Increasing array size for panels that won't tilt

Endurance

You all recently gave me what I consider great advice to add solar panels rather than rely on a generator to give better life out of a new battery bank of about 1200 Ah. This might also be good timing since I understand NAWS has a sale coming up on the Midnite Classic 150 controller I have had my eye on. It looks like I need to make a decision on panels to pair with that Classic 150. Since this is on a boat, there are some design constraints:

1. The battery bank has to be at 12 volts to operate a lot of existing 12v equipment.

2. Because the boat moves around, the panels have to sit flat rather than face the sun.

3. Safety says I need to keep the wiring between the array and the controller pretty close to about 50 volts.

4. My boat is at Lake Powell where typical daytime highs are over 100 degrees F. during the boating season.

I have used Midnite Solar's string sizing tool for a set of sample panels: six Trina Solar 240 watt panels (also on sale at NAWS right now). Here are their specs:


Number of Cells: 60
Tolerance: +3%/-0%
Vmp: 30.4 volts
Imp: 7.89 amps
Voc: 37.2 volts
Isc: 8.37 amps

The string sizing tool says that if I use three strings of two panels, I will be maxing out a Classic 150 controller's amps but not bad enough to need a second controller. That sounds like what I want, but I wonder if I should tweak the results a little. Since I have to keep my array voltage on the lower side, there's no way I will get into a hyper VOC situation. Maybe rather than the 1,440 watts that six of the 240 watt panels would give me, I wonder if I'd get more hours of production per day with something like six 290 watt panels. I realize that I would sometimes be "wasting" some panel capacity. I also realize that in the real world, completely flat panels at high temperatures don't often give their rated output. Besides, panels are getting pretty cheap these days.

I also wonder how happy a Classic 150 will be running at or close to its maximum output in a high ambient temperature for most of the day.

I asked this question at Midnite Solar and get conflicting answers. Some say that because panel output is diminished, larger panels will be a good idea to get more usable power out of the "bad" parts of the day even if it means wasting a little panel capacity out of the "good" part of the day. Others point out that at around 25 degrees C (about 77 degrees F) the Classic 150 will start to turn itself down to avoid overheating. That would lead to further "waste" of any panels that have a capacity that the Classic 150 can't use.

What do you all think?

Cariboocoot

Re: Increasing array size for panels that won't tilt

You will have an issue with panel Voltage falling off due to heat, but not current. This won't really be a problem because the panel Voltage is much above the system Voltage.

Yes, the heat will cause the MidNite to throttle back maximum current. No way around that. Getting it up to peak output before the day gets hot is viable, however you have no way of predicting those conditions.

Since you will be using three strings and thus requiring circuit protection on all, my plan would be to wire six of these SolarWorld mono 250 Watts http://www.solar-electric.com/solarworld-sunmodule-sw250-monocrystalline-solar-panel.html all in parallel. The lower input Voltage will reduce conversion loss and the mono panels will take up less space for the power. Then you'd have a 1500 Watt array capable of 94 Amps maximum (7.8%).

Since this is only peak potential current and not what will always be seen all the time, heat dissipation should not be too much of a problem. And you will gain charging in the less-than-perfect sun times without going too oversized on the array and budget.

Endurance

Re: Increasing array size for panels that won't tilt

Thanks. That makes sense. As Murphy's law would have it. NAWS has only one of the SolarWorld mono 250 Watt panels in stock. It looks like I could jump up to 265 watt panels for not that much more money, though.

You have me thinking about whether to run three strings of two vs. six strings of one. I know I'll need bigger wire from the array to the controller at lower voltage. Will I give up much in terms of the time that all six panels in parallel exceed 14 or so volts in the morning to start charging vs. the time it takes three strings of two panels to bump above 14 or so volts?

Cariboocoot

Re: Increasing array size for panels that won't tilt

Endurance wrote: »

Thanks. That makes sense. As Murphy's law would have it. NAWS has only one of the SolarWorld mono 250 Watt panels in stock. It looks like I could jump up to 265 watt panels for not that much more money, though.

You have me thinking about whether to run three strings of two vs. six strings of one. I know I'll need bigger wire from the array to the controller at lower voltage. Will I give up much in terms of the time that all six panels in parallel exceed 14 or so volts in the morning to start charging vs. the time it takes three strings of two panels to bump above 14 or so volts?

The panels you are looking at are Vmp 30-ish; they will be above charging Voltage for a 12 Volt system practically all the time. Certainly at any time they are producing any current. Sufficient Voltage isn't an issue.

The current could be, as it will near 50 Amps at full output. As such that wire from the combiner to the controller should be at least 8 AWG. However, depending on the exact layout of panels/combiner/controller the size for current handling and minimizing Voltage drop will vary. It may be beneficial to have the combiner nearer the controller. This is one of those plot-the-component puzzles. Especially when checking the physical size of the panels to see where they'd fit. That is probably paramount on a boat.

