Hi everyone,
This is a great list and have learned a lot by reading all the posts.

I am building a small backup PV system in part to see if I can and to gain knowledge from the experience since I enjoy doing this kind of stuff as a hobby. I am looking to the group to tell me where I am 'way off base' or maybe suggest better ways of doing what I am doing since my knowledge is all 'book' knowledge and not a lot of practical knowledge.

First of all, I am starting with building a small off-grid PV system that will be able produce around 150 amp hours of charge capacity a day (50 % of my needed battery resource). I live in North Central Massachusetts and when I looked up the Max. Sun for my area and it was an average of 3.9. Based on 3.9 hours of good sun a day I figure I need a system that would output around 40 amps (150/3.9). I realize that I won't always get 3.9 hours of sun, but that is what I used as a basis for the sizing what I thought I needed.
I am planning on getting 1 Outback MX60 Charge Controller to charge my 4 75 amp hour batteries. The real question comes to size and number of panels. My initial guess is that I should start with 4 BP Solar 140 Watt panels (560 watts max - 17.5 V @ 8 amps according to spec) wiring them as a 24 V array. Now here is where I think my calculations will probably go wrong...since the MX60 is MPPT charge controller I am guessing my optimum output on the 560 watt array to be 44.8 amps (560/12.5) and reducing that by 10% for efficency loss in the system still gives me 40.3 amps. Seems to work since I need 40 amps. I know that this system may not operate at this rate in the dead of winter and will do a little better than this in the summer, but am I at least in the ballpark or do I just need more/different panels?

Again, this is just a hobby and the whole purpose of this project is to learn about how to put together a system like this so believe me I am willing to take any advice all of you may have including 'going back to the design board' and starting over.

Thanks in advance for any advice!

A handy place where you can look up and calculate your data monthly/yearly, both by direction the panels face, their tilt, and if they are tracking or not is here:

http://rredc.nrel.gov/solar/codes_algs/PVWATTS/version1/

Was quite accurate for my location (near SF, CA).

You may also want to look at other than 12 volt system... 24 or 48 volt systems make the wiring job much smaller, and you can increase the ultimate panel capacity of your MX60 by 2x-4x...

Efficiency wise... The link I gave above accounts for panel and charge controller losses. Add in ~85% efficiency for inverter and ~80% for wet cell battery efficiency, you will get closer to your "useful" power output.

Example:

77% (from link) * 80% battery * 85% inverter = 53% overall panel rating to AC load efficiency...

The minimum useful panel to battery efficiency is that your panels should be about 3% of your battery capacity...

300AH*0.03= 9 amps or ~ 162 watts of panels (assuming little active use)

On the high side, about 10-15% is usually the recommended maximum panel size (more than that, the batteries cannot accept the current without overheating)... So, unless you have daytime loads, it is usually not worth the money to exceed that percentage of panels...
300aH*0.15=45 amps or ~810 watts (at 18 volts)...

Anyway, these are some rough rules of thumb that sort of bracket your system design and not hard and fast rules...

-Bill

Wow...thanks BB for this great information.
I guess my efficiency 'guess' was a little bit off
Based on a 53 % efficiency rating it seems like I need to go up to 6 panels (840 W) and maybe increase the size of my battery bank to flatten out the highs and lows in production and thus not have to charge so fast. I will have to run some calculations with a bigger bank but a slower charge rate. Thanks!
Obviously I have lots to learn.

Another way is to get more panel for the same money

www.sunelec.com the sun branded evergreen seconds are the best deal ( 3.34 watt ), great panels, or if its a rebatable system, the Evergreen seconds for a bit more.

So spend the $$ for that was for the 4 BP panels and get 6 Sun/Evergreens instead

Also, if you want a unit with true mppt and no fans, look into the Xantrex WX60-150, FYI I a principal engineer for that product

Thanks Solar Guppy!

I'll take a look at the Evergreen panels (they are a local company here in MA so it would be good to buy from them anyway).
One question though (which may be a stupid question), but the reason I have stayed away from the higher wattage panels (170 +) is because they are producing higher voltage and lower amps. To me it seemed like I wanted higher amps and lower volts (or a high ratio of amps to volts) so that that controller didn't need to transform it. Of course I know that is the point of MPPT controller so maybe it is truly better to have the higher volts and lower amps and let the controller do the work. Obviously it could take me to a smaller wire size which would be a good thing since I will have a decent run (50 + feet). Opinions?

