New System Design Questions

System

Hi, I'm a complete noob to the solar family but having purchased property in the Canadian Gulf Islands & not being on the grid I sharpened my pencil, did some calculations & ended up with a number of questions. If anyone can answer them or add constructive comments it would be greatly appreciated.

Assumptions:
1) Kyocera panel @ 130 watt - 7.4 amp & 17.6 V
2) attempting to equal an on grid consumption of 300 KWhr / month - I think this may be low for a permament residence but multiples of this value are probably arithmetic
3) assuming 4 hrs complete sun per day - again may be low

Questions:
1) using panel tracking I've found references that indicate possible gains of 15% winter & 35% summer raising the effective # sun hrs & therefore the panel output (possibly 150 - 170 watt equivalent) --- true??
2) using MPPT get gain 20-45% winter & 10-15% summer -- equivalent avg say 170 - 145 watt -- true
3) If 1 & 2 are true are the effects cumulative?? say tracking @15% & MPPT @20% = 130*1.15*1.2=180 watt equivalent
4) based on the above panels could I put 24 in // series to get 48V array (my calc say this would be 44.4 amps & 70.4 V fot approx 3130 watt - then use an MX60 controller (specs are 3200 watt at 48V - or would I have to use the 70.4V number)
5) If I had a 24V battery bank can I tap it with 12V inverters or do I have to use 24V (can I combine both) - want to use cheap mod sine for most load but maybe a good sine for critical applications & also use to connect generator directly for charging
6) better/cheaper to have stand alone generator for high load & use separate charger off generator for charging as compared to item #5??
7)Can I supply a 240V AC breaker panel with 24 breakers using 2 different inverters (not stacked) by connecting both to the panel common & one to each of the 120V poles feeding 12 breakers with each inverter? Can it be taken further by cutting the breaker common hot bar - have say 4 inverters feeding 6 breakers each?
8) If I want to drive a 240V deep well pump can I take one of the line out from the panel (120V) & use a step up transformer - thus having a 240V line on a 120V breaker?

Sorry for all the questions. Any help appreciated.

Thanks
Warren

halfcrazy

Re: New System Design Questions

i would say go for a good sine wave inverter as you dont want two 12 volt inverters on a 24 volt bank. the mx60 is limited by its output not its input so the 3130 watts at 24 volt would overload the mx 60. and no you cant hook two inverters into the subpanel to get 240 unless they are made to do that and stacked like the bigger sinewave inverters do. but if you get a decent sinewave inverter you can get a transformer to make 240 for a well pump so you could get by with 1 inverter.

niel

Re: New System Design Questions

warren,
i'll do my best here to answer your questions.

1> i don't know the exact gains you may realize, but the figures you list seem a bit high. you may gain more with more pvs than a tracker and you have to watch that the winds wouldn't destroy it as has happened in another thread from somebody in california.
2> this i know is high as the average mppt gain is about 10% reguardless of the season. it is good to keep losses low and using a good pv design voltage. ie not too high and not par.
3> yes. if you gain on both it multiplies by both.
4> i'm not 100% sure what you're asking here, but 4 pvs are needed in series for a 48v pv system. the max current would be 6x7.4=44.4amps. stc wattage rating for all is 24x130=3120w total. this would be doable for the mx60. a better way may be 5 pvs in series times 5 as you will have some cloudy day gainbacks along with more power recouped from the mornings and evenings.
5> you would be better off using the same voltage for both inverters as imbalances in the battery bank could occur and it would be more costly as it would be running more electronics to a run the 12v. if you intend 24pvs as stated in your previous 48v question this is fine, but be aware the controllers are limited by their current capabilities so at 24v outputted you would have half the wattage as for the 48v outputted system. example: with 48v at 50amps this is twice the wattage as 24v at 50amps. the 48v battery system affords you a benefit in needing less controllers to do the job as you'd need 2 for the 24v battery system to handle the same current.
6> i'm not sure of your question. is it my guess that you'd like to use a built-in charger like in many gridtied inverters to charge the batteries via generators?
7> i have heard that it has been done if i remember right, but i'd advise to go with the same inverters properly stacked. if out of phase with each other no 240vac is possible from 2 inverters. a high voltage transformer can bump it up to 240vac from 120vac and would be limited by the transformer's capacity or the inverter's capacity whichever is lesser.
you can split breakers all you want to, but you still only have whatever power you input to the panel to work with and drawing on all breakers could pop the main breaker/fuse. more than 2 inverters have been utilized in systems, but i'll refer you to boB if he'll chime in for you as he's designed and worked with systems like that. he may also shed better light on some of the answers your looking for.
in seeing halfcrazy had posted i see a much simplified version of my attemp at answering and maybe he understood some of your questioning better than i did.

