Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...

danielh
danielh Solar Expert Posts: 69 ✭✭✭✭
I am close to finalizing a 6.5kw PV system. My default plan is to grid-tie it. But the notion of providing UPS protection and backup for outages is appealing, especially after the 3 day outage I just experienced here in Maryland. But it isn't hugely appealing: if the cost is high, I can continue using my 6.5kw Generac generator/transfer switch, along with consumer grade UPS's on individual circuits.

After some research, and with a lot of help in a prior thread (http://forum.solar-electric.com/showthread.php?16354-Grid-interactive-what-benefits-make-it-worthwhile), I have identified a few options that might be worthwhile. I set a cap of $3000 extra (after fed rebate); measured as extra costs over and above the default grid-tie system. I figure 3.5kw AC (5kw surge) minimum for the "critical loads" subpanel to be supported by the inverter, 4kw or so of PV input, and a 9.6kw battery pack. Note that I may be undersizing the battery pack, but the goal is not off-grid living. Rather it is to make outages easier to deal with (somewhat less then 24 hr power availability is acceptable), and provide a good most-of-house UPS.

So the question is how to do this. Which leads me to the question:
What is better: outback vs xantrex vs .. magnum vs .. ?
[I know these kinds of opinions have been given on threads spread around this site, but perhaps it is a good time to reassess current thinking].

While I am probably missing some good alternatives, and may be dead wrong on my assumptions (especially regarding costs that are "extra"), three possibilities seem doable. Doability is effected by my cost cap, which precludes the more elegant choices, such as SunnyIsland/SunnyBoy or Outback Radian.

a) Outback Flex Power One (GVFX3648, Mate, FM 80 charge controller)
(around $4000 + $2000 for batteries - $2000 from dropping grid-tie inverter)
b) Xantrex XW6048 and 80A charge controller
(around $4500 + $2000 for batteries - $2400 from dropping grid-tie inverter)
c) Magnum MS4448PQE (ac coupling)
(around $2000 + $2000 for batteries)

The Outback might be a bit less expensive and seems to have good tech support, but has less capacity -- it would probably not support the full set of "critical loads" currently wired into my transfer-switch. And getting my current generator to work with it may not be possible (though "exchanging" the Generac for a nice Honda has its own appeal).

The Xantrex is a better size. Somewhat more expensive, might have some small generator support issues.

The Magnum is interesting, especially since it allows me to put the bulk of the decision off (since it could be retrofitted in at a later dater). But whether or not it is ready for a novice like me (i.e.; how to deal with excess current production when the grid is down) is not at all clear.

I am leaning toward the Xantrex, but it would be great to hear other's experiences.
And are my "extra" cost estimates are way too optimistic? (note these estimates assume panels not on a central inverter will be on microinverters).

Comments

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...

    Do you need 120/240 VAC support?

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • danielh
    danielh Solar Expert Posts: 69 ✭✭✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...
    BB. wrote: »
    Do you need 120/240 VAC support?

    -Bill

    Interesting question. It would make life a bit easier to have 120/240 output, since I could then use my current transfer switch as the critical load panel (it has 240V input).
    Otherwise, I do NOT intend to support any 240VAC loads (I accept that the central Ac, stove, and dryer will not be available)

    So 120V support only, like the Outback, would work -- but 240V (like the Xantrex) would probably be more convenient.

    OTOH, and I have no idea if this practical or feasible, I wonder if I could feed one leg (at 120VAC) of the transfer switch into the Outback. That would make it easier
    to switch between grid power and generator power, and I won't sacrifice the current loads capable of being run during an outage (eg; with the genset on, I can use
    the microwave and the toaster at the same time). This may be pushing the design limits of the transfer switch, but if safely doabie t would make the outback more attractive.
    Actually, there is still the problem of whether the GVFX3648 would work with the Generac (in its flexpower one environment), so maybe this idea is irrelevant.

    Maybe I am worrying too much about "repurposing" my current transfer switch, since I put it in only 6 months ago and it would be embarrasing to yank it out now that my standards have changed.
  • mike95490
    mike95490 Solar Expert Posts: 9,583 ✭✭✭✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...

    I think if you can get someone here on the forum, to translate into english, the proper way to configure the XW for load shaving, and not to be cycling the batteries all the time, it would be the "stronger" of the choices, except for the beastly: RADIAN GS8048 inverter, which is internally, 4 smaller inverters daisy chained together.
    http://www.outbackpower.com/products/sinewave_inverter/radian/

    Also, check on the POWER FACTOR (PF) of the battery chargers, for boosting the batteries if the power restores at night or if you hook a generator to it.
    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 ,

  • SolaRevolution
    SolaRevolution Solar Expert Posts: 410 ✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...

    Your budget for installing any of the systems mentioned is very optimistic and none of the systems you mentioned would work with your desire for a 6.5 kW array.

    An additional budget of $3k is tight for the equipment alone. The labor for a back up system is a much bigger project than for a non-battery system. Even a skilled DIY installation will have additional costs which can require you to spend twice that much or more.

    The XW6048 is the only inverter you mentiond with the capacity to handle the 6.5 kw PV array. It could work as the battery side of an AC coupled system (poor charge regulation when the grid is out). If you stay with DC coupling, an 80 amp charge controller can only handle around 4000 watts. So you need more charge controller capacity or a smaller array. Going with a smaller array may be considered as a trade off for the higher cost for batteries but be sure to check how it will effect your desired return on investment, rebates or other incentives.

    If you want to "put off the decision" and add a backup power later, your battery inverter will need to have adequate backfeed capacity for the PV rating. It should also have the AC voltage to match the non-battery PV inverter. If your PV array feeds one 6kw non-battery inverter it will not be able to be used to AC couple with a Magnum MS4048PQE or with an Outback GVFX3648. You could possibly use multiple inverters (or microinverters) and seperate the AC outputs so you could have part of the PV array feeding the back up system and the rest connected directly to the grid. This could work with the 120/240vac Magnum but the 120vac Outback would still need an autotransformer to accept the PV inverters AC output.

