Running two XW6048 inverters

Brent
Brent Solar Expert Posts: 64 ✭✭✭✭
I am curious if anyone is running two Xantrex XW6048 inverters. I have two for peak load reasons but my solar panel output runs around 6500 watts. One inverter works all the time and indicates around 5800 watts AC sell to grid at peak times. The slave always remains idle.

There have been edge of cloud effects where the master inverter indicated 6600 watts but the second inverter [slave] never seemed to kick in.

I am thinking of adding another 3 kw of DC to an additional charge controller [have two now] and was wondering at what level the second inverter starts supplying power.

Thanks,

Brent

Comments

  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Running two XW6048 inverters

    Not certain, but likely after the Master inverter has had more than 6kW output for longer than its surge time rating. These things will pas like 12kW for up to 10 or 20 seconds IIRC.
  • SolaRevolution
    SolaRevolution Solar Expert Posts: 410 ✭✭
    Re: Running two XW6048 inverters

    That is a question for Tec support. The XW system has had many firmware revisions which would deal with this in different ways. When you call be sure to have the revision #s of both inverters and the SCP at hand.

    It may be as simple as settings or it could require new firmware.
  • Brent
    Brent Solar Expert Posts: 64 ✭✭✭✭
    Re: Running two XW6048 inverters

    Thanks for the response. I have found Xantrex tech support to be generally incompetent. I have received conflicting answers to the same question many times. It seems that their training is severely lacking and their understanding of basic engineering principles is non-existent. There may be exceptions but I have not found them yet.

    Brent
  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Running two XW6048 inverters

    In master/slave configuration with parallel inverters the XW's operate independently. Slave unit will start to support loads when the load on the master exceeds 60% of it's rating for 300 seconds. In order for that to work, Search has to be disabled on the master, and enabled on the slave. If it is not working try disabling Search on the slave. If it works then, you either have the Search feature configured wrong, or the load is not "search friendly" and is failing to bring the inverter out of search mode into invert mode.

    The inverter does not care whether the load is loads in your house or the grid. But it does care about bringing a slave unit out of Search for a temporary "surge" and then having to put it right back to Search again. So there is built-in delays to prevent that from unnecessarily happening.

    Schneider tech support, in my experience, has been very competent and helpful with configuration issues or questions I have had on "how do I set this up?" There may have been some issues during the initial transition from Xantrex to Schneider, but once they got "up to speed" after taking over the RE end of Xantrex their tech support has been courteous, prompt and more than competent here.
    --
    Chris
  • Brent
    Brent Solar Expert Posts: 64 ✭✭✭✭
    Re: Running two XW6048 inverters

    Thanks Chris,

    I will try setting the slave unit to search.

    Would I expect to see the power output on the first inverter drop somewhat and the slave unit show a power output value? Currently there are 3 dashed lines on the slave power indicator.

    Also, if the slave unit is now sharing power generation with the master would I expect to see an additional ~8% drop in power generation on the slave inverter due to efficiency issues. If so it might make more sense to just keep only the master producing which is never more than 5800 watts. I have two inverters because I need the additional passthru amps.

    Thanks,

    Brent
  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Running two XW6048 inverters

    The power output of the master should drop if the load exceeds ~3,450 watts and the slave picks up the difference. There is a 300 second countdown in there before the slave will start supporting the load on grid sell. And once it starts supporting the load, the load on the master has to drop to 1,150 watts or less before the slave will go back into Search.

    In your Operation Guide it kind of explains how it works and Search much be disabled on the master and enabled on the slave.

    The XW is most efficient once the inverting load gets above about 1,100 watts and the efficiency curve is pretty flat up to about 4,000 watts. So the main thing to avoid is running the slave unit too much below 1,100 watts. It's not a huge thing because the slave only draws 28 watts idle, inverting. But for peak efficiency it should be operating at 1,100 watts or more. And this is what should be happening with your setup - I would expect to see ~3,500 watt load on the master and around ~2,300 on the slave under normal operation once the slave starts "waking up" like it should when the load on the master exceeds ~3,450.

    The three dashed lines on the inverter's display mean it is in pass-thru and is doing nothing.

    Edit:
    Brent - to clarify on the efficiency issue, your peak efficiency should be achieved when both inverters are operating from 1,100-4,000 watts. If the master is running at 5,800 watts, then you are on the downside of its efficiency curve. Efficiency should actually improve if you reduce the load on the master to below 4,000 and have the slave pick up the difference above 1,100 watts.
    --
    Chris
  • Brent
    Brent Solar Expert Posts: 64 ✭✭✭✭
    Re: Running two XW6048 inverters

    Thanks for the great answer Chris!