System2

Re: New System Design Questions

Thanks for the reply halfcrazy
Regarding my original #7 question - I don't want to have an available 240V at the AC subpanel but rather 2 distinct 120V sources one on each side of the panel - each fed by a separate inverter (inverters sharing only the subpanel common) - I realize the 120V sides would not be in the correct phase.
regarding #4 if the mx60 is limited by the output then couldn't I use a 48V battery bank instead of 24V

I am looking to do this in the most cost effective way realizing the biggest cost is going to be the array.

Thanks again

System2

Re: New System Design Questions

Thanks niel

RE
#1) expected with high cost of PVS the $ effective gain would be better realized with tracking - may not be true I guess
#4) if I interpret you correctly you suggest a 5X5 array - 37Amp & 3250 watts - using the mx60 is OK
#5) I believe you are suggesting a better alternative would be 60V into the controller from the array & out to a 48V battery bank - again supported my the mx60 - I believe you are saying that if I go with a 48V battery bank I can forget using cheap 12V mod sine inverters?
#6) again I was interested in the cheapest not still reliable way of supplying generator charging to the batteries when required & supplying extra AC when demanded by heavy loads - My alternatives appeared to be supplying 120AC from the generator directly to a combo inverter/charger OR driving a remote charger than directly feeds the batteries.
#7) I really don't care about having true 240V at the AC subpanel - 2 (or more) separate mod sine 120V is fine. My intention was to use a number of 12V mod sine inverters. Looks like I'll have to revise that & see if I can find resonably priced mod sine inverters that can be fed from 48V. The 240V single circuit for the well pump can still be obtained from a single 120V cicuit out from the subpanel, correct? If not I may look more closely at using a 120V pump (deep well @ 300 Ft).

Thanks again for the input, you've given me much food for thought.

Regards
Warren

niel

Re: New System Design Questions

when doing this stuff it's almost never cheap. a 48v battery system does allow more input watts to be utilized and that you comprehended. now for using cheap 12v inverters you could put one on either side of the breaker box, but you may then need more controllers as the 12v output would be limited to the current maximum of the mx60. i believe boB mentioned they have then up to 75amps, but out of the box the max is 60amps. to make it more complex the nec limits the input current when voltages are par to 48amps or 25% less. when you mod the mx60's programming this can up it to .75x75=56.25amps.
do know most cheap modsine inverters don't stack, but the more expensive ones might. you might as well go with something of quality if you can, but i understand about the finances part. look over the inverters at naws' store. see what's available at what voltages for what $. as to the generator you could charge the batteries through the built-in chargers of the better inverters from a generator. some have had bad luck with output amps on some of the cheaper ones. others can kick some feedback as to what works and don't for them. you could also use a cheap inverter that has no charger and buy a charger or chargers seperately.
as to #7 again you could have 2 random phased modsinewave inverters on a breaker panel. they will just operate independantly of each other and no 240vac capabilities. if you need big amounts of power beyond the capability of the 2 cheap modsinewave inverters a more expensive stacking of sinewave inverters may be needed unless you wish to get another breaker panel and run 2 more cheap inverters independantly. as was said a transformer can give you the 240vac. do some looking and you'll grasp more on the abilities of the products out there and you can throw something together on your own if you know what you want will be served by the equipment.

crewzer

Re: New System Design Questions

Warren,

I think we need to clarify certain specifications and performance parameters for you, and you then need to review items 1 - 4 with a newly sharpened pencil.

The Kyocera specs are based on so-called Standard Test Conditions, or STC. The key condition is that the PV module’s cell temperature is 25 C (77 F). Since the full-sun cell temp can be ~35 C above ambient, that means the Kyocera can only be expected to deliver full power at mid-day on a bright, sunny and cold (~-10 C) day and if clean and properly aligned.

As a practical matter, the module will deliver full current but reduced voltage in normal ambient conditions. Accordingly, you can expect to see ~85% performance(~15 Vmp) in the Summer, and ~95% (~17.25 Vmp) in the Winter.

A tracker can indeed help improve an array’s performance. However, rather than think of the improvement as more Watts (power), the correct view is that of more energy (Watt-hours, or Wh). For example, a fixed 130 W module over four hours of “full Sun” in the Summer might produce 130 W x 85% x 4 hours = 442 Whrs gross. If the tracker extends the day to, say, six hours of “full Sun”, then the calculation is 130 W x 85% x 6 hours = 663 Whrs gross.

MPPT gains are factored on top of instantaneous performance of a controller without MPPT, and not with respect to manufacturer’s specs for their PV module. MPPT works best on a cold, bright day (high PV array voltage) when the batteries are low. In other words, if there’s “additional” PV voltage to convert to extra charging current, the MPPT controller will help.

