figure there could be a problem with my battery banks and more seriously, I'm concerned that the 4 XW6048s don't play nice together with a properly functioning battery bank, and I can expect crashes whenever I start to put a fair load on the system. Can anyone provide insight into this problem?

I apologize for this long post. There is a vastly longer story behind this failed installation. I'll be glad to answer any questions on my System, and supply more information as needed. Thanks.

BB. · March 2013

Re: Xantrex System Failures

Wow--That is a huge system for off grid...

First, the battery bank... is that using Trojan L16RE-2V @ 1,110 Amp*Hours cells (two parallel strings of 24 cells in series)?

That would be a 2,220 AH battery bank--And would be just under the 2,400 AH @ 48 volt system that we would recommend as the minimum battery requirement for a 24 kWH system (100 AH per kW of inverter load or solar array).

So, the battery bank is a wee bit on the small side--But should not be the issue here. However, if these are 6 volt cells (different model) of 225 to 375 Amp*hours each-That would be a seriously under powered bank for a 24 kW of loading.

What is your electrical wiring? A 24 kW system should be wired up to manage:

24,000 watts * 1/0.85 inverter efficiency * 1/42 volts minimum battery voltage = 672 Amps maximum continuous current
24,000 watts * 1/0.85 inverter efficiency * 1/42 volts minimum battery voltage * 1.25 NEC derating = 840 Amp rated minimum circuit design

I would check the battery voltage with a good quality volt meter at the battery bus and at each inverter. See if the under voltage fault is "real".

You could also use a DC Current Clamp Meter (like this one) to make sure the two battery strings are each sharing their current (roughly) equally.

Sorry to hear that the inverters were down rev... That "implanter" has gotten quite a few people.

Regarding your solar array, if I guessed correctly on your battery bank capacity, you are running around 5% charging current from your solar array--About the minimum we would recommend for that size battery bank.

This is about as far as I can help you--I do not have any experience with off grid system configurations to give any hints there.

-Bill

Cariboocoot · March 2013

Re: Xantrex System Failures

I don't do Xantrex (OP would be one reason why) but I see three other potential problems in addition to Bill's comments on the battery bank:

1). If all inverters' outputs are coupled the DC side needs to be as well; do not use separate battery banks.
2). The AC wiring may not be right: sounds like one inverter's AC IN is connected to the others' AC OUT
3). Programming may be wrong (in addition to firmware upgrade needed). Sure hope it's got the SCC in that mix somewhere.

Also, 8640 Watts of panel would be about 138 Amps of current needing at least three charge controllers and being able to support only about 1380 Amp hours of 48 Volt battery - far below the 2400 Amp hour minimum Xantrex recommends for that much inverter. I don't think this is as critical for off-grid application, but it sure sounds like someone was guessing about components.

Trojan 6 Volt L16's come in a variety of Amp hour capacities from 320 to 390 (possibly more - can't remember them all). If there are only two strings of these you've got like 640 Amp hours; barely enough for one inverter. Can you give us more details about the batteries?

NorthGuy · March 2013

Re: Xantrex System Failures

glasstone2 wrote: »

The installer intended to turn the off-gird system over to me over a month ago, but when they brought up the 4 inverters, 3 were supplying AC power to the house loads and the 4th was charging the batteries from the AC output of the other three--Perpetual Motion!

If everything is wired correctly, this situation is impossible. Unless you want to achive some special effects, all four outputs must be wired together, and since this is off-grid, there should be nothing on the input of inverters except generator(s).

glasstone2 wrote: »

This continued for 10 or 20 seconds until the 4th inverter said it was supplying some of the load, and immediately the whole system crashed showing an F48 fault.!

When you apply loads on batteries, the voltage of the batteries will drop. If it drops too much, below so called LBCO (Low battery cut-out voltage) you get fault 48. It may be because LBCO is set too high (e.g. 47V). You can decrease LBCO, and it can go as low as 40V. This will prevent the fault 48. However, if LBCO is already low (say 45V), it could be that the load is too big for the batteries. It depends on the battery bank size.

How big are you battery banks (in Amper-hours)? How are they shared between inverters?

westbranch · March 2013

Re: Xantrex System Failures

the following is from the MN tech people in response to a question about 2 banks and 2 linked Classic CC's

Yes this will work as long as the battery negatives are both common to each other. You would run the Follow Me cables but not turn on Follow Me, this will allow you to address the Lite so it is viewable from the Classic. But the 2 battery banks MUST have common negatives.

They just need to be at the same potential (IE no voltage present between them) as long as they are bonded in one place (IE on the battery side of the shunts or through 2 separate GFP devices life will be good)

Are your batteries bonded?

Cariboocoot · March 2013

Re: Xantrex System Failures

westbranch wrote: »

the following is from the MN tech people in response to a question about 2 banks and 2 linked Classic CC's

Yes this will work as long as the battery negatives are both common to each other. You would run the Follow Me cables but not turn on Follow Me, this will allow you to address the Lite so it is viewable from the Classic. But the 2 battery banks MUST have common negatives.

They just need to be at the same potential (IE no voltage present between them) as long as they are bonded in one place (IE on the battery side of the shunts or through 2 separate GFP devices life will be good)

Are your batteries bonded?