    Since you already have a generator and transfer switch sized for 6.5 kw at 120/240vac, it could help make the installation easier especially if you go with the XW6048. Just be aware that although it can be done there can be some "challenges" to making the Generac and the XW AGS get along. It may be easier to remove the transfer switch and let the XW system operate in it's place. (There will still be some AGS issues requiring some creative solutions)

    I don't want to discourage you from battery back up. The ROI and $ numbers have a tendancy to do that on their own.:cry:


    -Alex Aragon
  • danielh
    danielh Solar Expert Posts: 69 ✭✭✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...

    Hello solar revolution.
    Your budget for installing any of the systems mentioned is very optimistic and none of the systems you mentioned would work with your desire for a 6.5 kW array.
    An additional budget of $3k is tight for the equipment alone. The labor for a back up system is a much bigger project than for a non-battery system. Even a skilled DIY installation will have additional costs which can require you to spend twice that much or more.
    I was wondering how much these costs would be. Mostly, I am trying to get some approximate sense, to see if it is even worth talking to the contractors (this is not a DIY project for me).
    The XW6048 is the only inverter you mentiond with the capacity to handle the 6.5 kw PV array. It could work as the battery side of an AC coupled system (poor charge regulation when the grid is out). If you stay with DC coupling, an 80 amp charge controller can only handle around 4000 watts. So you need more charge controller capacity or a smaller array. Going with a smaller array may be considered as a trade off for the higher cost for batteries but be sure to check how it will effect your desired return on investment, rebates or other incentives.
    I should of clarified that I was anticipating segregating the panels in to two groups. One group, say using 17 of the 260w (helios) panels (that are on order) would feed a "hybrid" central inverter. The other group (say, 8 panels) would either feed a standard grid-tie inverter, or would use microinverters. I am thinking that even if I don't do the batteries now, that I would still go with this segregation, so as to make it easier to revisit the issue in a year or so (i.e.; if ac coupling solutions become simpler). Note that the panels will be physically spread over several locations on my somewhat discontinuous hip-roof-with-flat-roof-addition with has intermittent shading issues; and my contractor is fine with this kind of segregation.

    If you want to "put off the decision" and add a backup power later, your battery inverter will need to have adequate backfeed capacity for the PV rating. It should also have the AC voltage to match the non-battery PV inverter. If your PV array feeds one 6kw non-battery inverter it will not be able to be used to AC couple with a Magnum MS4048PQE or with an Outback GVFX3648. You could possibly use multiple inverters (or microinverters) and seperate the AC outputs so you could have part of the PV array feeding the back up system and the rest connected directly to the grid. This could work with the 120/240vac Magnum but the 120vac Outback would still need an autotransformer to accept the PV inverters AC output.
    I should of read your 2nd paragraph before responding to your first!

    That was my thought with the Magnum, which is why I would size the grid-tie inverter to be around 4k (my contractor is recommending the Aurora One 4.2).
    The magnum documentation states "The maximum power rating of the renewable energy source must be no greater than 90% of the continuous power rating of the Magnum inverter",
    so that suggests max PV output of 3.96kw if I use the 4.4kw MS4448PAE.

    Let me ask: should that "max" be computed using the STC rating or PTC rating (for the helios 260, that is 234). At the PTC, 17 panels would work; at STC, 15 panels.

    BTW: I wasn't considering using the Outback in AC coupling world. Maybe it would be useful (the gt inverter would replace the Outback charge controller), but I was thinking that the Outback would be a "install batteries now" solution, not a "install them later".
    Since you already have a generator and transfer switch sized for 6.5 kw at 120/240vac, it could help make the installation easier especially if you go with the XW6048. Just be aware that although it can be done there can be some "challenges" to making the Generac and the XW AGS get along. It may be easier to remove the transfer switch and let the XW system operate in it's place. (There will still be some AGS issues requiring some creative solutions)
    Eh, AGS is not an issue (the generac is manual start anyways). I need to move the generator from storage to the exterior plug when I use it anyways, so auto start wouldn't be that usefu
    I don't want to discourage you from battery back up. The ROI and $ numbers have a tendancy to do that on their own.:cry:
    Yes indeed. That is why I am trying to figure if there is a way of doing this at moderate cost; sacrificing some convenience and accepting less than 24/7 coverage, and without jeopardizing the main purpose of the installation (to generate solar power the 99% of the time the grid is working fine, with much of it sold-to-grid-during-the-day-and-epurchased-at-night).

    I will say I am back and forth on whether or not to bother with this battery backup; but I seem to be moving to a "if I can do it for ~ $2500 (a bit more than 10% of the post-rebate system cost), I will". Above $3000, nope.

    It seems I am close, but it will depend on whether the contractor is willing to collaborate.




    -Alex Aragon[/QUOTE]
  • ggunn
    ggunn Solar Expert Posts: 1,973 ✭✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...
    danielh wrote: »

    I will say I am back and forth on whether or not to bother with this battery backup; but I seem to be moving to a "if I can do it for ~ $2500 (a bit more than 10% of the post-rebate system cost), I will". Above $3000, nope.

    It seems I am close, but it will depend on whether the contractor is willing to collaborate.

    To cut to the chase, take a look at what would be your critical loads, how long you would expect to have to run them in the case of an outage, factor in minimum battery state of charge (normally 50%), and from that calculate how much battery you would need (and price it). That is normally the first step in designing an off-grid/grid backup system; all other design considerations flow from that analysis. That's a major difference between the two types of systems (grid tied and battery backup/off grid); the starting points for the design process are very different. If you design a grid tied system and then decide you want it to be a battery based system, you start over.