You might see a 30% current gain the morning, but array current is low so the absolute gain is low. Also, note that MPPT is typically of no gain when charging in the Absorption- or Float stages, as the controller (MPPT or PWM) is actually reducing charge current in these modes. So, depending on other factors, you may see a daily net energy production gain of 5% to 10% in the Winter, and perhaps 0% to 5% in the Summer, both with respect to a non-MPPT controller.

The combination of a tracker and an MPPT controller will not raise the PV module’s effective power output from 130 W to 180 W. What they will do is allow you to produce extra energy during the day, especially in the Winter, when the tracker will help make the most of what little sunlight there is, and when the cold PV array is delivering higher voltage that can be converted into additional charging current.

The MX60 can handle a ~3200 W array when charging a nominal 48 V battery. For a nominal 24 V battery, it can handle an array rated at ~1600 W. A stable and reliable configuration for your location would be three modules wired in series (3 x 17.6 = 52.8 V Vmp STC; 65.7 Voc STC), and 4  series strings wired in parallel (4 x 7.39 A Imp = ~30 A) –- a total of 12 PV modules -- all connected to the MX60 controller. In addition to being an MPPT controller, the MX60 is also a DC-to-DC converter, so it'll take a "high" PV array voltage and down-convert it to the nominal battery voltage. Under ideal conditions, this 1,560 W array could deliver ~53 A while charging your “24 V” battery bank at 29 V. I’d expect typical summertime charge current to be ~45 A.

This search string may help you find insolation data for your location is Canada.

Following up…

300 kWh/month = 10 kWh/day. Assuming 4 hours/day of “full Sun” and ~67% overall system efficiency, you’ll need a PV array rated at 10 kWh/day / ~67% x 4 hours/day = ~3730 W STC. With a nominal 48 V battery bank, a single MX controller could handle a 3,250 W STC array assembled from 25 KC130 modules wired in a 5 x 5 configuration and stay within the electrical specs and safety margins for the MX60 and the U.S. National Electrical Code. You’ll need to check on Canadian code requirements.

The typical summertime charge current from the controller and such an array would be ~48 A at 58 V. An appropriate battery bank size would be ~1,000 Ah at 48 V nominal. The effective capacity would be ~50 kWh, which means you could autonomously deliver your 10 kWr/day for three days (no Sun) before the batteries were discharged to a 40% state of charge (SOC). That’s a significant discharge and shouldn’t be done too often. At that point, you’d need Sun, a generator, or both.

HTH,
Jim / crewzer

niel

Re: New System Design Questions

just 2 notes on what crewzer has said. 1 is i have been using 90% overall for the ptc average for simplicity, but as jim pointed out it is lower in winter and higher in the summer. 2 is i would use a 12v module in series for every 12v multiple of the battery bank. 48v battery bank equals 4 pvs in series. i told you of using 5 to gain a slight advantage during cloudy times and morning/evening times.

crewzer

Re: New System Design Questions

Niel,

It’s “higher in Winter, and lower in Summer”…

There’s good technical reasoning behind my recommendation for five "12 V" PV modules per series string when using the MX60 controller and a "48 V" battery bank. Trojan’s recommended EQ voltage for a 48 V bank of their flooded-cell batteries is 62 V. :-o Allowing for a voltage drop in the PV array wiring, the MX’ DC-DC converter, the controller-to battery cabling and the various breakers, fuses and/or shunts, the PV array voltage should be in the 65 V range as measured at the array. This means just 13 V per PV module is required from a five-per-series-string configuration, but it means 16.25 V per PV module in a four module per-series-string configuration. This higher number may well be difficult to achieve in the Summer when relatively hot ambient conditions reduce the modules’ output voltage.

As you are aware, it’s a different story with AGM batteries. The end of Bulk/MPPT voltage for a 48 V bank is ~57.4 V, so the array voltage would need to be in the 60 V range, or just 15 V per PV module in a four-per-series-string configuration. This shouldn’t be too difficult to achieve from a relatively high Vmp module like the KC130 in mild Summer weather in Canada.

Another advantage to the five-per-series-string configuration is that for a given amount of power, the voltage is a bit higher and the current is a bit lower, which leads to reduced voltage drop and power loss in the array-to-controller wiring.

HTH,
Jim / crewzer

niel

Re: New System Design Questions

jim,
sorry i must have misread you, but we appear to be in agreement on using the 5x5 arrangement. as to the "It’s “higher in Winter, and lower in Summer”…" i do know that and i don't know what i was thinking when i put the reverse in. thanks for catching it though.