I didn't see any mention in the OP about MidNite charge controllers.

This seems to be basically an inverter problem, and I'm pretty sure Brand X like Brand O wants all inverters to share the DC (positive and negative) is the AC outputs are coupled.

Frankly it looks like there's more than one problem here.

Vic · March 2013

Re: Xantrex System Failures

Just as a side note.

Trojan recommends a commissioning EQ for their batteries -- this is done after a complete Absorption cycle.

In addition to all of the above, the battery bank needs to be commissioned. Some batteries from distributors arrive at the user's site with 50=75% SOC, and should be commissioned immediately, IMHO.

Wish you well. Schneider seems to never update the FirmWare of any of their products that are in their warehose, or in distribution, expecting their customers to do it for them. Wish you the very best with this large system. Vic

glasstone2 · March 2013

Re: Xantrex System Failures

Wow, I was gone for a few hours, and there is considerable information on this Thread. Thanks to all!

I'm going to try to answer the questions so far here. Let me know if I miss something and need to cover it:

The batteries are the 2V Trojan L16RE in two 48V banks for 2220AH as correctly mentioned by BB. The batteries are arranged into two 48V battery banks with 24 batteries per bank. I believe the two banks are wired in parallel at the inverters, but don't hold me to that. We do have the 3 charge controllers (XW MPPT60) for the solar panels deduced by Carlboocoot, and after 6 months, I agree with his "guessing" comment. BTW, I did specifically ask my installer if these banks were sufficient to support these 4 inverters and was assured that they are. Whether or not that is true is something I'm trying to get at to address these troubling Faults.

As NorthGuy suggests, there may be an Inverter wiring problem giving rise to the Perpetual Motion scenario. As far as I can tell, the installer has not detected and corrected any wiring problems. Sometime, I'll have to tell you about the initial wiring at the circuit breaker box supplying 1/2 of our house that floored all of us (back in October). I will ask about the LBCO as I expect our installer simply took the defaults on these inverters especially considering how he failed to even look at the firmware version while installing them. However, I think NorthGuy is getting to the heart of one of my concerns which is will the 4 Inverters load the batteries enough so as to fault and give me frequent crashes in normal use.

I have to look deeper into the questions by westbranch as I'm unfamiliar with bonded batteries. If I can figure out how to post pictures to this forum, I could take pictures of critical components and systems so that people with better knowledge and experience can see how wiring etc. are physically set up. Let me know if that would be useful, and what you need to see.

As to commissioning the batteries: I don't know what the installer has done exactly. These batteries have been hooked up and attached to at least 1 inverter since October 2012. On occasion, the whole system has been turned off when it was obvious that it is not behaving properly like charging the batteries from the AC output of the other 3. Right now, 2 Inverters are running to maintain the charge on the batteries. We have never been able to actually run the house on this system. I only very recently received instructions from my installer on how to maintain the batteries even though I have asking for that guidance since October.

I think I've touched on everything so far. Let me know if otherwise. BTW, I'm not going to tear into the actual equipment much more than I can do visually or change things on my own. I don't want to make the case that I'm the cause of the System failure by tinkering! However, there is a lot I can get to without tinkering. To anticipate some future questions: This IS a large off-grid system supplying two buildings comprising our house (basically the house is 1 story with the virtual 2nd story being an attached, but electrically isolated (separate feeding circuit breaker panel etc.) building--each building can be taken off the grid separately). Before installing the Solar Power System, we retrofitted the HVAC with a Geothermal Heat Pump System--2 tons on one building and 3 tons on the other. Our HVAC installer which is part of the company installing the Solar Power System, tells me that our Geothermal HVAC will consume, at most, 4.5kW.

Thanks again.

BB. · March 2013

Re: Xantrex System Failures

I believe, that on rare occasions in the past, Xantrex (Schneider) has sent an engineer out to locations with unresolved issues. After the firmware update and you measuring battery bank voltage (and voltage at the inverter DC input) while loaded--If you are getting below ~46.0 volts at the battery bus or the inverter input(s) while loaded--Then you may have a battery bank/wiring issue (too light/long of wire runs, batteries not outputting enough current for whatever reasons, etc.).

If the firmware does not "fix" your problems and you (and your installer) cannot see any "obvious" issues--I would ask the installer to seek more engineering support.

Otherwise, take photos and write up your notes/story and contact Schneider/Xantrex for help and see what they can/will do.

You do not want to get into the hardware details--That size battery bank is deserving of respect. Typically measuring voltages and using a DC Current Clamp is a pretty safe thing to do (just be careful with the volt meter leads).

You should have a log of Specific Gravity readings and volt readings (per cell) too. You have a lot of "identical" cells/parallel wiring runs. Any "differences" during heavy charging/heavy loads, etc. is usually worth looking at (things will not be identical, but they should be pretty close--cell to cell voltage--differences of less than a 0.10 of a volt, if current sharing in the battery bank is over ~20-50% difference between high/low parallel string, etc. do matter).

-Bill

Cariboocoot · March 2013

Re: Xantrex System Failures

Should I point out that an XW6048 is capable of a 12kW peak on its own, so four of them could power 24kW continuously with a peak to 48kW? You just might have two inverters too many as it is.