I will also give the Xantrex XW60-150 a look. I like Xantrex since I have the GT inverter for my grid-tied system and it has been working flawlessly since it was put in.

Thanks for the advice!

i'll take a small stab at this one. i understand your concern and is based on efficiency as it seems wastefull to you starting higher to go lower and i'd agree with you if you weren't using an mppt controller as they are manipulating the voltage and current anyway. now do understand that yes there is a small reduction in efficiency the farther from the desired output voltage you go to, but 1 step up will not hurt the efficiency much and you will gain, as you are aware, through the lowered overall losses through the wires with the higher voltage.
now if i missed something or screwed up this answer, i'm sure sg will let me know about it.

Thanks niel!

Actually that answers my question perfectly. Since Solar Guppy's answer I have started running the numbers on the higher voltage panels to see what the cost saving look like (if any). This does open a few more options for me though since the lower voltage/high amp panels seemed to be few and far between.

Thanks again!

yw, also, don't forget that when dealing with voltages that are doubled, at say x amps, that when downconverted by the mppt controller that the current tends to jump up by the same ratio, in this case 2x amps, and you get a small percentage more in recouping some charging losses that mppt is known for doing too. i say this due to the output current limitations the controller. you probably won't reach that limit with a small system, but keep it in mind somewhere in the back of your mind.

The typical MPPT Charge Controller does have losses that limit its usefulness to minimum system size of ~400 watts of solar panels (below that, the energy needed to run the MPPT circuitry exceeds that would be gained by MPPT)...

To get a better feeling of why an MPPT is good... MPPT is basically a modern switch mode power supply that can convert from high voltage/low current to low voltage/high current very efficiently. Basically, is is almost the DC equivalent of an AC Transformer...

So, instead of "wasting" the extra voltage drop from higher voltage panels to a lower voltage battery--it simply "transforms" following the P=IV equation (neglecting losses)... So, if if voltage is needs to be dropped by 1/2, that means the current will increase by 2x...

Of course, this is only useful when the charge controller (or grid tied inverter) is passing full available energy from the panels to the load.

In the case of batteries, once you have charged them to roughly 80-90% of their capacity, the charge controller will taper off on available current and slowly bring the batteries to 100% charge then drop the current/voltage back even further and just "float" charge the batteries.

During the time that batteries are finishing up their charging--this is frequently a good time to bring on optional loads (washing machine, water pump to tank) where you can use the excess power (not needed by the batteries) to do real work for you (the charge controller will supply the extra current to your loads).

And, this is why a Grid Tied inverter is so neat--it always is transfering the maximum power available to the home's wiring--and back out to the grid (turning the meter backwards)--so that they can get the maximum power form the solar panels without regards to other limitations (like battery charging requirements).

It may be too late--but since you already have a grid tied inverter and are going to spend the money anyway for this (smaller?) off grid system... You might look into the new Xantrex XW system to replace your current GT system and reuse your panels and wiring...

Pull the GT inverter, rewire the panels to run XW charge controller (140 vdc max vs 550-600 vdc max for the GT inverter), add batteries and wire in the new Grid Tie/Off grid capable inverter and you have the best of both worlds.

The efficiency of the Grid Tie inverter and the backup of off grid power...

(I think I have this right--the XW system is a combination on-grid/off-grid capable system)

Even though you lose the cost of the GT inverter (or sell it on Ebay), it will still be less expensive than needing to purchase 840 watts of panels to run your new system (assuming it is all for the same address). And the new hybrid system will be much more capable than either system alone (you may need more batteries to use the system properly as the XW inverters are probably much larger than you planned for this small test system).

-Bill

that minimum is typical of somebody we know's responce and he thinks mostly in an mx60 mode, but there are mppt controllers that do allow for working with smaller systems and there is more than just the likes of outback. i'd like to hear guppy's general recommendations on a minimum for the xw too, but in light of tbottger's needing about 300-400w off-grid that this is the borderline area for the mx60's operations. my sb50 does not share that same minimum as the mx60 and it operates just fine with 128w of pv stc, but i would not recommend going with the sb50 unless guppy can confirm that it does not downconvert linearly.