The results are not living up to the math

Hi guys,
I decided to take some of the systems around here and put them on their own array and battery bank. Ran the numbers, put the system together and sat proudly back and waited for the sun to go down. Only thing is, I am not getting the power I should be getting.
I am at a loss here and am hoping someone out there can tell me what I left out. I am running 1 led light, the Mac all-in-one, Printer, a tv, security lights (10) and a small 12v on demand water pump. According to the Kill-A-Watt and my inverter everything but the pump and security lights have a total draw of 190w. The security lights run at 6.5w each so 65w there. the pump might see 2 hours use a day at around 30w. So... The Math (yay)

everything but the pump runs 10 hours so

1900w inside
650w for the lights
60w for the pump

2610 total watt hours

taking into account the inverter loss of 15%
2610/.85 = 3070 watt hours

Since watts = amps * volts divide the watt hours by the voltage of the battery to get amp-hours of battery storage

3070 watt hours/12v=256 ah

Now we all know never to got below 50% of the battery so we need 256*2 or 512ah

We live in west texas so we have 300+ full sun days, and we use a generator and 45amp charger to on the few days the sun doesnt shine so for this project the extra day storage is not being calculated.

I have 855 watts of soar panels and 5.1 hours of full sun so 4360 watts (not counting morning and late afternoon production)

since this system is not on my mppt controller the 80% efficiency of my control must be added in so 4360*.80= 3488 watts. (I know its tight and another 285 watts is about to be added)

Anyway, I have 7 90ah 12v batteries so 630ah total storage.
Because I need 512 and I have 630 I should have no problems getting the 10 hours i need out of this system, correct?

I am at 12.2v or 50% in 7-8 hours!

Any thoughts?

Great work up of your numbers first off...
Do you have a meter on your solar charge controller?  I would want to verify production for sure.
Next  I would see if I could track down the inverter efficiency curve. If you are in its bottom 10 percent of inverting capacity efficiency falls off. Hmm... maybe verify your battery specs to make sure the amperage quoted is the 20hr rating not the 100 hr rating. Battery companies tend to muddy up the waters some when it comes to numbers.
Cheers, David

Hi David, Thanks for the quick reply. They are Duracell 27DC Marine and RV. Rated 90ah@20hrs. I am not concerned with the charging side of the problem because I can throw another panel at it to widen the gap. However when starting the night with what I know to be full batteries or at least reading 99% on the controller and 12.8 on the meter, I should have enough to run 10 hours, right? The inverter is a new PowerDrive 1500w. not the most efficient but far from the worst. The only thing I can come up with is perhaps I wired the batteries up wrong. Currently I have a single string with the sensing wire on the middle battery. Charging is positive on battery 1 and neg on battery 7 and I am pulling it out neg on 1 and positive on 7.  Perhaps add another battery and run 2 strings? I figured the meter should be in the middle of the bank in case charging was not even but the inverter and the meter read the same voltage for the most part. sometimes the meter reads .1v less (12.5 on inverter and 12.4 on meter) but only for a few minutes.

Oh and only the 1900w or 190wph is passing through the inverter, pump and security lights are 12v so run direct. while i am above the 10% inverter capacity, it not by much lol

13v is a full battery. 12.8 sounds close but it isn't when used as an indication of battery level. Remember also that for every kwh you want to draw from your lead acid battery you need to put 2kwh in. LiPoly is a bit better but not much.

Joseph Johnson wrote:Hi guys,

I am at 12.2v or 50% in 7-8 hours!

Any thoughts?

12.2V is 50% SoC wth NO LOAD and that is a generic figure.  You need to look up what the Soc vs voltage under load is for your specific brand/type of battery.  Also note that the listed value will be for a NEW battery, as they age the voltage under discharge will start falling.

Calculating SoC from voltage is at best a guess.  You really need to measure AHs instead.  Get a battery meter designed to keep track of AH in vs AH out.  Something like the Xantrex Link-10 or perhaps one of the cheap Chinese clones.  Or you could build one using an Arduino, etc. if you are into that sort of thing.

Some thoughts, unplug the printer (and anything else that draws power) except when you need it.

Do you only use electricity at night?  If you run anything during the daytime, you have to subtract that power from the output of the MPPT because it's being used BEFORE it goes into the battery.

Finally, you did not take into account the charge efficiency of the batteries.  You only get about 0.9 AH out of a battery for every 1 AH you put in, but it's worse than that because you charge at a higher voltage than you discharge, so for every watt that you put in you only get about 0.75 watts out.
And with Lead-Acid batteries you need a periodic equalization charge.  Equalization charges run at around zero % efficiency.  I.e you put a lot of extra energy in, but basically get nothing extra out (except a healthier battery that lasts longer)

P.s. those are crap batteries to use for off-grid solar.  When they die and you need to replace them in a year or two, buy some decent quality golf-cart batteries or better yet some batteries designed for use off-grid solar.  Marine batteries are not really deep discharge batteries, they are starting batteries that are designed to last a little longer than normal starting batteries when used occasionally for deep discharge applications.

A couple additional thoughts.  
When estimating hours of sunlight for off-grid, use worse case average.  Plug the specs for your location, solar array, etc. into http://pvwatts.nrel.gov/ and then look for the month with the lowest output, use that number instead of the annual average.