Batteries: 2220 Amp hours, eh? The panels will have a hard time charging that much; peak rate around 6%, which is barely above minimum. Especially as Trojan specifically recommends 10% peak charge rate.

So you could get up to about 48kW hours out of them. Do you need that much? It's three times what I use here in town. And do you really have combined loads that will demand 24kW all at once? That's two electric stoves on full. That battery bank is only good for about 9.6 kW "hit" all at once.

In theory each inverter should have its own DC breaker. Shut two of them off and see if things don't go better as a result. Although the firmware needs updating no matter what else is done.

mike95490 · March 2013

Re: Xantrex System Failures

westbranch wrote: »

.... Are your batteries bonded?

Just the housekeeper..... I didn't know there are issues with the batteries. When cells go bad.... 8)

westbranch · March 2013

Re: Xantrex System Failures

you're losin' it there Mike....somethin' about your former self... must be galloping old age...

inetdog · March 2013

Re: Xantrex System Failures

NorthGuy wrote: »

If everything is wired correctly, this situation is impossible. Unless you want to achive some special effects, all four outputs must be wired together, and since this is off-grid, there should be nothing on the input of inverters except generator(s).

Since this is only optionally off-grid (and I will not go into the economics of that kind of system now!), the AC inputs of all 4 Xantrex would be connected to the grid unless the transfer switch(es) were flipped for the test. If there were some incorrect connections in the transfer switch, transferring them all could have resulted in one of the AC in wire pairs being connected to the grid (showing powered from grid) or from the AC out of one of the other three inverters. It would not have started to charge until the battery voltage dropped under load, and when it finally switched to inverting it might have come on out of phase with the other inverters because of the wiring error instead of in phase, causing who knows what. Is the same inverter always the one that is fourth and causes problems? Turn one inverter off of the other two (not the master, and open its AC-in!) and see what happens when the questionable one comes on as third.
As Northguy said, unless something is powering the AC-in of the fourth inverter (grid or other inveters), there should be no way that it can charge the batteries.
But remember that Xantrex also has a generator support mode if AC2 is wired that could do really weird things if not all 4 are playing the same game.

The only configurations I can think of (four at the moment) to wire the four together without major problems would be:
1. To use a single disconnect rather than transfer switch between the grid and the four AC inputs joined together, with the assumption that opening this switch would open the transfer switch inside all 4 inverters. Then all 4 inverters' AC out would have to be paralleled and connected to both of the two house loads.
2. Connect two of the inverters' AC out to each building's individual load panel. Input configuration same as before.
3. Connect a normal transfer switch between the grid and the four inverters, such that the grid and the inverters are never connected together, the house loads go to one or the other, and the AC-in of all four inverters is not used. This would eliminate the possibility of GTI selling when the loads are less than the panel output and the batteries are charged.
4. Similar to 3, but with two inverters per building and two independent transfer switches. If there were ever any cross connects between buildings, this will release magic smoke.

YehoshuaAgapao · March 2013

Re: Xantrex System Failures

As NorthGuy suggests, there may be an Inverter wiring problem giving rise to the Perpetual Motion scenario. As far as I can tell, the installer has not detected and corrected any wiring problems. Sometime, I'll have to tell you about the initial wiring at the circuit breaker box supplying 1/2 of our house that floored all of us (back in October).

If you are doing all xantrex-stock-configuration, Four inverters requires 2 power distribution panels each servicing their own electrical panel because a PDP can only combine up to 3 inverters' output (and with 3 inverters, the bypass needs to be external/custom electrical becuase Xantrex only makes jumper bars for one or two inverter setups). You really need to make sure all output goes only to the laod panels. You probably have input and output cross-wired in the PDP or whatever custom electrical cabinet you might have that is combining the 4 inverters.

I will ask about the LBCO as I expect our installer simply took the defaults on these inverters especially considering how he failed to even look at the firmware version while installing them. However, I think NorthGuy is getting to the heart of one of my concerns which is will the 4 Inverters load the batteries enough so as to fault and give me frequent crashes in normal use.

I have 16 conventional L16 batteries (2 strings). What I like about the PDP is that its DC bus bars can also act as paralleling bus bars for multiple battery bank strings. I wonder if you are taking advantage like I am? Plus if you are going to be sucking 600+Amps out of them batteries, I would not hesitate to get some 500KCMIL battery wires and have your installer put (or send out too have put) his own connectors on them. 500KCMIL is the code minimum (http://www.electriciancalculators.com) for 336 amps continuous

24,000 watts * 1/0.85 inverter efficiency * 1/42 volts minimum battery voltage = 672 Amps maximum continuous current

With two sets of inverter cables (using the PDP DC bus bars as the battery parellel bus) that 672 amps is divided mostly evenly into two to 336 amps. 4/0 will not cut it - they will both cook, and suffer excessive voltage drop when you reach full loads. 2 500KCMil wires will fit in a 2.5" liquid-tight flexible metal conduit (the size of the bottom knockouts on the PDP). Your battery cables (and wiring to load panel) actually need to handle the rated capacity of the inverter continuous plus the surge capacity non-continuous.