    An even shorter route to the chase may be this: I do not believe that you can do it for anywhere near $3000.
  • danielh
    danielh Solar Expert Posts: 69 ✭✭✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...
    ggunn wrote: »
    To cut to the chase, take a look at what would be your critical loads, how long you would expect to have to run them in the case of an outage, factor in minimum battery state of charge (normally 50%), and from that calculate how much battery you would need (and price it). That is normally the first step in designing an off-grid/grid backup system; all other design considerations flow from that analysis. That's a major difference between the two types of systems (grid tied and battery backup/off grid); the starting points for the design process are very different. If you design a grid tied system and then decide you want it to be a battery based system, you start over.

    An even shorter route to the chase is this: I do not believe that you can do it for anywhere near $3000.

    I should do a more formal design exercise, but I think the specs I noted are close to providing sufficient improvements. Again, I am willing to accept less than normal life during an outage (no 240Vac appliances, hour or so time periods where there are no power, staggering what I use when there is power). IOW: living with some inconvenience during short outages, and more inconveniences (and being mindful) during longer outages. Maybe I am being blithe, but the goal is to improve over the current haphazard coping strategy (admittedly, the current genset-only strategy is a lot better than the old one of hoping it doesn't get too cold or too hot)

    I should stress: the $3000 is NET of unpurchased equipment and fed rebate. I.e.; I won't buy a + $2000 grid-tie inveter, and then I also get the fed rebate.Which means a budget closer to $6500 for necessary EXTRA equipment and extra labor (i.e.; ignoring costs that don't change, such as panels and racking). Perhaps you are correct, this is way too low. I would be real nice to get a breakdown of actual expenses--- but that is something my contractors can do.

    Of course, if someone wants to do a cost breakdown exercise, or has one handy... that could be useful to see.

    I think I am now in a position to make a coherent set of requests to my contractors (they are willing to listen). Maybe the costs will be close, and I will be enticed to raise my limit. Maybe not, in which case I stick with my "segregated" system, and wait for prices to drop/ technology to mature.
  • mtdoc
    mtdoc Solar Expert Posts: 600 ✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...

    My 0.02 FWIW:

    I faced similiar choices last year before installation of my GridTie with battery back up system. Specifically Xantrex XW vs Outback GVFX. I did not have any plans for any strictly grid-tie only panels and did not consider Magnum with AC coupling. Here's the way I looked at it:

    Xantrex XW:

    Advantages: Higher 240 VAC output and wider generator input options (possibly?), PFC charger
    Disadvanges Higher cost, poor customer service, recent recall, no easy or inexpensive options for online monitoring.

    Outback GVFX:

    Advantages: Proven relieability, great customer service, good, inexpensive online monitoring software (Wattplot or
    Greenmonitor). Less expensive
    Disadvantages: No 240VAC output without stacking, Lower output, non PFC corrected charger. Need for inverter generator
    input.


    For the Outback I looked at their Flexpower prewired solution and Midnite Solar's prewired Epanel system (GVFX with Classic 150 CC).

    In the end I went with Outback and the Midnite epanel (link: http://www.solar-electric.com/misoprousy3w.html) and I've been very happy with my choice.

    Several points - the Midnite Epanel has the advatage of their top of the line SPDs, more flexible inputs for future expansion and of course the Midnite Classic 150 CC which IMO has several advantages over the Outback and Xantrex CCs. I highly recommend adding an Outback Flexnet DC to this system for battery monitor function and to control the Sell- Back parameters.

    The generator input issue was not a big deal since I already owed one Honda eu2000i (and it sounds like generator input on the XWS may be just as picky as well). I purchased a second eu2000i to parallel and to have enough juice to provide full battery recharge and load support if needed. And FWIW it turns out the non PFC charger issue on the Outback GVFX is a non issue. It seems even without PFC the GVFX charger charges a depleted battery bank at about a 0.9 PF.

    The lower output of the GVFX 3648 compared to the XW has not been an issue since my 120 loads never approach 3600 watts. The one 240 VAC load I have backed up is my septic pump (using an Outback PSX240 autotransformer) - this is a 1 HP pump with high inrush current and I was worried that the GVFX3648 might not handle it. It only comes on for about 15 seconds once a day. The GVFX handles it fine.

    Also - FWIW with the GVFX3648 wired as a grid-interactive back up system, selling excess PV production, there is no significant cycling of the batteries with the grid intact. They just are held at float and the grid power is "passed through" the inverter. The PV power obviously passes through the batteries to the inverter before supporting loads or being sold back to the grid but there is no draw down of the batteries with this.

    Bottom line - The XW is a proven system with lots of happy users, but the Outback GVFX inverter with the Midnite epanel and charger was the right choice for me.

    I would think a total- before rebate- budget of $6000.00 may come up just a little short for any on these depending on what the electrician charges and what you spend on batteries. The Midnite Epanel with FNDC will cost about $4000 and may reduce some of the "balance of system" and installation costs associated with other options.

    Good luck with your decision.
  • danielh
    danielh Solar Expert Posts: 69 ✭✭✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...

    Hello MtDoc.
    mtdoc wrote: »
    Xantrex XW:

    Advantages: Higher 240 VAC output and wider generator input options (possibly?), PFC charger
    Disadvanges Higher cost, poor customer service, recent recall, no easy or inexpensive options for online monitoring.

    Outback GVFX:

    Advantages: Proven relieability, great customer service, good, inexpensive online monitoring software (Wattplot or
    Greenmonitor). Less expensive
    Disadvantages: No 240VAC output without stacking, Lower output, non PFC corrected charger. Need for inverter generator
    input.


    For the Outback I looked at their Flexpower prewired solution and Midnite Solar's prewired Epanel system (GVFX with Classic 150 CC).

    In the end I went with Outback and the Midnite epanel (link: http://www.solar-electric.com/misoprousy3w.html) and I've been very happy with my choice.

