Thoughts on my 48v Off-Grid Setup

Alaska Man wrote: »

*Reminder*

If you are going to be off-grid, you will need a generator too........ especially if you are building.

Cariboocoot wrote: »

If the charge rate is over 15% you're basically using PV to heat the batteries; charge acceptance will be much lower the higher you go (actual performance depends on the particular batteries of course). But a good charge controller like the Classic can limit the maximum current, so the 'extra' is only used on days when the sun is being uncooperative.

<10% peak rate is okay unless the concurrent loads bring the total down to where you can't recharge in enough time. This usually isn't a problem. We target 10% but don't demand it.

An occasional dip to 50% SOC (or even lower) won't kill the batteries right away. Habitual deeper discharged reduces the cycle life. On average around 25% DOD works out to the best value for the batteries; less discharge is almost identical life but without as much realized power. (Note this applies to basic lead-acid batteries and not AGM's or forklift types.)



PV Watts is really for GT systems but it does give you an indication of what insolation you may expect and properly interpreted will tell you how likely your off-grid system is to work. The thing is an off-grid system has to have at least some of the power stored in and taken from the batteries, which reduces efficiency. So the numbers may be good for a GT system but if the loads are heavy at night (no sun hours, as is often the case) then it's not so good.

I don't think you meant "69 kW hours AC". Six point nine maybe. That would indicate a lot of daylight, like 8 hours. Normally you'd plan for a worst-case scenario like this: 1040 Watts of array * hours of equivalent good sun (usually 4) * 52% over-all efficiency. So those numbers come out around 2.1 kW hours AC per day which is just about what you need. If your daylight is longer, all the better. If your system efficiency is greater (as in load use after batteries are charged but panels still producing), all the better.

On the whole it looks like your system is going to work fine.

zoneblue wrote: »

If your sure about upgrading then that works. But its not usually a good idea to run mmpt controllers under half their rated capacity because of the tare losses.

9.5% is close enough for government work mate. 0.15C for flooded is a little high. Remember these are rules of thumb, that yeild balanced systems. The lower the C rate the less pv you have the longer they charge, the deeper they cycle, the less chance fo a recharge in a weather clearing. Everything is connected. 77% is just another rule of thumb. Generally youd account 90% controller efficiency, maybe 5% BOS losses, so i guess the remaining 8% is airmass/angle of incidence/hot panels. Sure on a nice breezy day, bang on the equinox, those 8% will go away. Sure, max charge current only occurs for a briefish period prior to entering absorb. Sure, depending on your design, that might not even occur at noon. All of that can be modelled using daily weather data. But the 77% nominal charge effciency / 0.1C charge rate / 25% DOD rules takes care of most of all that for you, and still contains a bunch of fudge factors that give you some leeway.



The idea of the 25% DOD rule, is that a typical day with better than "average" insolation, and a typcial nights loads yeilds 25% DOD. That allows for a worst case single day of zero insolation, or two or three days of marginal insolation, either of which gets you to 50%. We also assume that all your loads come from the battery, even though day time loads come from solar direct, thats one of the fudge factors.



Those figures are averages. So 2.7 sun hours in the winteryest month, just means thats the average for that month. Thus there will be higher and lower days. How low depends on the climate there. If you download a daily insolation dataset, you can then start to get an idea about the spread/std dev if you want. But as i said the rules generally work out ok.

Heres my daily insolation at 40 degs latitude.
Attachment not found.

Means and std devs aside, there are always outlier events. And it should be mentioned that the 25% rule generally means you need a genset to cover the outlier events. In milder climates say within 35 degrees of the equator and going with more like 80-90% SOC you can avoid gensets altogether, at the cost of faster battery aging from grid corosion. Age kills the battery either way, so we figure you may as well use them.