As to commissioning the batteries: I don't know what the installer has done exactly. These batteries have been hooked up and attached to at least 1 inverter since October 2012

If your batteries have been sitting without getting charged by that one inverter, you probably got yourself nice case of battery sulfation. If your system sits idle without charging batteries, you get more sulfation. My batteries had a moderate amount of sulfation sitting in storage for 6 weeks and getting non-corrective default-voltage maintenance charges every 2 weeks over 3 months.

I would also recommend getting more solar panels. Batteries are expensive and are the shortest lived asset on a off-grid or mixed-grid (bi-modal; grid-tie with backup) system. Solar panels is one of the longer lived components and have much better cost and actually make power (and if you have grid-tie you can sell your excess). Batteries are expensive and only store power and don't like being under-charged and die very quickly if not maintained or left heavily sulfated / undercharged. You got 24KW worth of inverters, I'd get 24KW worth of solar panels (triple your 8KW), assuming you have room for them.

As to my installation, only one installer in my area I would trust for a full service install and they wanted gobs of money and would relinquish little control. James is a crack union (predominantly commercial) electrician but he's only done 4 systems and they were all straight grid tie. I the design, configuration, plans, SRP, and network-related low-voltage were all DIY by me (I actually had to take a few days off of work to guide him on where to put the wires in the PDP). My system cost me $45.5K (incl batteries, subpanel, generator). Dependable Solar would've wanted 60K (they rough quoted 50K) with all the pre-design for expansion and general over-engineering (tilt mounts, standoffs, x-large PV wire sizing).

YehoshuaAgapao · March 2013

Re: Xantrex System Failures

So you could get up to about 48kW hours out of them. Do you need that much? It's three times what I use here in town. And do you really have combined loads that will demand 24kW all at once? That's two electric stoves on full. That battery bank is only good for about 9.6 kW "hit" all at once.

My whole-house electricity usage (all-electric, no gas) is 30-60 KW-h per day in Oct/Nov/Mar/Apr, 40-70 KW-h per day in Dec/Jan/Feb (heat pump), 50-80 KW-h per day in May/Jun/Sept (pool pump, 24/7 swamp cooler) and 60-150 KW-h per day in July/August (too humid for swamp cooler, stuck with A/C). My 6.4KW system is running 37KW-h per day (6 panels are flat-mounted and act like 4 panels so the 6.48KW system is really 6.0KW).

This person, having 4 inverters and 2 strings of 2-volt batteries (about 100KW-h of storage, 60-80 KW-h usable), obviously wants to be able to run his whole house off grid, and with the poor PV capacity, it looks off-grid and not bi-modal. I only have off-grid capability for my non-kitchen, non-dining-room, non-laundry 120V circuits and the swamp cooler (even in humid july 110+ day, the swamp cooler will still get the house below 90, close to 85). During an outage (short or long term) Kitchen outlets (except refrigerator), dining room outlets, dishwasher, garbage disposal, washer, and any 220V circuit (dryer, electric water heater, range, pool pump, A/C) are non-operable during off-grid use.

NorthGuy · March 2013

Re: Xantrex System Failures

glasstone2 wrote: »

The batteries are the 2V Trojan L16RE in two 48V banks for 2220AH as correctly mentioned by BB.

24kW load, assuming 90% efficiency will be 555A at 48V. This is 25% of the bank capacity. Such load can easily drag down battery voltage below 46V (the Xantrex default for LBCO), especially when batteries are not fully charged.

Also, with such high current, if wiring is not adequate, there could be a significant voltage drop between inverters (or one of the inverters) and batteries, as high as 1-2V. In this case, when batteries are at 47V, the inverter may think they are at 46V, or even at 45V. This can contribute to fault 48 too.

glasstone2 · March 2013

Re: Xantrex System Failures

There has been some speculation on the wiring of my System. It is as follows:
1. There is NO grid connection to the Inverters. We are not selling power back to the utility, and my installer tells me that even though we can set the Inverters not to feed current into the grid with a software switch (turn off the Sell Mode), we cannot make that one-way grid connection without me doing the paperwork to sell back. It would be more convenient to have the grid connected to the Inverters as this would simplify the switch between solar power and grid power, but I cannot have that.
2. There are 2 PDPs with 2 Inverters each. The output of the PDPs are combined in another panel and the 240V, 2-phase wiring goes out to our own utility pole where the grid's meter is mounted and we have 2 distribution panels--one distribution panel per building comprising our house.
3. There is an 8.5kW backup generator. Its main purpose is to make up charge on the batteries. The breakers are now arranged in the PDPs so that we can bypass the Inverters and feed the house directly from the generator in an emergency.
4. The Transfer Switch is far more complicated than I would like. The solar power system 240V, 2-phase wire feeds the two distribution panels (1 per building of our house) at our utility pole via a 125A per phase 240V circuit breaker in each panel. The installer has put in a mechanical interlock device (a sliding, shaped piece of metal) as is found in older backup generator installations. To switch to the solar electric in each panel, we throw the circuit breaker from the meter, slide up the interlock, and switch in the solar electric circuit breaker; to go back to the grid, we reverse this procedure. I had wanted a single transfer switch with minimal interruption, but my installer indicates this is impossible?

