Looking for opinions on an 'non ideal' setup

Hi Folks,

 I'm looking for a bit of a review of the system I've set up, for the sake of my own perspective, and maybe heading off any potential lurking oversights.

The core system itelf comprises of this:

- 2x Fullriver DC 400 6v AGM batteries (wired series for 12v)
- 6x 100W '12v' panels (the most I can fit on the roof)
- Bogart Engineering Trimetric & SC2030 PWM solar controller


It's all installed in cargo van I've outfitted, with the aim of having a fully-solar electric mobile home. Meaning solar power for heating, cooking, washing, charging, refrigerating - the lot.

I can hear the wincing already.

I'll admit - the config is not ideal. 12v clearly isn't best for things like a 2500W convection oven,, and with limited roof real-estate, one would think MMPT and 48v would be the more or less mandatory. However, the ambitious aim of this build was to use cheaper recycled or 2nd hand materials wherever possible. Consequently, the batteries were scored from a tax writeoff ($400 USD for both), and the solar panels were repaired factory 2nds ($300 for the lot), along with several meters of 35mm^2 copper wire, scavenged for free. So, the rest of the system was built around this (ie backwards).

The two DC400's tested fairly close to the Fullriver specs - 6.2V each after a 25% nominal Ah discharge measured over a shunt. I know - OCV is only an approximation, but being AGM the hydrometer is no use, it seems to indicate no *major* signs of sulfation, so they were a good price for a lot of reserve. Downside is there was only 2 - meaning the system had to be 12V. So, I  doubled-up on 25mm2 wire for all main runs, limited their length to under a meter, and used the chassis as a common earth.

I knew the panels, in theory, could give around 600/13=46 amps of charge with MPPT. Real-world, thats obviously never going to happen, but it still meant an MPPT controller approaching that range would cost prohibitive $$$. So instead, I did some math.

Firstly, after compensating the I-V curves for real-world factors, I found that  for my conditions  the MPPT efficiency gains would be <10% (as seems often the case with 12v systems), especially considering a 400Ah bank would likely mean long periods in absorb. Moreover, I estimated that even in winter, with proper tilt I'd still get about 5.3a per panel from a PWM controller (vs about 5.7A from an MMPT), or 32a total - still enough to cover 150% of my estimated typical daily use in the winter 3-hour solar days . Consequently, I went with the Bogart Trimetric/Sc2030 combo, figuring full control over charging parameters and accurate metering would pay off more than ~10% gained from MPPT in bulk.


The controller itself was one of the few capable of Fullriver's recommended 4-stage IUIU profile: max bulk amps up to 14.7V (@ 25C), hold that until current drops to 8A (2% of C20), then feed that current until voltage rises to 'finish absorb' of 15.2, for a quicker absorb (technically could go higher, but 15.2 is as high as my appliances will allow) . It's held  there until its recharged 110% of the last discharge, then floated.

So -  did it work?

Well, after a year of testing and use, the answer seems to be yes. While I do see close to 200A when cooking sometimes, it's typically for less than 20 mins, with no detectable wiring heat, nor has voltage drop (max .5v) ever tripped the inverter.

My charging math was pretty darn close. With good panel adjustment, even in winter conditions, I typically recover the estimated ~100Ah on a clear day. The major issue is the lengthy absorb of 400Ah batteries, and I occasionally simply run out of sun in the last 10% or so. But usually with some mild energy discipline and  good weather timing, it's topped off the next day. Generally, the Trimetric indicates nightly  dips to 80% SoC (based on Ah, not OCV), floats by mid arvo, and never goes under 60% even with multiple cloudy days. Now summer's comes round, I'm rolling in watts - typically reaching the point where the battery simply will not absorb any further by a bit after lunch.


So - for anyone dedicated to get through all that, firstly, thanks for reading. Secondly: although it's apparently working fine, I'd like to hear opinions on the setup, as an 'acknowledged compromise'.

 For my part, I'm well aware that, ideally , it should have a 24v battery array, and consequently higher volt panels and an MPPT controller of around 60amps to take advantage of them. Unfortunately,  thats would cost almost twice again the present system did, and provide power well in excess of actual needs. So in the absence of glaring problems, I'm inclined to continue as-is.


Thanks very much!