    Several points - the Midnite Epanel has the advatage of their top of the line SPDs, more flexible inputs for future expansion and of course the Midnite Classic 150 CC which IMO has several advantages over the Outback and Xantrex CCs. I highly recommend adding an Outback Flexnet DC to this system for battery monitor function and to control the Sell- Back parameters.


    I wasn't aware of that option!

    I am still having a bit of difficulty wrapping my head around how to match "watts from PV" to "charge contoller capabilty". I had assumed that the Outback's FM80 would handle 3.6kw (since that is the inverter's rating), but perhaps that is incorrect. Similarly, will the Midnite Classic handle 3.6kw of PV power from my panels?
    Note that the helios 6t 260 STC specs are: Rated Power PMPP=260W, MPP voltage=30.85V, MPP Current=8.46A.
    mtdoc wrote: »
    The generator input issue was not a big deal since I already owed one Honda eu2000i (and it sounds like generator input on the XWS may be just as picky as well). I purchased a second eu2000i to parallel and to have enough juice to provide full battery recharge and load support if needed. And FWIW it turns out the non PFC charger issue on the Outback GVFX is a non issue. It seems even without PFC the GVFX charger charges a depleted battery bank at about a 0.9 PF.
    As much as I don't-really-like the 6500 Generac, it does work. Spending $2000 on dual EUs is a hurdle (even assuming I can sell the like new generac for $600 or so). Getting just one EU I could see, but I suspect that would be pushing the bounds -- I would run into trouble with smallish generator and a smallish battery bank.
    By trouble, I mean overloads that cause damage. Overloads that mean lights flicker, or generator cuts out without damage (one just restarts it)... I can live with (since they won't happen too often if one is careful about starting loads).

    Note that the Xantrex seems to have problems accepting power from the Honda EU, while the Outback only likes the HondaEu. That's kind of funny
    mtdoc wrote: »
    The lower output of the GVFX 3648 compared to the XW has not been an issue since my 120 loads never approach 3600 watts. The one 240 VAC load I have backed up is my septic pump (using an Outback PSX240 autotransformer) - this is a 1 HP pump with high inrush current and I was worried that the GVFX3648 might not handle it. It only comes on for about 15 seconds once a day. The GVFX handles it fine.
    No 240 VAC for me.. except for my current transfer switch!

    With the Outback, I might have to make a few choices (as to which of my critical loads remain critical).
    mtdoc wrote: »
    Also - FWIW with the GVFX3648 wired as a grid-interactive back up system, selling excess PV production, there is no significant cycling of the batteries with the grid intact. They just are held at float and the grid power is "passed through" the inverter. The PV power obviously passes through the batteries to the inverter before supporting loads or being sold back to the grid but there is no draw down of the batteries with this.

    Question: how well does the Outback act as a "most-of-house-UPS". For example, is the switch over time quick enough to prevent computers from crashing? The xantrex seems to get good marks on that score.

    Thanks for the info -- it helps!
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...

    When you're talking about grid-tie, the output to the grid is limited by the inverter's capacity. So you could have the 3.6 kW OB inverter with 10 kW of array behind it and you'd still only get 3.6 kW out. You'd get it even under slightly less than ideal insolation, but the inverter could never pass the full capacity of the array.

    So you have a charge controller capable of 80 Amps @ 48 Volts or 3840 Watts. The array can be slightly larger than that to accommodate normal panel efficiency losses. Manufacturers usually have their own array sizing tools and input limit figures.

    Then you have an inverter capable of 3.6 kW maximum, or about 200 Watts less than the array. Unless that array is on a tracker it will not produce its maximum all day long anyway. Allowing for that you might run a 4kW+ array and allow the inverter to "clip" any surplus power from the panels. This is one of the balancing acts that is dependent on local conditions.

    You also will have a battery bank that must be large enough to handle things like AC ripple current and output/load fluctuations. The number typically used on 48 Volt systems is 100 Amp hours per 1kW of inverter, or 360 Amp hours @ 48 Volts minimum. Since that battery bank requires only about 2.2 kW of array to recharge it you can increase the capacity to have more power available when the grid goes down (up to, say, 800 Amp hours with maximum array on the controller).

    Not easy to decide what to get.

    By the way, the Outback's switch over is as near to instantaneous as can be. Although I use the off-grid version, I can disconnect the generator and there is not even a blip in the lights as the inverter picks up the load.

    And the Outback can have 240 via an autotransformer or through stacking a second inverter if need be. It does not care much what kind of generator it's hooked to either, so long as the Voltage and frequency are close to right. It's highly programmable and more tolerant of error than Xantrex in my opinion. The only bit I don't like is the lack of end Amps on the built-in charger, but the accessory equipment can solve that.

    If you do go with Outback, you should consider going with both their controller and inverter and connect them via the HUB. Their equipment integrates better than any other I've seen or heard of.
  • mtdoc
    mtdoc Solar Expert Posts: 600 ✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...
    danielh wrote: »
    I am still having a bit of difficulty wrapping my head around how to match "watts from PV" to "charge contoller capabilty". I had assumed that the Outback's FM80 would handle 3.6kw (since that is the inverter's rating), but perhaps that is incorrect. Similarly, will the Midnite Classic handle 3.6kw of PV power from my panels?
    Note that the helios 6t 260 STC specs are: Rated Power PMPP=260W, MPP voltage=30.85V, MPP Current=8.46A.

    Either CC should handle 3.6 kw of PV. The exact amount of PV will depend on the array configuration which will depend in part on the Voc of your panels, temperature extremes, etc. Based on the numbers you give, I would guess 5 strings of 3 panels = 3900 watts would work well, WIth derating the inverter is unlikely to see over 3600 watts. On the Midnite Classic you can set the output amp limit to be sure (edge of cloud effects could potentially put you over 3600 watts on a cold day). I'm planning on using that function myself later this year when I add another PV array. I'm not sure if the FM80 lets you limit the output amps.