    It looks like I can pick up a few watts by running the slave. With 6250W of DC into a single inverter I get 5800 watts AC out @ 93% effic [according to handbook and is correct when I measure it with a clip on ammeter].

    Assume the output is evenly split between two inverters then I have 3125W of DC into each inverter so at 95% effic I should get around 2970W from each inverter or 5940W combined.

    Let me know if this makes sense.

    Thanks,

    Brent
  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Running two XW6048 inverters

    Yep, that's basically how it works, Brent. The efficiency curve is in the electrical specs, I believe, in the manual and your numbers look about right. And even though we don't have grid power here, your numbers coincide with what I have seen for efficiency at various loadings on the inverter. Keeping them operating in what I call the "sweet spot" definitely gets better use of your PV power.
    --
    Chris
  • Brent
    Brent Solar Expert Posts: 64 ✭✭✭✭
    Re: Running two XW6048 inverters

    Chris,

    Can you or anyone suggest a reliable way to measure the efficiency of the two inverters?

    If I use the system DC out value from the chg controllers [120A @ 51.8V] minus battery float amps [1A] and the AC readings on the inverter panels [which I suspect are not too accurate] I get around 89% with the master indicating 1900W and the slave 3600W for a total AC of 5500W: 5500 ACW/6164 DCW = 89% for a combined efficiency for both inverters.

    Thanks,

    Brent
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: Running two XW6048 inverters
    Brent wrote: »
    Can you or anyone suggest a reliable way to measure the efficiency of the two inverters?

    The numbers displayed by XWs are not very accurate. To measure efficiency, you need relatively precise measurement tools and you need to measure both DC and AC at the same time. Clamp meters are not very precise - usually no better than 2%. Voltmeters on DC side are more precise, but on AC side an inexpensive voltmeter is usually no better than 2% neither. You also need to figure out a power factor on AC side. When you put all these measurements together and do all the math you will combine all the errors, so you probably will have precision no better than 5%. Which means that if you measured it to be 89%, there's a good chance that a true value is between 84% and 94%.

    In short, it is not an easy task to measure efficiency with a sufficient precision.
  • Brent
    Brent Solar Expert Posts: 64 ✭✭✭✭
    Re: Running two XW6048 inverters

    Thanks NorthGuy,

    I suspected that the RMS errors would add up and cause an inaccurate result. I think I will live with the Calif approved inverter list which has the XW6048 at a weighted 92.5% efficiency. Any good design should be at least that. I am thinking of writing a software GUI to display the systems real time input/output data in which case I will install some accurate shunts for current measurements etc. I have enough equipment to allow calibration of errors to about 1%.

    Thanks,

    Brent
  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Running two XW6048 inverters

    I checked mine with my Sun AVR on the DC cable and my Fluke meter on the AC. I found that the SCP readings are fairly accurate at higher loadings anyway, so you can probably use those too. Then I calculated the watts in vs watts out. It agreed with the SCP readings within a couple percent.
    --
    Chris
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: Running two XW6048 inverters

    When I use numbers from SCP, I get peak inverting efficiency of 85% (less at smaller loads). I don't think it's true. Comparing to Fluke, AC Volts/Amps seem to be relatively Ok. DC Amps that XW shows during inverting are usually considerably higher that Fluke measurements on the shunt. This might be somehow related to pecuarities of the batteries.

    I'm also working on getting precise measurements so I can get efficiency withing 1% or better, but since I have lots of other things to do, it's moving on very slowly.
  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Running two XW6048 inverters

    What I have measured before is 80-90% from very light loads up to 500-600 watts. 96-97% at 1,100 watts. 92% at 6,000. 84% at 7,200.
    --
    Chris
  • Brent
    Brent Solar Expert Posts: 64 ✭✭✭✭
    Re: Running two XW6048 inverters

    Both interesting observations. I know that there are two kinds of efficiency ratings: marketing efficiency and engineering efficiency. I would be happy with something around 93-94% over the inverter power mid range.

    Can anyone tell me why I get two slightly different reading on the SCP readout for AC power and VA for a given inverter? I sometimes see a 200-300 watt difference where they should read the same.

    I did an inspection on a 500,000 watt PV system earlier this week. The inverter has a 97% efficiency at 500kw so it burns up 15,000 watts!!

