Thoughts on My First Solar Design

[buccscott]

Let me preface this by stating I am a Radio Engineer and am generally very good at network design, specifications, and requirements. However, trying to design my first solar system has me a bit out of sorts. This is most likely due to the fact that it is my money I am investing and power generation/storage is a bit out of my wheel house. That being said, I thought I would throw my draft design out for review by the more knowledgeable members of this forum.

I have watt metered my daily usage and it is pretty minimal. I will be running a fridge and an attic fan or two daily (live in FL). I will also be charging some 18v tool batteries, and a small battery charger/conditioner, etc. from time to time. I also want to have the capacity to run a portable A/C unit and a few other small items in the event of a power outage. I don't want to have to rely on a gas powered generator when the next hurricane hits.

Here is what I have come up with:

4 x 300 Watt 24v ~9.5A Renogy Panels (series/Parallel) ~ 48v ~ 19A home run to controller

1 x Midnite Solar Classic 150 Charge Controller (to allow for a bit of expansion if I get addicted to solar)

8 x 215AH 6v Batteries Connected in Series (48v @ 215AH = 10320 Watt Hours) This gives me ~ 5k per charge @ 50% Discharge to work with.

1 x 1500 Watt 48v Inverter (My AC pulls 1000 - 1100 watts on high leaving me a bit of breathing room to charge phones in an emergency, run a few LED bulbs, etc.)


My biggest questions are:

Do I need to fuse the leads on a 4 panel (2x2) Series/Parallel configuration? I plan to insert a DC breaker before the controller to isolate the panels for maintenance.

I have a basic understanding of the MPPT controller's function (fuzzy on the algorithm at this point) but want to ensure the output will provide sufficient current to charge the batteries at a C/10 rate. Do I have enough input to provide this? I think the output will be 25A (1200 Watts / 48v). The C/10 rate should be 21.5A ( 215AH / 10).

Will I have overhead to allow for battery equalization, or is this a function charging at the same current for a longer period of time?

Any Input regarding the rough design or even recommendations for specific RELIABLE parts (inverters, breakers, fuses, etc.) would be greatly appreciated.

Every time I tweak the design to get what I want it seems to cost more $$$ Man!

[Logan005]

very nice set up, 48 volts is really the best way to go for a battery power set up. You also need terminal fuses on your battery bank. I use BlueSea terminal blocks with buss fuses.

[buccscott]

Thanks Logan. I had the fuses on the output of the CC (to batts) and the output of the batts (to inverter) planned. I will check out the Blue Sea products as I didn't identify a manufacturer or part numbers yet. Thanks for the tip. Like I said, the deeper I get into this, the deeper I have to reach into my wallet. It's like owning a boat.

Where I'm really confused is the protection on the panel side. I planned to breaker it according to wire size near the input to the charge controller. My current understanding is that if all the panels are serial, you don't need to fuse it and can use a simple switch to isolate when necessary. Since I have planned for 2 serial strings (2 panels each) into a parallel feed, I have no clue as to where to fuse before the charge controller. Should I fuse at the top near the panels, individually by panel, by string (before paralleling), or where I originally planned on the single feed into the charge controller.

Any thoughts?

[sensij]

You don't needs fuses on the panels until you have at least three strings.

1200 W of panels are never really going to produce that much for any meaningful period of time, since the conditions at which they are rated are not real-world. Depending on how the panels are oriented, you might peak out on a clear day at around 1000 W or so, or 21 A, right at the C/10 current. If your batteries are FLA, you can watch the specific gravity with a hydrometer and see how close you are getting to fully charged each day.

In general, designing for 50% daily discharge is a bad idea. The battery will not live as long discharging more than that, and on cloudy days, since you won't be able to get back to fully charged, you will have less capacity for discharge the next day. 25%-30% is a better design guideline, giving more "autonomy" before you need to fire up the generator to get the battery charged back up again.

The 1200 W of panels are probably not enough to recover 5 kWh daily, definitely not in winter and maybe on only the best days in late spring. One problem is that you need to put more into the battery than you take out... 5000 Wh out might take 6667 Wh in to replenish. You can use PVWatts to get some idea of the expected insolation on your array (given your location and array orientation), but the default loss factor there is for grid-tied systems and doesn't account for battery efficiency, which can be around 75%.

[buccscott]

Thanks Sensi. So I can just use a small disconnect or DC breaker to isolate the input to the CC? No need for fusing (assuming properly sized conductor)?

My real world daily consumption will only be around 2.7 - 3k putting me at about 25-30% discharge. Much of that will be during peak hours (attic fans on timers from 9am-4pm) so if I'm topped off the panels should bear the brunt of the load (correct?). My 50% value is the max I would utilize and only occasionally If I were working on my car in the garage with an air compressor, etc. I just meant I would feel comfortable going to that level (not below) at times. Now, when a hurricane hits, or the power is out during an emergency... I will squeeze every bit of juice I can out of it I didn't really go into this level of detail in my post, but I do appreciate your comments.

I think I will over design (typical practice in my field) my conductors, etc... so I can possibly add an additional two panels on later (if/when necessary). The I will have three strings. Where would I fuse it at that point?

[sensij]

Thanks Sensi. So I can just use a small disconnect or DC breaker to isolate the input to the CC? No need for fusing (assuming properly sized conductor)?

Correct.

I think I will over design (typical practice in my field) my conductors, etc... so I can possibly add an additional two panels on later (if/when necessary). The I will have three strings. Where would I fuse it at that point?

You would fuse each of the three strings at the point they are combined. Something like this is typical:

http://www.midnitesolar.com/productPhoto.php?product_ID=78&productCatName=Combiners&productCat_ID=9&sortOrder=4&act=p

[buccscott]

Man, I'm glad I found this forum! I will plan my wire routing so I can cut a combiner into the system down the road when I expand my panel count. Thanks again!