    Midnite and Outback have string sizing tools you can play with to decide the best array configuration. Midnite's is here: http://www.midnitesolar.com/sizingTool/index.php

    As Cariboocoot states Outbacks integration is very good and using the FM80 CC does have some advantages that way. Honestly though I have not missed that capability at all with the Midnite CC and Epanel. The CC and the inverter communicate via the battery. With the FNDC (and Hub), the control of battery voltage settings, charging stages, end amps and selling parameters works very well and the Midnite Classic follows along just fine. And the Midnite local app is a great tool to monitor your PV production and battery charging,etc.

    As much as I don't-really-like the 6500 Generac, it does work. Spending $2000 on dual EUs is a hurdle (even assuming I can sell the like new generac for $600 or so). Getting just one EU I could see, but I suspect that would be pushing the bounds -- I would run into trouble with smallish generator and a smallish battery bank.

    One nice thing about the Midnite Epanel is that the AC input has a bypass switch to bypass the inverter. So you could use your Generac to power your critical loads through the Epanel- crtitical load subpanel - without going through the Outback inverter. You would not be able to charge your batteries with the GVFX under this scenario but could charge them directly with a stand alone battery charger powered by your generator.


    Question: how well does the Outback act as a "most-of-house-UPS". For example, is the switch over time quick enough to prevent computers from crashing? The xantrex seems to get good marks on that score.

    Thanks for the info -- it helps!

    As Cariboocoot says, the outback works great as a UPS - no problem with computer reboots, etc. In fact the only reason we've noticed when the power has gone out is that the thermostat/controller on our heatpump/furnace (which is not backed up) turns off.
  • danielh
    danielh Solar Expert Posts: 69 ✭✭✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...

    Hello cariboocoot

    I realize I might need to take the battery-sizing rules more seriously. So, I think of your comment:
    You also will have a battery bank that must be large enough to handle things like AC ripple current and output/load fluctuations. The number typically used on 48 Volt systems is 100 Amp hours per 1kW of inverter, or 360 Amp hours @ 48 Volts minimum.

    If this 100AH rule is dangerous to break, ugh. To use half of what my system pv output (~3kw) would require 300 AH; using sun-xtender AGM that would cost around $3200. Or to use the 4.2kw I was thinking of, $4500.

    Now, that size of a battery bank may be better for practical purposes, but it starts to make me wonder if it is worth it.

    I could cut back on the panels connected to the batteries -- say, to 2kw -- and install 200AH (for about $2400). That bends the budget, but not impossibly.
    But then I only have 1/3 of PV power during outages; and part of the reason for batteryizing is to not lose the advantage of PV panels during outages!

    Let me posit an alternative view. The battery chargers (either the PV connected charge controller or the inverter-embedded one) are smart enough to only
    charge when necessary. In a grid-interactive setting, where outage backup and UPS protection are the main tasks, battery usage will be infrequent. Hence,
    battery charging will be infrequent.

    And, if you limit charging, ripple effects won't occur very often, so violating the 100AH rule isn't so onerous.

    Thus, by oversizing the PV power relative to battery power, one trades off "more PV power during outages" for "less battery life".
    But since in this grid-interactive setting battery life is probably going to be effected more by aging then by usage, this may be a beneficial tradeoff.

    IOW: perhaps the 100AH rule is something that off-gridders need to pay attention to. Less so for grid-interactive applications.

    Or am I just in denial?
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...

    Hybrid inverters do not function like standard GT inverters when in GT mode. Leave out the battery capacity and watch it stop working. There have been numerous posts on the forum of, by, and for people who have installed XW 6048 inverters with insufficient battery capacity and all the trouble that comes with trying to get the thing to do a stable sell-to-grid. Off-grid systems have no qualm about battery bank size other than being able to support loads: no AC ripple coming back from the grid.

    As far as I know there's no law that says you can't have part of your GT system as "standard" and part of it as "hybrid" if such meets your needs. Just enough battery-based system to keep critical systems going during outages and the big money in regular GT.

    Or you can go with the SMA system; install the GT Sunny Boy inverter(s) now and add a Sunny Island & batteries when possible.
  • danielh
    danielh Solar Expert Posts: 69 ✭✭✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...

    Hello mtDoc
    mtdoc wrote: »
    Either CC should handle 3.6 kw of PV. The exact amount of PV will depend on the array configuration which will depend in part on the Voc of your panels, temperature extremes, etc. Based on the numbers you give, I would guess 5 strings of 3 panels = 3900 watts would work well, WIth derating the inverter is unlikely to see over 3600 watts. On the Midnite Classic you can set the output amp limit to be sure (edge of cloud effects could potentially put you over 3600 watts on a cold day). I'm planning on using that function myself later this year when I add another PV array. I'm not sure if the FM80 lets you limit the output amps.

    Midnite and Outback have string sizing tools you can play with to decide the best array configuration. Midnite's is here: http://www.midnitesolar.com/sizingTool/index.php
    Still up in the air, but probably use a two strings (9 and 8 ).
    mtdoc wrote: »
    As Cariboocoot states Outbacks integration is very good and using the FM80 CC does have some advantages that way. Honestly though I have not missed that capability at all with the Midnite CC and Epanel. The CC and the inverter communicate via the battery. With the FNDC (and Hub), the control of battery voltage settings, charging stages, end amps and selling parameters works very well and the Midnite Classic follows along just fine. And the Midnite local app is a great tool to monitor your PV production and battery charging,etc.

    mtdoc wrote: »
    One nice thing about the Midnite Epanel is that the AC input has a bypass switch to bypass the inverter. So you could use your Generac to power your critical loads through the Epanel- crtitical load subpanel - without going through the Outback inverter. You would not be able to charge your batteries with the GVFX under this scenario but could charge them directly with a stand alone battery charger powered by your generator.
    Intersting (say, in the evening when the sun is down). About $300 for an IOTA (or the Meanwell)? In fact the Outback people suggested that(their tech support really is nice, then answer my dumb questions even though I am not an owner)!