    Brent
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    Re: Running two XW6048 inverters
    Brent wrote: »
    Can anyone tell me why I get two slightly different reading on the SCP readout for AC power and VA for a given inverter? I sometimes see a 200-300 watt difference where they should read the same.

    disclaimer: I don't know my way around xantrex and the SCP.

    From a general perspective, why should Power and VA be the same? Power in an AC circuit with inductive or reactive loads is not the same as VA. The difference is Power factor. Watts (power) = Volts X Amps X power factor.

    Try running only a pure resistive load (in which pf=1) and see if power = VA. Fans, LED or CFL lights, motors, many electronic devices, etc are inductive or reactive loads.

    vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Running two XW6048 inverters
    Brent wrote: »
    Can anyone tell me why I get two slightly different reading on the SCP readout for AC power and VA for a given inverter? I sometimes see a 200-300 watt difference where they should read the same.

    Volt-amps is power factor corrected AC power. So they will always be different, according to the power factor. The XW measures DC current in and AC current out on each leg. Everything else is calculated in software. It uses a hall effect sensor to measure DC in (see attachment). I thought I had a photo of the CT sensors in it for the AC side but I can't find it right now.

    Anyway, apparent power as shown on the SCP is pretty accurate, even at lighter loads. The AC power is sometimes a little off at lighter loads and it gets progressively more accurate as load is applied to the inverter.
    --
    Chris
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: Running two XW6048 inverters
    ChrisOlson wrote: »
    I thought I had a photo of the CT sensors in it for the AC side but I can't find it right now.

    Here it is.

    There are two more.
  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Running two XW6048 inverters
    NorthGuy wrote: »
    Here it is.

    There are two more.

    Thanks NG! I knew I had uploaded those to another thread once, but couldn't find it in the Advanced uploads.

    The hall effect sensor is on the processor board. The CT sensors are on the AC output board.
    --
    Chris
  • Brent
    Brent Solar Expert Posts: 64 ✭✭✭✭
    Re: Running two XW6048 inverters

    Thanks for the clarification. I did not realize that the AC power measurements in the inverters were sophisticated enough to take into account polar magnitude in the calculations.

    Brent
  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Running two XW6048 inverters

    Yeah, all they do is measure RMS voltage and RMS current, so the volt-amp reading on the SCP is pretty accurate +/- the error value in most RMS meters. The real power value can be off because it's a calculation without a true watt meter and there's a lot of factors that can affect that, like they probably don't know what the phase angle is between between voltage and current, whether it's leading or lagging. But as you add load to the inverter the watts reading will get more accurate with more load.

    I think they just use a "programmed in" set of values for Power Factor but those values might not always be accurate, depending on the type of load. But for the most part, it's "close enough for government work", as we always say in these parts.
    --
    Chris
  • Brent
    Brent Solar Expert Posts: 64 ✭✭✭✭
    Re: Running two XW6048 inverters

    I thought I would plot the DC/AC levels of the two inverters now that I have the slave running in the search mode.

    Seems to work ok and is much easier on the inverter temps. I was under the impression that the slave would kick in when the master reached 60% but according to the plots it does not.

    Brent
  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Running two XW6048 inverters
    Brent wrote: »
    was under the impression that the slave would kick in when the master reached 60% but according to the plots it does not.

    I'd have to re-read all the multi-unit configurations in the manual. But going from memory the slave should stay in idle (not inverting) mode until the master reaches approximately 60% of its rated output. Then the slave will "wake up" and they should share the load until the load on the master drops below 20% of it's rated output.
    --
    Chris
  • YehoshuaAgapao
    YehoshuaAgapao Solar Expert Posts: 280 ✭✭
    Re: Running two XW6048 inverters
    ChrisOlson wrote: »
    Volt-amps is power factor corrected AC power. So they will always be different, according to the power factor. The XW measures DC current in and AC current out on each leg. Everything else is calculated in software. It uses a hall effect sensor to measure DC in (see attachment). I thought I had a photo of the CT sensors in it for the AC side but I can't find it right now.

    Anyway, apparent power as shown on the SCP is pretty accurate, even at lighter loads. The AC power is sometimes a little off at lighter loads and it gets progressively more accurate as load is applied to the inverter.
    --
    Chris

    The amps on my AC2 is often wrong. Often get Volts*Amps less than watts. My house is full of gadget power bricks and I heard that gadget bricks often just clip at the tops of the sine waves. Not sure if its that or if the inverter just can't measure amps well. Also the current reading on the DC is always high (but watts aligns) when the inverter is selling (AC1+AC2 watts about 80% of DC watts), but more reasonable if it is in load shave or invert mode (battery watts within 90% of AC2 watts).