In the tests I described, both the perpetual motion and crash, the generator and the solar panels were cut out. The Inverters were running purely off the batteries. When and if we had success on just batteries, we would (I would insist) cut in the other subsystems to see if they work. YehoshuaAgapao's comment on the possible condition of the batteries echoes the concerns I have been voicing to my installer for months; at various times, over 2 week and longer periods, he has either left the system off entirely or left some combination of Inverters maintaining battery charge. BTW, when my installer starts his demonstrations that always fail, I note that the 5 green battery status LEDs on all the XW6048 Inverters are all lit, when they start supplying load, it quickly goes down to just the left 2 LEDs lit. I realize these LEDs are a crude indicator, but is this behavior expected?

YehoshuaAgapao mentioned his usage. Like him, my house is wholly electric. Most of the year with little or no HVAC (the Texas Hill Country is a fairly mild climate) our average usage is 54 kWH/day. In summer, it averages 62 kWH/day, and winter 69 kWH/day. However, those numbers reflect our previous HVAC (conventional split Heat Pump). That HVAC has been replaced with a Geothermal Heat Pump which should reduce the summer usage and greatly reduce the winter usage. Moreover, we have since replaced 1 of our 2 Hot Water Heaters (the most power hungry one) with a heat pump hybrid model; the remaining one is a Marathon. As a result, I expect all of the above usage numbers to be reduced.

Thanks for the suggestions on monitoring the DC side of the system for the next time my installer's people show up; we have no next date. I would think that these are the sorts of tests they should be doing to diagnose the problem, but I suspect they are hoping to get a problem-free demonstration and dump the thing on me.

NorthGuy · March 2013

Re: Xantrex System Failures

glasstone2 wrote: »

1. There is NO grid connection to the Inverters. We are not selling power back to the utility, and my installer tells me that even though we can set the Inverters not to feed current into the grid with a software switch (turn off the Sell Mode), we cannot make that one-way grid connection without me doing the paperwork to sell back.

That may not be entirely accurate. There are three modes - Sell mode (where you pump power back to the grid), Grid Support mode (where you combine inverter output with grid, but do not sell) and Regular mode (where XW doesn't connect its output to the grid, but still can use grid power for charging). In this last mode, inverter acts like a regular load (the same as if you would connect a charger for your EV car). For this mode, you probably wouldn't need to do the paperwork.

glasstone2 wrote: »

3. There is an 8.5kW backup generator. Its main purpose is to make up charge on the batteries. The breakers are now arranged in the PDPs so that we can bypass the Inverters and feed the house directly from the generator in an emergency.

This may be too small. When you're charging batteries with generator, you need to get power into your house from somewhere. Clearly, you cannot get it from batteries, so you need to get it from generator. If you want to preserve your 24kW capabilities and get 10kW to batteries, it would need to be 24kW + 10kW = 34kW.

If you have 2 separate banks, you may charge one bank with generator and use other bank to power loads, but you will only be able to use half of your load during these times. If it is 6 hours for one bank, and then 6 hours for the other bank, it's pretty much the whole day (or whole night). This may be to harsh on the generator.

glasstone2 wrote: »

4. The Transfer Switch is far more complicated than I would like. The solar power system 240V, 2-phase wire feeds the two distribution panels (1 per building of our house) at our utility pole via a 125A per phase 240V circuit breaker in each panel. The installer has put in a mechanical interlock device (a sliding, shaped piece of metal) as is found in older backup generator installations. To switch to the solar electric in each panel, we throw the circuit breaker from the meter, slide up the interlock, and switch in the solar electric circuit breaker; to go back to the grid, we reverse this procedure. I had wanted a single transfer switch with minimal interruption, but my installer indicates this is impossible?

The purpose of the transfer switch is to switch load between two sources and make sure that these two sources are never connected together. One of the common way to do that is to use mechanical interlock with breakers so that it's impossible to switch them both on at the same time. There are lots of other models which look like a box with switching lever attached to it etc.

glasstone2 wrote: »

I note that the 5 green battery status LEDs on all the XW6048 Inverters are all lit, when they start supplying load, it quickly goes down to just the left 2 LEDs lit. I realize these LEDs are a crude indicator, but is this behavior expected?

That's perfectly normal. Xantrex is not capable of detecting SOC. When voltage sugs under the load the number of LEDs go down.

glasstone2 wrote: »

YehoshuaAgapao mentioned his usage. Like him, my house is wholly electric. Most of the year with little or no HVAC (the Texas Hill Country is a fairly mild climate) our average usage is 54 kWH/day. In summer, it averages 62 kWH/day, and winter 69 kWH/day. However, those numbers reflect our previous HVAC (conventional split Heat Pump). That HVAC has been replaced with a Geothermal Heat Pump which should reduce the summer usage and greatly reduce the winter usage. Moreover, we have since replaced 1 of our 2 Hot Water Heaters (the most power hungry one) with a heat pump hybrid model; the remaining one is a Marathon. As a result, I expect all of the above usage numbers to be reduced.

You're probably in a very sunny place, but, even without losses, your solar array is probably not capable of producing 18MWh of electricity per year. You will have to run the generator for long times.

Your batteries, if cycled between 50 and 90%, can store about 40kWh. With 69kWh/day consumption this may not be enough to take you through the night even in sunny weather.

If you want to go off-grid, you would need to lower consumption (which may be easier than you think. We went from 32kWh/day in winter to 14kWh/day without any problems) or increase the size of your system.