    That said, per my response to carib... I am now wondering how seriously I need to take the 100AH per kw rule (it might just be one hurdle too many).
  • danielh
    danielh Solar Expert Posts: 69 ✭✭✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...
    Hybrid inverters do not function like standard GT inverters when in GT mode. Leave out the battery capacity and watch it stop working. There have been numerous posts on the forum of, by, and for people who have installed XW 6048 inverters with insufficient battery capacity and all the trouble that comes with trying to get the thing to do a stable sell-to-grid. Off-grid systems have no qualm about battery bank size other than being able to support loads: no AC ripple coming back from the grid.

    As far as I know there's no law that says you can't have part of your GT system as "standard" and part of it as "hybrid" if such meets your needs. Just enough battery-based system to keep critical systems going during outages and the big money in regular GT.

    Or you can go with the SMA system; install the GT Sunny Boy inverter(s) now and add a Sunny Island & batteries when possible.

    Ah, I was wondering where the AC in AC ripple came from (the grid!). Makes a little more sense.

    So you are saying: even if the battery capacity is rarely used, sizing does matter -- improper sizing won't just effect battery life, but also grid sales.
    And grid sales is the bigger apple.

    Also: instead of $MA, some other AC coupling solution (like the vapor warish solutions posted by Magnum) could be effective (assuming it advances beyond the vapor stage).
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...

    The "ripple" we are worried about is the DC ripple voltage on the battery. A single phase inverter outputs its power in a (sine)2 (sine squared) wave form... Basically, when the voltage/current is at Zero, there is no power transfer. And when the voltage/current is at the + or - peak of the sine waves, the power flow is a maximum (basically a 120 Hz sine squared wave form).

    It turns out that to store energy locally, you typically use a high voltage capacitor (i.e., for standard computer supplies, the capacitor votlage is around 370 volts peak. Or:
    • storage energy in capacitor = 1/2 * CV2
    From the equation above, to store enough energy inside a inverter input capacitor at 12-48 volts or so would require huge input capacitors. Instead, the inverter vendors just use the battery bank as their energy storage.

    Typically, the "sell voltage" or inverter operational voltage is somewhere around a minimum of 52-54 VDC (for 48 volt bank, or ~13.0 to 13.5 volts for a 12 volt bank). This allows the GT/Hybrid inverters to pull the voltage down to no lower than ~12.7 volts (basically the "surface charge" and battery capacitance needs to keep the ripple voltage minimum above ~12.7 volts--otherwise, you are "micro cycling" the battery bank at a 120 Hz (for 60 Hz AC system).

    If you have too small of battery bank, your issues become:
    • micro cycling (not enough surface charge/battery capacitance to keep battery from "discharging/charging" during an AC power cycle)
    • Surge capacity (typically around 2x maximum inverter Watt Rating) is insufficient--causes inverter to hit battery cutoff voltage when trying to start a well pump, etc. AGM batteries do have very good surge capabilities (and even the ability to accept very high current levels during charging). However, even with AGMs, they well stop accepting current when "fully" charged and go "high voltage" if there is too much current (why I still suggest the 100 AH per 1kW even for AGM when charging at 10-13% rate of charge).
    • During charging--It appears that (at least some) MPPT charge controllers go "uncontrolled" output during the MPPT search cycle (trying to find Pmp=Vmp*Imp array operating point). Basically, the charge controller to find Vmp-array appears to dump Imp*Vmp into the battery bank for a short period of time while doing the search. With too small a battery bank, the batteries "lose regulation" and we have read about >72VDC battery fault voltages with too small of battery bank (AGM or Flooded Cell). A second theory I have is that the control loop on the MPPT charge controllers may be on the order of 2-5 seconds... If there is a inverter surge load, cloud across array, etc., the controller may dump maximum available current into the battery bank for a hand full of seconds--Again, if the battery bank is already fully charged, the bank voltages goes uncontrolled (over 72 VDC on a 48 volt bank--which should normally be around 62 VDC max during normal/equalization operation). Again--this is my guess--I have not talked with anyone who really knows what is going on with the various brands of charge controllers.

    Anyways, my personal observations on why it is a "bad idea" to ignore the 100 AH per 1kW of load (and available charging power) for standard system designs.

    If you have a special need--Such as a UPS type operation (large loads, and "small" recharging current relative to battery AH capacity), then the rules can be "bent".

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • mtdoc
    mtdoc Solar Expert Posts: 600 ✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...

    Wow, great explanation Bill! I've learned so much reading yours and others posts here.
    danielh wrote: »
    Hello mtDoc

    Still up in the air, but probably use a two strings (9 and 8 ).

    ?? If you mean one string of 8 panels in series and one of 9 in series that won't work. Depending on the Voc of your panels your strings would likely need to be between 2 and 4 panels for any of the mppt CCs you're considering (except the 600V Xantrex CC option).
    Have a go with the string sizing tools.

    Coot and Bills points about battery bank size are important. In the final analysis doing a "Grid Interactive" aka "grid tie with battery back up" system will be unlikely to ever make sense from a purely economic standpoint. It can make sense from a peace of mind- reliable power perspective. As a bonus some even find it fun!
  • danielh
    danielh Solar Expert Posts: 69 ✭✭✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...
    mtdoc wrote: »
    ?? If you mean one string of 8 panels in series and one of 9 in series that won't work. Depending on the Voc of your panels your strings would likely need to be between 2 and 4 panels for any of the mppt CCs you're considering (except the 600V Xantrex CC option).
    Have a go with the string sizing tools.
    !
    That was the plan: the 600V Xantrex.