When I said your batteries will die in a year or two, I was being optimistic.

Hi Guys,
Thanks for taking the time to reply to this thread. The PVWatts Calculator lists us at 6.37 this month and going up from here to 8.09 for June and down to 5.79 for Oct.  The lowest is 4.11 in Dec but this system will not be used as much by then. As for wether or not we use power in the daytime, as originally stated this is a secondary array, not the primary. My primary is a 24v system with 3000 watts in the array and a 3600AH (Surrette) battery bank. We are in the process of building the cottage up by the road and are at the point where we are spending time there more at night, but not were we are ready to move freezers and such and therefore not the main array at this point. The 7 batteries are brand new (installed 4 days ago). Once we have completely moved into the cottage, this array will be moved into the livestock area where it will be used to power the 12 security lights up there and a small water pump for watering the chickens, turkeys, goats and the cow so it will only be drawing out 1000w a day. At only 20% drain on its capacity these batteries will last far longer than a year or two. Right now my issue is getting through the next 5-6 weeks until we are ready to finalize the move up to the cottage and take the main array with us. As for the SOC under a load, this is the first time I have encountered that idea. Any clue where I might find that info? Seems as though that would be a great resource to share with my friends here in the valley as many of them are finally seeing there is more to solar than overpriced Harbor Freight "complete" (junk) systems.

It's often hard to find discharge curves on marine batteries, the manufacturers don't want to commit to anything.  It's even harder for batteries from companies like Duracell, because they don't actually make the batteries, they buy them from other companies and slap their label on them, sometimes from multiple different companies at the same time.

The voltage on Lead-Acid batteries sags under load.  The higher the load, the more it sags.  Put a high enough load on the battery and it will instantly sag to 10V or less.  For example CCA (cold cranking amps) are determined by measuring how much current you can draw from a freshly charged battery for 30 seconds without dropping below 7.2 volts.

50% SoC is typically around 12.1 to 12.2 (depending on the battery) measure with no load and the battery sitting for at least 5-10 minutes unloaded.  One way to determine what the voltage under load is would be to discharge the battery at a steady current equal to 1/20 of the rated current for 10 hours and measure the voltage.  To double check, disconnect the load and wait at least 10 minutes and then measure again.  If it comes up to around 12.2V, that is a good indication that you hit 50% SoC, whatever the voltage measured just before you disconnected the load would be what it should measure with that load, it will be lower with a higher load (perhaps even significantly lower) and higher with a lower load.
For a C/20 load I would expect the difference between loaded voltage and unloaded voltage at 50% SoC would be only perhaps 0.2V or maybe less, but that is for a deep cycle battery, for a marine battery (basically a hybrid between a deep cycle and a starting battery) I'm not sure what it would be.
It would also depend if it was an AGM battery, AGM batteries tend to have a higher voltage and less sag (lower internal resistance)

Note: that Lead-Acid batteries don't reach their foll capacity until they have been cycled 20 times or so.  If the battery sits un-used for more than a couple days it will also have reduced capacity until it has been cycled a couple times.

I second Peter on this. Voltage is only good as a general reference for state of charge unless the battery has sat for at least an hour.

'12.8 is great' is my saying for battery check. Might be 13.1v or so at rest.

If you are curious, a battery monitor is a good addition. You could track state of charge in percentage, amp hours in and out and the difference between the two, at a glance. This will also give an indication of charge and discharge efficiency.

7 batteries in parallel is a lot. The connections have to be good throughout, if there is difference in resistance at any of these, that battery will suffer and the ones adjacent to it will also.

Bright and tight, check resistance between the battery terminals and the cable lugs at each connection. Also, you could revise your charger settings. I find wide difference sometimes, between suggested profiles and ones that produce a full charge indicated by specific gravity.

It is amazing what difference temperature has on voltage setpoints and capacity.

frank li wrote:

If you are curious, a battery monitor is a good addition. You could track state of charge in percentage, amp hours in and out and the difference between the two, at a glance.

Any recommendations on an inexpensive but effective one.

https://www.altestore.com/store/meters-communications-site-analysis/meters-battery-monitors/ammeters-voltmeters-battery-monitors/bogart-engineering-tm-2030-a-trimetric-monitor-p11865/

This is one option. Probably my favorite is a whizbang jr and midnite classic charge control combo, not cheap.

The Zantrex Link series may be a good choice, Bogart Pentametric.

I found this real cheap. Has the features.

http://vi.raptor.ebaydesc.com/ws/eBayISAPI.dll?ViewItemDescV4&item=170483294165&category=4678&pm=1&ds=0&t=1505993659000&ver=0

frank li wrote:https://www.altestore.com/store/meters-communications-site-analysis/meters-battery-monitors/ammeters-voltmeters-battery-monitors/bogart-engineering-tm-2030-a-trimetric-monitor-p11865/

This is one option. Probably my favorite is a whizbang jr and midnite classic charge control combo, not cheap.

I have the classic 150 on my main array. Perhaps I should give midnite a call and get the wizbang jr to go with it

I'm using the Bogart trimetric and I would recommend it.
It tracks actual amperage instead of generalizing with voltage. And it gives me a fuel guage like percentage of full to read.
Two thumbs up.

Its a good monitor!