Blackcherry04 · March 2013

Re: Xantrex System Failures

NorthGuy wrote: »

That may not be entirely accurate. There are three modes - Sell mode (where you pump power back to the grid), Grid Support mode (where you combine inverter output with grid, but do not sell) and Regular mode (where XW doesn't connect its output to the grid, but still can use grid power for charging). In this last mode, inverter acts like a regular load (the same as if you would connect a charger for your EV car). For this mode, you probably wouldn't need to do the paperwork.

He might not be able to get with this if he has a Smart Meter, it'll register any activity on the buss as a use. That said, you might still leave a separate sub panel on the grid as a backup source. Charging Batteries would be a lot cheaper from the grid. I left a side line tap on my meter base and a box for emergency's, even at $10.00 a month it's a fall back of last resort.

Cariboocoot · March 2013

Re: Xantrex System Failures

Technical point: your power is single phase, not "2 phase". For all practical purposes there's no such thing as two phase. The North American standard for households is split single phase (240 VAC on the phase, divided into to half-phase waves of 120 VAC).

Otherwise we're all just speculating here until more information is available. The comments on battery bank size, wiring, and array size are all correct. The AC wiring part could still be done incorrectly as could the programming.

Unfortunately it sounds like your installer is not actually familiar with the work he's doing. If he were, it would all be sorted by now by him.

NorthGuy · March 2013

Re: Xantrex System Failures

Blackcherry04 wrote: »

He might not be able to get with this if he has a Smart Meter, it'll register any activity on the buss as a use. That said, you might still leave a separate sub panel on the grid as a backup source. Charging Batteries would be a lot cheaper from the grid.

That's what I have now. The grid enters into a small panel, which has a main breaker and one load breaker. Xantrex's AC1 input is connected to this load breaker. When I turn it on, Xantrex passes the grid power to the output (where all disctribution panels are connected) and goes to charge mode and charges batteries. When it's done, it stops charging and does nothing. Loads are still powered fro the grid. Since the "Grid Support" is turned off, it cannot go to invert mode while grid is on. Then I switch the breaker off and it immediately goes to inverting.

The Meter only sees it as a load when charging. When it is inverting, everyhting is completely disconnected from the Meter.

YehoshuaAgapao · March 2013

Re: Xantrex System Failures

There is an 8.5kW backup generator. Its main purpose is to make up charge on the batteries. The breakers are now arranged in the PDPs so that we can bypass the Inverters and feed the house directly from the generator in an emergency.

Too small. Should be close to inverter's continuous rating so you can get through the night. Ideally, because you're off-grid, the full rating of the automatic transfer switch of each inverter (240A worth of breakers but continuous = 192A = 46KW) to accommodate combined battery charging demand and load panel demand.

The Transfer Switch is far more complicated than I would like. The solar power system 240V, 2-phase wire feeds the two distribution panels (1 per building of our house) at our utility pole via a 125A per phase 240V circuit breaker in each panel. The installer has put in a mechanical interlock device (a sliding, shaped piece of metal) as is found in older backup generator installations. To switch to the solar electric in each panel, we throw the circuit breaker from the meter, slide up the interlock, and switch in the solar electric circuit breaker; to go back to the grid, we reverse this procedure. I had wanted a single transfer switch with minimal interruption, but my installer indicates this is impossible?

Xantrex XW (and Outback Radian) inverters have automatic transfer switches built in to them. Its why it has 3 AC connections - AC1-in (grid), AC2-in (gen), AC-out (loads). This complexity is not needed. Generator output does need to be y-split to each PDP.

In the tests I described, both the perpetual motion and crash, the generator and the solar panels were cut out. The Inverters were running purely off the batteries.

Check your AC wiring. May need to break out the wiring diagrams in the manual, take off the deadfronts, and check yourself. Your install seems it might be manual-stock outside the generator/ATS.

Do get your firmware versions syched up. XW-config can update firmware. Check your charge controller firmware versions also. Firmware 1.15+ (bugs fixed in 1.17) supports AC-coupled battery backup systems (accept AC input from AC-out from a grid-tied inverter), which can contribute to your perpetual motion/crash. Make sure you have your phase-synch cables installed. And make sure they are all networked together on the ethernet cables. It is possible that one pair of inverters is drawing from the other pair as if it were a grid-tied inverter in a AC-coupled battery system. The latest and greatest manual (November 2010) does say to connect the inverter load subpanels together when you have 2 PDPs. Maybe you have your master/slave configuration setup wrong? Or you don't have the phase-synch cables installed? Or the ethernet is not setup or setup correctly?

BTW, when my installer starts his demonstrations that always fail, I note that the 5 green battery status LEDs on all the XW6048 Inverters are all lit, when they start supplying load, it quickly goes down to just the left 2 LEDs lit. I realize these LEDs are a crude indicator, but is this behavior expected?

Both battery sulfation and undersized battery cabling can contribute to this. Sulfation at any non-trivial load. Undersized Wiring at heavy loads. If you have 2 inverter cable sets, anything less than 400KCMil is undersized.

