pv system connected to sub-panel

[Jest Waitn]

OP: NOMB or concern, but a 9kW system for an 8,400 - 8,500 kWh/yr. load seems maybe a bit oversized. How did you determine the array size ? Where are you located ?

#1) want FREE electricity. 90% of tier 1 grid-tied pv system = ~9kw. #2) make a little $ off of it.

[sensij]

9 kW is killing your design. Stick with 32 M250's or smaller, much more compatible with your existing electrical service. Please consider J.P.M.'s question carefully... how are you estimating the kWh output of your system?

[J.P.M.]

#1) want FREE electricity. 90% of tier 1 grid-tied pv system = ~9kw. #2) make a little $ off of it.

Unless you get the grid alternate means of supplying electricity for nothing in mat. and labor including the value of sweat equity, you will not achieve # 1.

On making a few bucks, if your POCO is like most that allow net metering, you'll probably find they pay a relatively small amount for overproduction vs. what they charge. Also, net metering is, in general, becoming less "net" than it once was.

Just a thought: Are you under the impression that a 9 kW system will only produce 9,000 kWh/yr.? If so, a suggestion: Do yourself a favor and educate yourself about how much a PV system will produce at your location. If you know, and don't care, forget I mentioned it.

Take what you want of the above. Scrap the rest.

[Jest Waitn]

Unless you get the grid alternate means of supplying electricity for nothing in mat. and labor including the value of sweat equity, you will not achieve # 1.

On making a few bucks, if your POCO is like most that allow net metering, you'll probably find they pay a relatively small amount for overproduction vs. what they charge. Also, net metering is, in general, becoming less "net" than it once was.

Just a thought: Are you under the impression that a 9 kW system will only produce 9,000 kWh/yr.? If so, a suggestion: Do yourself a favor and educate yourself about how much a PV system will produce at your location. If you know, and don't care, forget I mentioned it.

Take what you want of the above. Scrap the rest.

thanks for the input.

i don't see much difference between an 8kw & a 9kw system; total difference is about $1k. the framework for the 8kw system is only slightly less than the 9kw and the panels are dirt cheap now. bought mine for 25 cents/watt (gcl class a from sunelectric). the micro inverters are $100/ea.

but if there is a problem with over production as noted above then unless the virtual net metering is supported, i'll stick with the smaller system. i have no problem with this.

i expect to achieve ~5 hours average sun time daily * 365 = 1825 * 9kwh = 16,425kwh annually. with 2 arrays (6kw): 365*5*6 = 10,950kwh annually. so with 1 array (3kw): 365*5*3 = 5,475kwh annually. my belief is that the 3kw system is about a 60% system. it might work out that the 6kw system is too much given the distance and some of the inverters need to be disconnected from the end of the chain. time will tell.

[Jest Waitn]

Unless you get the grid alternate means of supplying electricity for nothing in mat. and labor including the value of sweat equity, you will not achieve # 1.

i don't understand what you mean by this exactly. i have shopped around and purchased materials at the lowest cost i could find. given the 30% tax credit on the system, this helps decrease the ROI time. i am doing the work myself whenever possible and when parity in the PV electrical vs. grid usage is achieved then the system will start to pay for itself. the 3kw is a 60% system; the 6kw is a 120% system. this should pay for itself in a few short years. so the 8/9kw will be ~ a 180% system, which will reduce the ROI time even more. given tax credits and system cost, i expect ROI to be ~ 6 years.

[Jest Waitn]

9 kW is killing your design. Stick with 32 M250's or smaller, much more compatible with your existing electrical service. Please consider J.P.M.'s question carefully... how are you estimating the kWh output of your system?

the gcl panels produce 290wdc and produce 250wac @ 1a through the m250 micro inverter. 12 are daisy chained, so this gives 12*250 = 3kw @ 1A/ea = 12A max output, ~13.2A peak.

[SunEagle]

i don't understand what you mean by this exactly. i have shopped around and purchased materials at the lowest cost i could find. given the 30% tax credit on the system, this helps decrease the ROI time. i am doing the work myself whenever possible and when parity in the PV electrical vs. grid usage is achieved then the system will start to pay for itself. the 3kw is a 60% system; the 6kw is a 120% system. this should pay for itself in a few short years. so the 8/9kw will be ~ a 180% system, which will reduce the ROI time even more. given tax credits and system cost, i expect ROI to be ~ 6 years.

It really isn't how quickly your ROI or payback can be that determines if you will ever get "Free" electricity. All it means is that you can generate power at a lower cost then purchasing it from your POCO. But during all the times you can't generate your own power will mean you will be purchasing the electricity so not "Free" but maybe cost avoidance depending on the Net metering contract you have.

[J.P.M.]

thanks for the input.

i don't see much difference between an 8kw & a 9kw system; total difference is about $1k. the framework for the 8kw system is only slightly less than the 9kw and the panels are dirt cheap now. bought mine for 25 cents/watt (gcl class a from sunelectric). the micro inverters are $100/ea.

but if there is a problem with over production as noted above then unless the virtual net metering is supported, i'll stick with the smaller system. i have no problem with this.

i expect to achieve ~5 hours average sun time daily * 365 = 1825 * 9kwh = 16,425kwh annually. with 2 arrays (6kw): 365*5*6 = 10,950kwh annually. so with 1 array (3kw): 365*5*3 = 5,475kwh annually. my belief is that the 3kw system is about a 60% system. it might work out that the 6kw system is too much given the distance and some of the inverters need to be disconnected from the end of the chain. time will tell.