    But I think I have learned the error of my ways: that getting cheap on the AH to kw-of-PV-input is a no no. So I am thinking of scaling back to 1.5 to 2.5kw of panel, to a 150 or 200 AH battery pack. Not sure how I would size the strings (the layout of panels is odd, sets o 3 5 4 4 and 9 scattered over the back of the house ... but that's a whole other story); but that should be solvabble.

    I did try the Midnite string sizing tool, but I couldn't really figure out what it meant. The inputs made sense .. but the output ... lots of rows, several columns, 0 documentation.
    mtdoc wrote: »
    Coot and Bills points about battery bank size are important. In the final analysis doing a "Grid Interactive" aka "grid tie with battery back up" system will be unlikely to ever make sense from a purely economic standpoint. It can make sense from a peace of mind- reliable power perspective. As a bonus some even find it fun!
    Yeah, I see that. And both peace of mind and fund ARE worth something -- but not many k$ of something. Hence the search to find some kind of compromise between effectiveness and cost. Again, the goal is to do better than now, with the expectation that outages will still be a pain in the neck that have to be actively managed.


    BTW: would AC coupling avoid some of these issues, at least to the extent that one could have immediate use of all PV power when the grid was down, even if only a fraction of such power could be used to charge a moderately sized battery bank? Of course, this would require a cost effective means of implementing "only a fraction of such power ... " (where SMA is a bit to expensive for my threshold of cost effective).
  • stephendv
    stephendv Solar Expert Posts: 1,571 ✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...

    If I were in your shoes, I'd take a cautious approach to this project. What you're 100% sure of is installing a grid tie system. So install a Sunny Boy inverter for the unshaded part of the array and keep it to about half the total array size. Install micro-inverters or another sunny boy for the shaded part. Make sure you can easily access the AC from the sunny boy, i.e. if you want to only connect half the array then you can.
    Then once you have the system installed and all is right with the world, consider AC coupling part or all of it. Since you have a Sunny Boy(s) you still have the option of using a sunny island if you want (remember for a 6kW array you only need a 3kW sunny island), or if not you can always using the xantrex or magnum solution with only half the total array capacity. Or you could even use a dump load controller and AC coupling with any bi-directional inverter as another option.
    Point is, I wouldn't base the system on being a battery based system when that's the part you're least sure about, and will be the least used part of system.

    2 cents worth ;)
  • stephendv
    stephendv Solar Expert Posts: 1,571 ✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...

    Just to expand a bit on the AC coupling, you basically have 3 options:
    1. Variable frequency power control ala SMA sunny island + Sunny Boy
    2. Disconnect GTI through over frequency ala Xantrex, Magnum (I believe) and Victron
    3. Send excess power to a dump load across the battery using any bi-directional inverter/charger: Xantrex, magnum, victron, outback and probably more.

    This last option would work as follows in a pure off-grid setup:
    GTI feeds the inverter/charger with all available power. Inverter/charger passes it on to the battery. Dump load controller (like morningstar tristar or xantrex C60) diverts any excess power into DC dump loads which are usually just chunky resistors. The dump load controller then controls the battery charging by throwing away excess power during the absorb and float stages. So if you don't trust the frequency shifting control that the Magnum's offer you could use this as an alternative configuration.

    To use this in a grid interactive setup, you're going to have to have to use a transfer switch to switch the grid from the AC in connection on the inverter/charger to the output of the inverter/charger when the grid goes down. So with this configuration you could use a 4kW magnum for $2100 + half your array on a 4kW GTI + battery of > 400Ah + 2 x Xantrex C60 dump loads ($320 total) + DC dump resistors capable of dumping 4kW). Using this configuration it'll work as follows:
    While grid is up:
    - GTI feeds directly to the grid.
    - Grid is connected to the AC-in of the inverter/charger
    - Batteries charge normally using the regulated charger of the inverter/charger
    - Output of the inverter/charger goes to a SEPARATE load circuit in your house. When the grid goes down, these loads will continue to be powered.

    When the grid goes down:
    - Throw the transfer switch to move the GTI to the output of the inverter/charger
    - Batteries charge by being backfed through the inverter/charger and dump controller throws away excess power

    Looks like you've hit the 4 cents bonanza today!
  • danielh
    danielh Solar Expert Posts: 69 ✭✭✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...
    stephendv wrote: »
    Just to expand a bit on the AC coupling, you basically have 3 options:
    1. Variable frequency power control ala SMA sunny island + Sunny Boy
    2. Disconnect GTI through over frequency ala Xantrex, Magnum (I believe) and Victron
    3. Send excess power to a dump load across the battery using any bi-directional inverter/charger: Xantrex, magnum, victron, outback and probably more.
    A Magnum 2010 white paper states (http://www.wholesalesolar.com/pdf.folder/inverter%20pdf%20folder/MagnumACcoupling.pdf):

    Note 4: The battery bank voltage can be managed by installing a
    diversion controller and load capable of absorbing the majority of
    the expected surplus energy, using: 1. DC diversion controller
    and DC resistance loads; 2. AC diversion using AC resistance
    loads driven by DC controlled relays; or 3. Magnum’s AC Load
    Diversion Controller. Now in development, Magnum’s AC Load
    Diversion Controller will allow customers the ability to maximize
    the use of onsite generated renewable power and provide
    multiple-stage battery charging with temperature compensation
    using AC loads. For more information on the release of this AC
    Load Diversion Controller, continue to monitor our website at
    www.magnumenergy.com
    Unless Magnum has made progress in 2 years (one hopes they would have), even they aren't sure!
    This last option would work as follows in a pure off-grid setup:
    GTI feeds the inverter/charger with all available power. Inverter/charger passes it on to the battery. Dump load controller (like morningstar tristar or xantrex C60) diverts any excess power into DC dump loads which are usually just chunky resistors. The dump load controller then controls the battery charging by throwing away excess power during the absorb and float stages. So if you don't trust the frequency shifting control that the Magnum's offer you could use this as an alternative configuration.
    How does the dump load controller know when enough is enough -- does it monitor the battery SOC?
    To use this in a grid interactive setup, you're going to have to have to use a transfer switch to switch the grid from the AC in connection on the inverter/charger to the output of the inverter/charger when the grid goes down. So with this configuration you could use a 4kW magnum for $2100 + half your array on a 4kW GTI + battery of > 400Ah + 2 x Xantrex C60 dump loads ($320 total) + DC dump resistors capable of dumping 4kW). Using this configuration it'll work as follows:
    While grid is up:
    - GTI feeds directly to the grid.
    - Grid is connected to the AC-in of the inverter/charger
    - Batteries charge normally using the regulated charger of the inverter/charger
    - Output of the inverter/charger goes to a SEPARATE load circuit in your house. When the grid goes down, these loads will continue to be powered.