At defaults, worse case scenario continuous battery current from 4 Xantrex inverters is 580A, split in two is 290A (continuous). Minimum wire size is 400KCMil for that current. This is all wire-temperature safety. If you got more than 5-10 feet of battery cable, you need to go beyond code for voltage drop sake. Now if you wanted to be fully safe (and its great for voltage drop), you would need wire rated for 290A continuous plus 290A non-continuous (for the Xantrex inverter's massive surge capacity), you will need 1250KCmil wire.

At 290A continuous on 10 feet of 400KCMil, your voltage drop is 0.19V (using the VB calculator at http://electriciancalculators.com - the 2 main ones are table 310.16 & voltage drop and Raceway/trough fill). half that for 5 feet. Battery cable needs to have as little voltage drop as cable. You can fit 500KCMil in 2.5" conduit. Just use 500KCMil (voltage drop 48V 10' is 0.15V). 1250KCmil 10' will have 0.06V voltage drop. 2x 1250KCMil needs 4" conduit. If you have a start capacitor in all your big motors (A/C, well pumps, etc..), you probably won't need the surge capacity of the inverters and can get by with 500KCMil for your battery cables. 4/0 has 0.35V voltage drop (48V, 10 feet). If you for some reason had 2AWG going to the batteries, the voltage drop rises to 1.13V. I doubt you got less than 2/0 or 4/0 - 2AWG would burn up its insulation and short circuit at that kind of current. If you had all 580 amps going on a single inverter cable set at 4/0, your voltage drop rises to 0.71V and the wire would get quite hot.

Are your 2 battery strings combined somehow or is each isolated to its own PDP & inverter-pair?

Vic · March 2013

Re: Xantrex System Failures

Regarding battery health, have not noticed anyone mentioning that a Hydrometer -- a good glass one -- is essential for determining SOC of EACH cell of each battery. These SG readings should be recorded in your battery Log Book.

Your batteries should have used some water by now. It is also very important that the electrolyte level and SG of each and every cell be monitored at least monthly, until YOU get a feel of the battery bank needs. You can identify a/several Pilot Cell/s which will help you quickly do a spot check on SG more frequently.

You seem to have many issues with the system, BUT, please take it on yourself the monitor and get a feel of what your batteries need in the way of maintenance.

Agree with others, that genset seems way too small. Hope you have room for an additional/other genset. In your probable climate, an adequately-sized genset is a real life saver at times. Good Luck, Vic

YehoshuaAgapao · March 2013

Re: Xantrex System Failures

Regarding battery health, have not noticed anyone mentioning that a Hydrometer -- a good glass one -- is essential for determining SOC of EACH cell of each battery. These SG readings q should be recorded in your battery Log Book.

SG (Specific gravity) from a hydrometer is the only accurate measure of battery state of charge. Unless you are super rich and hire your installer to do this for you and he is full-service and customer-oriented (this seems questionable), you should look after your batteries and system in general.

Definitely don't neglect watering. Exposed plates are bad news (exposed plates immediate and probably irreversibly sulfate). Heavily exposed plates = battery death.

Agree with others, that genset seems way too small. Hope you have room for an additional/other genset. In your probably climate, an adequately-sized genset is a real life saver at times. Good Luck, Vic

Aggreed. Get more PV too. Generators are gas guzzlers and gas isn't cheap. PV is getting cheaper as time goes on (Fed money printing may reverse this though), batteries are getting more expensive. PV lasts much longer than batteries (and inverters) do. Load up on the PV. You should be able charge your batteries fully in a day on top of the feeding the house demand in a day. 5% (C/20) rate blah. I prefer 20-25% (C/5 to C/4), about double the recommended charge rate - enough for both your loads and your batteries simultaneously.

Northern arizona Wind & Solar Battery FAQ - Flooded (recommended maximum): C-20/8 (12.5% of AH capacity at 20-hour rate)
Northern arizona Wind & Solar Battery FAQ - AGM (recommended maximum): C-20/4 (25% of AH capacity at 20-hour rate)
Surrete (recommended maximum): C-6/6.67 (15% of AH capacity at 6-hour rate, which translates to about C-20/9 or 11% of AH capacity at 20-hour rate)
Trojan flooded (recommended): C-20/10 to C-20/7.7 (10-13% of AH capacity at 20-hour rate)
Trojan AGM (recommended): C-20/5 (20% of AH capacity at 20-hour rate)

My system (6.48KW, 800AH) has a max charge rate of 141A (STC, LBCO), which is C-20/5.7
Your system (8KW, 2220AH) has a max charge rate of 174A (STC, LBCO), which is C-20/12.8
My system (6.48KW, 800AH) has a typical charge rate of 103.7A (80% STC, ReCharge Volts), which is C-20/7.7
Your system (8KW, 2220AH) has a typical charge rate of 128A (80% STC, ReCharge Volts), which is C-20/17.3

You should get to 12KW minimum, 24KW if you can (especially if you go bi-modal and sell to the grid; Xantrex favors drawing from grid but can tinker it not to with load shave. Outback radian favors mini-grid which minimizes grid use and only uses grid in place generator which guzzles expensive gas)

Ken Marsh · March 2013

Re: Xantrex System Failures

Cariboocoot wrote: »

Technical point: your power is single phase, not "2 phase". For all practical purposes there's no such thing as two phase. The North American standard for households is split single phase (240 VAC on the phase, divided into to half-phase waves of 120 VAC).

