Generators and autonomy

tabbycat

There seem to be many recommendations for a generator backup as well as several days of battery backup (autonomy). These two recommendations seem to be contradictory when you are told you should have a generator even with the several days of battery capacity. If you need a generator with the extra battery capacity then why not just rely on a generator for backup and live with one day (50% DOD) of battery capacity? It certainly would reduce maintainence and cost. The generator may run more but that is a good way to insure you have a reliable generator. Thanks.

vtmaps

One of my most important design goals is to minimize my investment in battery.  So I do pretty much what you suggest... but I don't go as low as 50% SOC.

If I get to 100% SOC in the afternoon, the next morning I am at 75%.  If it's a dark day (minimal recharge) I will run the generator long enough to get to 90% SOC.   That assures me I won't be below 65% SOC the next morning.  There are a few stretches every year where I have to run the generator daily for 5 or 6 days in a row. 

There are a couple of reasons why I don't go as low as you suggest (50% SOC) on a daily basis.  First, if I use half of my battery's capacity overnight, I will need to get back to 100% the next day.  Not something I want to do with generator.  Getting to 90% SOC is pretty efficient use of the generator.

Second reason is Peukert factor.  If I reduce my battery capacity to twice my overnight load, there will be more voltage sag with moderate loads.  For example, my washer draws 500 watts when agitating or spinning.  In my 24 volt system that's about 22 amps from my 370 ah battery.  That's more than the 20 hour rate. (My battery is rated to provide 18.5 amps for 20 hours).  I think that 22 amps from a smaller battery would cause problems.

--vtMaps

Dave Angelini

On large systems the extra days are insurance also if the owner/skilled operator is not there.
In systems in the south west US or similar world locations, the generator is not needed if the solar is large enough and the skilled operator understands what the meters and displays are telling them.
Some have the generator also because in a large system it is cheaper to burn fuel or they choose to rather than conserve in bad solar weather. Probably other reasons also.
The sun has quite a bit of power in it on a poor day and a larger than normal amount of solar panels can harvest quite a bit in a good location! It can be as simple as having enough clothing to put off laundry for 4 days. 
It can be more complicated by having an eastern and western array or tracking the sun.

Photowhit

I lean the other way, not wanting a generator to worry about, but I have suggested just what you are saying. It's good to know you have a reliable back up. So needing to run a generator every couple months is not a bad plan.

I have us solar for water heating and air conditioning, so during summer I have about 1 day's reserve. Lucky my heat comes with sun, so I rarely have a problem with reserve in the summer. Long periods over the winter can be a problem. We had 14-18 days with one hour of direct sun, fall before last. That get's pretty interesting. I'll use hot water off the wood stove and a sun shower and as Dave suggests in such situations, I am "over paneled" and am 3 years behind in being even more "over paneled" So I do get some charging even in the worse runs of cloudy days.

jonr

The cost tradeoffs aren't that clear to me.   Maybe $.65/kwh to run a generator at optimal output, $.35/kwh if you are using the heat?  Maybe $.10/kwh to add more panels/charger/inverter?   Both need to use batteries, but the generator is more flexible and creates less wear. (these are all guesses and highly subject to details).  To some extent, 2x the batteries means 2x the lifetime, so the cost increase isn't so much.

Maybe it does make sense to think more in terms of a solar/co-generator hybrid system and less a solar system with occasional use of a generator.   Would be interesting to see an optimization calculation.

mike95490

Since I have a long walk around the bridge to get to the generator, I time generator operations to breaks in the rain, when it's easy for me to get to it.  With NiFe batteries, they have high internal resistance, and will sag a lot when a big load kicks on.  So Wed AM, when wife needs the blow dryer, I go fire the genset, for the hair dryer, and then another 90 min to bulk up the batteries.   In heavy clouds, my 5Kw of PV, barely gives me 400w.  Each system is different, and I'll never get enough PV installed to not run the generator.  Last winter, we had over 10 days constant cloud cover & rain, I would run the generator 2x a day, for insurance if I could not get to it because of conditions. (and it's easier to hand crank the diesel when it's a bit warm)

Photowhit

jonr said:

Both need to use batteries, but the generator is more flexible and creates less wear. (these are all guesses and highly subject to details).  To some extent, 2x the batteries means 2x the lifetime, so the cost increase isn't so much.

Not sure I understand what you are saying? Are you suggesting that prompt charging would create less wear on the batteries?

If you are also suggesting that 2x the amount of batteries will last 2x the lifetime of the batteries, Basically NO... Batteries age from their point the are born. less cycling may lengthen their lifetime slightly, but 30-40% would be about the maximum gain anyone should expect.

jonr

Yes, prompt charging = less DOD and puts less wear on the battery.  See a DOD vs life graph - it's not proportional.  Not to mention that while the generator is on, there is typically little to no use of the battery - another source of reduced battery wear.   On the other hand, once you have the panels, solar is basically free and maybe worth waiting for (ie, a reason not to start the generator).   Somewhere there is an optimal balance.  

Check the graphs - what do you think happens to battery bank life if you double the AH and reduce DOD in some hypothetical situation from 80% daily to 40% daily?  The example I see says 1.85 years to 4 years (better than 2x!).  I expect that at those numbers, shelf life is not an issue.  On the other hand, some lightly loaded bank that already lasts 8 years - I wouldn't double AHs expecting to get 16+ years.  Not to mention they have to be paid for upfront, so there is a cost of money issue.

Photowhit

jonr said:

Yes, prompt charging = less DOD and puts less wear on the battery.  See any DOD vs life graph - it's not proportional.  Not to mention that while the generator is on, there is typically little to no use of the battery - another source of reduced battery wear.   On the other hand, once you have the panels, solar is basically free and maybe worth waiting for (ie, a reason not to start the generator).   Somewhere there is an optimal balance.  

Check the graphs - what do you think happens to battery bank life if you double the AH and reduce DOD in some hypothetical situation from 80% daily to 40% daily?  The example I see says 1.85 years to 4 years (better than 2x!).  I expect that at those numbers, shelf life is not an issue.  On the other hand, some lightly loaded bank that already lasts 8 years - I wouldn't double AHs expecting to get 16+ years.  Not to mention they have to be paid for upfront, so there is a cost of money issue.

Have you been designing systems for 80% DOD daily? Any system that is designed for normal operations with a DOD higher than 25% is a poor design, So taking a very bad area where a well designed system is drawn down to 50 or 60% after 2 days and you have to use a generator on the 3rd...

There are batteries made for float use, that will be used for 80% of capacity during catastrophic conditions, and batteries that can handle 80% DOD, such as forklift and Lithium. But to design a system that that is drawn down daily to 20% SOC...

Heck...  I don't want to explain, Do you live off grid? Might put your system info in your signature.