A respectful suggestion or two before you go further:

1.) Check out if net metering is possible w/ your POCO and if so, find out what the deal is, including reimbursement for over generation.

2.) Your sizing estimates leave a bit to be desired. If you're in a sunny climate, 5 kWh/day/m^2 (which is the technically correct version of "sun hours" and avoids a lot of neophyte confusion) is probably a decent est., but it's relatively easy to get a bit harder est. If you have not done so already, check out something called PVWatts on the net. Read the help screens and use a 10 % system loss parameter rather than the 14 % default rating. Get the azimuth and tilt as close as you can measure or estimate.

If that 5 kWh/day/m^2 average daily irradiance does turn out to be a reasonable representation of reality, and unless you have well defined and certain plans for a rather large increase in usage, a well designed 9 kW system will produce a lot more electricity than you're likely to use. If your POCO is like most, the meager amount they'll pay you for overproduction on a NEM arrangement is peanuts and nowhere near what you're charged for power, making overproduction a losing proposition in most cases.

Overproduction may feel good and suit yourself. I'm only suggesting that you're looking way oversized for your stated usage based on what I see, and IMO only, as a lack of information and familiarity w/ PV on your part. Kind of like oversizing HVAC equipment, except the penalty from oversizing PV is a lot more costly than oversizing conventional HVAC equipment.

Take what you want of the above. Scrap the rest.

[sensij]

the gcl panels produce 290wdc and produce 250wac @ 1a through the m250 micro inverter. 12 are daisy chained, so this gives 12*250 = 3kw @ 1A/ea = 12A max output, ~13.2A peak.

Yes. If you stick with less than 32 A total, on two strings of 16 panels, I think your existing sub-panel, sub-panel feeder, and main panel can be left unchanged. If you increase over that value, you'll have to derate your main breaker for sure, need to work through the sub-panel calculations appropriate for your specific NEC code cycle to figure out if the feeder size and OCPD are correct.

[Jest Waitn]

Yes. If you stick with less than 32 A total, on two strings of 16 panels, I think your existing sub-panel, sub-panel feeder, and main panel can be left unchanged. If you increase over that value, you'll have to derate your main breaker for sure, need to work through the sub-panel calculations appropriate for your specific NEC code cycle to figure out if the feeder size and OCPD are correct.

thanks, this is my belief. i haven't gone through these calculations in detail (just ball-parked them) yet, but will. FWIW, the max # inverters/circuit/array is 14 per enphase. i pulled wire for 3 circuits/arrays. the plan is to build an array each year for 3 years, if warranted.

[sensij]

thanks, this is my belief. i haven't gone through these calculations in detail (just ball-parked them) yet, but will. FWIW, the max # inverters/circuit/array is 14 per enphase. i pulled wire for 3 circuits/arrays. the plan is to build an array each year for 3 years, if warranted.

Are you in north america (16 max), or australia (14 max)?

[Jest Waitn]

A respectful suggestion or two before you go further:

1.) Check out if net metering is possible w/ your POCO and if so, find out what the deal is, including reimbursement for over generation.

2.) Your sizing estimates leave a bit to be desired. If you're in a sunny climate, 5 kWh/day/m^2 (which is the technically correct version of "sun hours" and avoids a lot of neophyte confusion) is probably a decent est., but it's relatively easy to get a bit harder est. If you have not done so already, check out something called PVWatts on the net. Read the help screens and use a 10 % system loss parameter rather than the 14 % default rating. Get the azimuth and tilt as close as you can measure or estimate.

If that 5 kWh/day/m^2 average daily irradiance does turn out to be a reasonable representation of reality, and unless you have well defined and certain plans for a rather large increase in usage, a well designed 9 kW system will produce a lot more electricity than you're likely to use. If your POCO is like most, the meager amount they'll pay you for overproduction on a NEM arrangement is peanuts and nowhere near what you're charged for power, making overproduction a losing proposition in most cases.

Overproduction may feel good and suit yourself. I'm only suggesting that you're looking way oversized for your stated usage based on what I see, and IMO only, as a lack of information and familiarity w/ PV on your part. Kind of like oversizing HVAC equipment, except the penalty from oversizing PV is a lot more costly than oversizing conventional HVAC equipment.

Take what you want of the above. Scrap the rest.

before starting this post i planned to ultimately build a 9kw pv system; one will be finished this year. system performance will be evaluated and the plan will be modified, if necessary.

i didn't use pvwatts, i used another site. http://www1.solmetric.com/cgi/insola...kup/lookup.cgi

any surplus of electricity produced will be bought at an 'approved rate'. this is all the detail given. it's probably set by the PSC.

[Jest Waitn]

Are you in north america (16 max), or australia (14 max)?

well, oops. i'm in North America. anyway, the pv frame holds 12, so its 12 for now.

[J.P.M.]

before starting this post i planned to ultimately build a 9kw pv system; one will be finished this year. system performance will be evaluated and the plan will be modified, if necessary.

i didn't use pvwatts, i used another site. http://www1.solmetric.com/cgi/insola...kup/lookup.cgi

any surplus of electricity produced will be bought at an 'approved rate'. this is all the detail given. it's probably set by the PSC.

So how did you get from the solmetric chart to 5hrs. of sun/day ? What's your zip ? array azimuth and tilt ?

Just wonderin'

[Jest Waitn]

So how did you get from the solmetric chart to 5hrs. of sun/day ? What's your zip ? array azimuth and tilt ?

Just wonderin'

zip: 32025, longitude ~30 degrees, tilt ~28 degrees, azimuth unknown, believed ~195 degrees.

the 5 hours is an estimate based on observation of sun in the area where the pv array is located.