ChrisOlson

Re: Increasing array size for panels that won't tilt

Endurance wrote: »

I asked this question at Midnite Solar and get conflicting answers. Some say that because panel output is diminished, larger panels will be a good idea to get more usable power out of the "bad" parts of the day even if it means wasting a little panel capacity out of the "good" part of the day. Others point out that at around 25 degrees C (about 77 degrees F) the Classic 150 will start to turn itself down to avoid overheating. That would lead to further "waste" of any panels that have a capacity that the Classic 150 can't use.

We put in 2.25 more kW solar capacity in the last month (nine 250 watt 30.8 Vmp panels) with six facing east on the 22.5 degree roof slope of our house, 3 facing west on the other 22.5 degree slope and the existing 3.75 kW (15 of the same panels) facing south. All on one Classic 150. And it works fine. I also see I have to update my sig after we put in the latest panels.

But anyway, due to bad weather we couldn't get all the panels on the roof slopes installed at once so I had 1.5 kW of it with three east and three west, and I put those on a XW-MPPT60 controller. But I had some problems with the XW-MPPT60 getting into an argument with the Classic over who's boss when it comes time to end absorb. So I put the 1.5 kW on a Classic Lite, which networks with the other Classics using Follow-Me charge stage coordination. That all works fine. But this last weekend when we were able to get the rails and last three panels on the east-facing roof slope I decided to try them all on the single Classic 150 to see what happens.

On the cloudy day it works fine because the panels work off "Cloud Shine" and they only run at maybe 50% rated output anyway. So we see around 50-55 amps. On the nice sunny day it comes up to 80 amps by 9:30 in the morning and holds 80-85 amps until at least 2:30 in the afternoon. But that's only IF we have enough loads to keep it at that output. The bank is done absorbing by that time so it starts to "throttle back" shortly after lunch time (unless my wife finds one of her high-draw appliances she can turn on to keep it at 80 amps).

We arranged our panels that way because we do not need the big "peak" at solar noon - that really accomplishes nothing for an off-grid system. I wanted a flatter output curve all day, with the biggest boost in the morning when we need the maximum amps the most to get the bank up to absorb voltage. By experimenting with it, I've found that two controllers are not needed with the way we have our panels aimed in the different directions because all the panels don't reach full output at the same time like they would if they were all facing south.

So, based on my experiment with doing this, which is similar to what you propose, I see no problem with it. It works perfectly fine here. This is a photo of the first six panels and how they're arranged - I got chased off the roof by snow before I got the other three east-facing panels on the rails

Attachment not found.

Attachment not found.

This is called "virtual tracking" - putting over-capacity on one controller and facing "banks" of panels in different directions to try to achieve a flatter output curve, and adjusting the size of the "banks" to achieve the highest outputs when you need it most. It was an experiment for me - and I've found that it works perfectly. It would not be recommended by any certified solar installers because they'd come up with all sorts of reasons why you can't do it, or why it won't work. And for a grid-tie system where you can use that solar noon "peak" on sellback to the utility it would not be an ideal setup. But with the price of solar power these days, and the characteristics of off-grid battery charging where a bank simply can't use the full capacity of the arrays anyway once it gets to the end of absorb and into float, it works fine with no moving parts (trackers) and you don't need more than one controller. It's like so what if one "bank" is off optimum Vmp by a few tenths of a volt?

So that's my 2 cents on it, based on experience actually doing it.

Edit:
One other note on our setup - the south facing 3.75 kW array is wired for 92.4 Vmp - three in series and five strings. When I put the first 6 panels on the east/west banks I had them wired two in series, three strings, with an east facing panel in series with a west. That did NOT work. The XW-MPPT-60 struggled with that because two panels in series not receiving equal solar insolation (any time other than solar noon) is not ideal. I rewired those first six panels for 92.4 Vmp - three in series east and three in series west. And then it worked fine. Adding the three additional east-facing panels this last weekend only improved the morning output and gets the controller up to its full rated amps sooner in the morning.

There is some amp limiting going on because we still get a solar noon "peak" and I see the controller unloading the panels and running them at 98-102 volts or so at solar noon. And at that time I could get more power from the arrays if I had two controllers on them. But until I see that I actually need that solar noon peak, I'm not worried about it.

The other thing is that our system is 48V so the Classic can't quite do the 94 amps peak that you'll get on 24V. But I still saw it at 86 amps a couple days ago. When it heats up it de-rates itself to around 80 amps or so. If the de-rate factor because of controller temp becomes an issue in hot weather then I'll put the Classic Lite on the east/west array banks to reduce the load on the 150.
--
Chris