Update July 2023: Battery test DONE with FV Panels, Inverter setup and 700W load. With minimal budget.
I finally did a full setup test, I was really delayed by:
1) The solar panels, mounting them on my balconies isn't easy because of the self-made metal structure which must be light enough, at the same time solid enough to survive 140KM/h wind (~95MPH only happens a couple of times a year, but it rips roofs off). My original plan wasn't good for the fact I was using metal bars for the whole structure, too heavy; the new project has metal bars for making an "L" shape to fix on the brick wall (not full bricks) and an aluminium part to hold the panels themselves;
2) The wife. Please use your imagination here!
For the test I simply put the panels on the balcony's floor giving the best inclination I could.
Setup configuration:
- One 14S20P 40A battery, 18650 cells are the ones I selected. 2000-2399mAh, IR 30-79mOhm, all ICR cells (no Sony V3, no ICR, no unknown cells). The second battery is ready and the third one just started; I don't have a busbar for the battery yet, that may seem a small detail but I must find an effective (and final) solution with what I have in my lab; Cables 16mm2 from battery to inverter;
- Hybrid Inverter 5KW to connect FV panels, battery, AC Input (it's my lower priority input source after Solar and Battery), AC Output to the 700W load; oh, I loved the inverter; cables for AC input and output are the standard here in Italy 2.5mm2;
- 4x 280W poli FV panels, 8.64A (Imp), 32.4V (Vmp), 39.0V (Voc), connected in series of two, the only configuration compatible with the 60-115V PV input of this inverter; for the test I connected 4 out of the 6 panels; not at the best inclination but the best I could get on my balcony without mounting the metal structure for now, maybe I reached 36-39° instead of the optimal 32° for my city. Orientation, not the best either, I have south and west orientations;
- Cables for FV: I had 10mm2 and 16mm2 for the 10mt (~30ft) distance from balcony to my room. So I used one for the negative and one for the positive (sorry, guys! It's a test!). From my formulas I'd have an 0.9V loss from the panels to the inverter using a 10mm2 cable. Depending on the price I can find, I'll chose 10 or 16mm2 for this wiring;
- Connectors: MC4 and SC10-6, possibly copper ones. In some points I had to use SC6-6 connectors, not full copper ones;
- Environment: outside temperature around 34°C (93.4°F), inside 3 degrees lower;
- Cooling: I don't think it's needed at all being the system inside.
Some photos.
Battery and BMS
Series are connected with 10mm2 copper wire. SC10-8 copper connectors, bolted together. I abandoned the cell-level fuse wire, too much effort for me to spot weld. So I chose a nickel strip solution and maintenance, if necessary, will be at pack level where removing a pack shouldn't be too difficult.
I will mount a relay to automatically disconnect a battery in a temperature or voltage anomaly situation. I bought 60A and 80A relays.
BMS is a dumb Daly 20A/40A (charge/discharge).
Nickel strips that connect series can carry around 20A. Fuses for each battery are 20A.
Spot welder: the Sunkko 709AD+ did a decent job but I did have some
problems. Until now I did maybe 5000 spots and had a severe problem twice: the spot welder just wouldn't weld. I pressed the pedal and it reacted as if there was no contact between the two tips. I never changed the tips yet, they are quite used up, there only a few millimetres left. So, what I did was to clean the tips, disconnect all the wiring, connect it up again, wait a bit and it started working again. Uhm... And another problem I had, and I don't really know the cause is that I used the same settings for these first two batteries (pulse 10, amp scale setting 5.75) but on the second battery it couldn't make it, as if it lost power. I had to increase to pulse 14, amp scale setting 6 to get some decent spots.
Ah, I did find out something about nickel welding, nickel on nickel: I did cut the the nickel strips to connect the series and then spot welded them. Now, by coincidence,
I cut a couple of strips in the other direction (vertically instead of horizontally) and... the welding can be done easier at a lower power. So I suppose the nickel molecules, at a chemical level, have an orientation (they're not round or symmetrical molecules / structures). Wow.
Panels and connectors
I then tilted the panels down to get them as near as possible to 32°, weren't really that "flat"; I could visually compare them to the other 4 small 50W panels I already had mounted and I was off by quite a bit, but it's OK for the test. The connectors had a really big impact, they were screwed up tight but I wasn't getting much power out of the panels. I tightened them really super-strong and it was ok. I think I'll avoid those connectors and bolt the wires up directly.
Inverter and connections
Black wires are 16mm2, blue ones are 10mm2. Minimal safety devices for the test, two 63A DC MCBs (I'll use 20A ones for the final setup).
With a 700W load inverter reported a 15% load.
First try to connect battery wires to the inverted without the MCB... connected the positive, ok; as soon as the negative touched the connected it sparked, did a loud bang and I jumped; I said to myself "I burnt the inverter!". But no, it was only that moment when a single strand of the wire touched the connector and sparked, that's no good. Use the MCB to connect the parts? Yes.
Final thoughts and results: I'm satisfied with the job, what a pleasure. What gets me is that nobody at my house can even imagine what's behind this DIY job, LOL
. Maybe I know one person, maybe two in all, that could share the pleasure with me! Of course, a part from you!
The battery is the biggest part of the work; and the details, in general, are really time consuming.
Output from FV reached a maximum of 55% of the nominal wattage (1120W) with full sun coverage, I think I can reach the full rating for 2-3 hours a day when the panel will be mounted high up on the balcony. Battery was cold during all the tests, 700W load isn't much at all; charging and discharging was what I expected in regards of the load and the battery capacity. Inverter worked first go, I did I quick setup, I'll give an update on what I didn't like much (for e.g. granularity in voltage settings, sometimes with 1V steps others with 0.1V steps). Cables cold during all test.
Next step, I'll have a week holiday in August dedicated to get things finished.
- Arduino/ESP8266 temperature control with DS18B20 ICs, already tested them in the "one-wire" configuration and they work perfectly; for temperature problems the Arduino/ESP8266 will disconnect the battery via a relay;
- Measurements: only DMM ones for this test and what the inverter told me. I'd like to have the measurements integrated in my website where the house measurement arrive for now (temperatures, AC loads and water tank level); I always have in mind
@Wolf 's superb Graphana dashboard, If I have spare time, I doubt it, I'll look into it (I used it in the past, it's super simple, but... time consuming!);
- Inverter's Software, I have a CD with the software but haven't installed it yet.
Hopefully I'll get them mounted soon and will be happy with seeing everything nice and setup.