    he white paper above shows the GTI feeding the "seperate load circuit" subpanel. So power from the GTI could either go to the subpanel loads, or be back fed into the inverter.
    Perhaps your setup reduces the chances of overcharging the battery due to unregulated current from the GTI (since there is no back feed into the inverter/charger when the grid is up)?
    When the grid goes down:
    - Throw the transfer switch to move the GTI to the output of the inverter/charger
    - Batteries charge by being backfed through the inverter/charger and dump controller throws away excess power
    Interesting. You have to take action (throwing a switch); so it isn't automatic. But so what, I have to do more than that now to switch to generator.

    Just to clarify (correct me if I am wrong):
    The SEPERATE load circuit will automatically/immediately get inverter power (from the battery) when the grid goes down -- so the inverter/charger acts
    as a UPS for these loads.
    Under normal circumstances, the GTI is feeding power to loads NOT on the critical panel (which includes out to the grid).
    When the grid is down and you flip the switch, the GTI is charging the batteries (albeit in an unregulated fashion) and feeding power to the subpanel (the inverter/charger's internal transfer switch will not permit backflow to the main cb panel).

    Overall, interesting. Only slight drawback is what happens to all this "dump resistor" heat. If the resistor is in the basement, and its a summer outage (like a week ago)-- who needs MORE heat in the house! Not a big deal... but could one safely just switch the GTI back to grid if one were to see that the batteries were fully charged (and switch it back when they are sufficiently discharged).

    One thing I want to check on: the "100 AH to 1kw of output" rule still holds.
    Thus ...
    ...since GTI output must be < less than inverter/charger output, and since the inverter/charger output should obey the 100AH rule,
    then ...
    ...the GTI output should be less than 1kw per 100AH of the battery pack
    If so ... that could be inconvenient, given that the GTI has nothing to do with the batteries 99% of the time.


    Looks like you've hit the 4 cents bonanza today!
    Is that better or worse than a wooden nickel?
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...

    Normal emergency back-up power with an off-grid inverter:

    AC Main --> OG inverter ---> critical loads

    AC Main goes down, inverter takes up critical load without a blip. While AC is available it will recharge the batteries through the inverter and run the loads. Just like most UPS units.

    Add GT into the mix and you have:

    GTI ---> AC Main ---> loads
    Where the loads would include the separate OGI and critical devices. When the AC Main goes down, the GTI stops working. You then need to manually switch the circuit to this:
    GTI ---> critical loads <--- OG inverter

    Any power the GTI produces in excess of the critical loads will back feed the OG inverter causing it to recharge its batteries. If the batteries are fully charged that excess power has to go somewhere, hence the need for two (by NEC regs) dump load controllers to take power away from the batteries and use (or waste) it somewhere else (resistive heating load in a hot water tank is common). There is no sure-fire regulation on this type of system and the dump load Voltage level is inevitably different from the Absorb set point of the batteries. This will interfere with proper charging, but for short-term is not devastating to the batteries. This differs from a hybrid inverter's function in that it "knows" when the batteries have been through the proper charging sequence.

    Not the easiest, safest, or most dependable method in my opinion.
  • stephendv
    stephendv Solar Expert Posts: 1,571 ✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...
    There is no sure-fire regulation on this type of system and the dump load Voltage level is inevitably different from the Absorb set point of the batteries. This will interfere with proper charging, but for short-term is not devastating to the batteries.

    Not the easiest, safest, or most dependable method in my opinion.

    'coot, there is sure fire regulation in the system: the dump load controllers. Every off-grid wind turbine installation uses this configuration with a dump load controller. What's more, the dump load controllers regulate the bulk, absorb and float stages exactly the same way a normal PWM controller does (in fact, they're the same controllers)- the only difference is that the PWM signal goes to the dump load, not to the battery. Perhaps you're referring to dumping on the AC side, which is more hit and miss?
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Cost effective grid-interactive choices: Outback vs Xantrex vs magnm vs ...
    stephendv wrote: »
    'coot, there is sure fire regulation in the system: the dump load controllers. Every off-grid wind turbine installation uses this configuration with a dump load controller. What's more, the dump load controllers regulate the bulk, absorb and float stages exactly the same way a normal PWM controller does (in fact, they're the same controllers)- the only difference is that the PWM signal goes to the dump load, not to the battery. Perhaps you're referring to dumping on the AC side, which is more hit and miss?

    Yes; I do not trust AC coupling on inverters that were never meant to be used this way. Even in normal operation there are sometimes issues with multiple charge sources causing momentary high current to the batteries. When a GT inverter tries to "test the grid" against one of these by pushing the Voltage up, the current to the batteries from all sources could rise suddenly and the battery Voltage climb. It would be important here to size the battery bank the same as you would for a hybrid GT and be sure the sump load controllers were capable of taking the max potential current just in case. The problem there is that an inverter such as the FX3648 can run 80 Amps DC (about 32 AC IN) through its charger, whereas most controllers are 60 Amps. Very important to get the sizing right on everything here. I don't think you'd want to run a GT system that is 3X the size of the OG inverter used.