A technical point on C Coot's technical point.
You are right. There is no practical two phase around today
but there used to be 50 - 75 years ago around Niagara falls and in England.
The two phases were spaced 90 (270) deg and it really worked good.
Google "Scott T".

And as to the original post.
You most likely did go out on low bat. voltage.
Particularly with Trojan batteries.
When they are new and if they have not been used for awhile and are cold, etc.
the kind of amps you are drawing will pull them right down.

There is nothing wrong with Trojan batteries.
They are delivered under developed.
You have to cycle them several dozen times before they get up to capacity.

I helped install a system with three XW-6048s.
It has a HUP battery in effect somewhat larger than yours.
We never could get it to put out full power for any amount of time.
But by realizing that you just can not run everything at once it is a useful system.

ChrisOlson · March 2013

Re: Xantrex System Failures

I don't understand why one inverter is being used to charge batteries with the AC output of the other ones. That seems pretty stupid to me?

I think the problem with the system going offline is probably something not wired right on the AC side. And not enough batteries to power four inverters even if they were wired right. We just got our new XW installed here and got everything kind of tuned this past week, and added some more panels. We did a "stress test" by loading the new inverter to a measured 29.7 amps on one leg and 31.1 amps on the other for 30 minutes with no gen support, no incoming power, and a fully charged bank. It held fine for 30 minutes at 7,300 watts output but the bank was down to 47.4 volts under load after 30 minutes, and my clamp-on meter showed 167 amps going to the inverter.

I would say we got the bare minimum for battery power to run a 6048 at full rated power for 30 minutes. So I don't see how 2,200 ah is even going to come close to running four of 'em and keep 'em lit - unless you got some serious incoming RE power to offset what has to come from the bank.
--
Chris

Blackcherry04 · March 2013

Re: Xantrex System Failures

ChrisOlson wrote: »

I don't understand why one inverter is being used to charge batteries with the AC output of the other ones. That seems pretty stupid to me?

I think the problem with the system going offline is probably something not wired right on the AC side. And not enough batteries to power four inverters even if they were wired right. We just got our new XW installed here and got everything kind of tuned this past week, and added some more panels. We did a "stress test" by loading the new inverter to a measured 29.7 amps on one leg and 31.1 amps on the other for 30 minutes with no gen support, no incoming power, and a fully charged bank. It held fine for 30 minutes at 7,300 watts output but the bank was down to 47.4 volts under load after 30 minutes, and my clamp-on meter showed 167 amps going to the inverter.

I would say we got the bare minimum for battery power to run a 6048 at full rated power for 30 minutes. So I don't see how 2,200 ah is even going to come close to running four of 'em and keep 'em lit - unless you got some serious incoming RE power to offset what has to come from the bank.
--
Chris

Sounds like you'v been busy swapping out your SW Inverter. I swapped out a Magnum yesterday to Outback G series to try out the generator support feature. If it works , it'll save me a bunch on fuel. It took for ever to re-work the sub-panel for a third Inverter with constant loads all on one Inverter, instead of spread around. Good Luck with yours.

Cariboocoot · March 2013

Re: Xantrex System Failures

ChrisOlson wrote: »

I don't understand why one inverter is being used to charge batteries with the AC output of the other ones. That seems pretty stupid to me?

It isn't supposed to. The most likely culprit here is that something has been wired wrong or at the very least programmed wrong.

Blackcherry04 · March 2013

Re: Xantrex System Failures

Cariboocoot wrote: »

It isn't supposed to. The most likely culprit here is that something has been wired wrong or at the very least programmed wrong.

Another thought is that it's probably a calamity of errors, maybe one part of the original install team set one up to keep the batteries charged and would change it back prior to commissioning. It's a complicated system for sure and a on going problem.

ChrisOlson · March 2013

Re: Xantrex System Failures

Blackcherry04 wrote: »

Sounds like you'v been busy swapping out your SW Inverter

We had a guy that wanted to buy the SW Plus and I wanted a bit bigger inverter so I can run my welder without having to pre-start the generator. I also got another Honda generator that I can use on the welder, but that ended up being a pain. So we bought the big XW. Still fiddling with some of the settings but got it fairly well tuned now to our loads. As good as the 5548 was for surge and overload, it don't hold a candle to the XW6048

Mike at Schneider tech support recommended that for our small generator that we use our Outback PSX-240 to prevent gen leg overload. This has worked fine - no problems with the inverter disqualifying the generator with leg imbalances. The 6048 can handle a substantial leg imbalance with no issues - but the generator can't.

So it was one of those things we decided to do, that didn't cost us very much money.
--
Chris

ChrisOlson · March 2013

Re: Xantrex System Failures

Blackcherry04 wrote: »

It's a complicated system for sure and a on going problem.

It actually sounds a bit scary to me. I get the distinct impression that the installer is attempting to get the inverters to run once in their present configuration and if that's successful go "She's good - have fun - see ya".

Plus I'm convinced there's not enough batteries for four inverters and the voltage sag when they start loading is severe so one kicks out with a F48 and it cascades the other ones. So I dunno, but when there's something wired in a loop so one inverter is charging batteries from the output of the other ones, there has been a serious misunderstanding in wiring them up.
--
Chris

Xantrex System Failures

Comments

Categories