DOD vs number of cycles for a car/starter battery?

[mab]

I've been thinking of replacing my deepcycle AGM with a car battery: I know it's practically heresy, but bear with me 

my small offgrid system has a small 24v 100Ah AGM battery which I've been using for several years even before I got the hydro up & running. Now I have hydro continuously feeding ~300w into the battery (and I run my larger sustained loads when the solar is working), I seldom take more than a % or two from the battery - it's mostly there to provide the surge for the fridge & freezer.

It occurred to me that I could run large short duration loads (say a toaster (1kw)or kettle (2.2kw)) for a couple of mins without deeply discharging a 100Ah battery, but the amps drawn (90-100A for 2.2kw) would be a bit hard on the 100Ah deep cycle AGM.

So I'm thinking it would be much better to use a regular car battery which can provide 100's of amps for a short duration, as it will be recharging as soon as the short duration load is removed (much as it is in its 'natural habitat').

I've done some experiments with a couple of old batteries that were lying around and they seem happy enough, but before I run the system in anger (or actually spend money on new batteries) I'd like to have an idea as to what sort of regular DOD would be acceptable for car batteries?

I've been looking, but battery makers don't seem to provide dod vs cycles charts for car batteries (no surprise there). There does seem to be a common reference to 'not more than 50% dod' but that's a bit vague, and I wouldn't expect many cycles to that dod.

In it's normal application a car battery manages probably several 1000 'cycles' of starting the car but I'm struggling to put a figure on what DOD an average engine start represents (maybe <1% for a warm engine in the summer to 3 or 4% for a cold start midwinter?).

So to the question: Does anyone have any references to DOD vs cycles for a car battery? or suggestions as to what DOD would be sensible maximum to keep to?

[joestue]

Car batteries should last a lot longer motionless on the floor, rather than having their positive plates broken off by vibration.
Most of the car battery failures i have witnessed are the positive plates breaking off the positive battery post. --due to vibration coupled into the post, also by the fact that, its different than the other five cells.

Have you added water to your agm batteries yet? do it slowly for sure but when you add enough water to fully saturate the fiberglass they start acting like flooded cells.

perhaps car batteries are double the normal rule of 30-40 amps short circuit amps times amp hours.
take a 500 amp car battery.. that implies 2000 amps short circuit amps which means it would be equivalent to a 60 amp AGM battery (all most all of which can push 30-40 times their amp hour rate, in short circuit amps)
a cheap 500 amp car battery is probably 60$ and only 30 amp hours.

I don't think they are worth the expense, perhaps they would be if you can find a flooded agm car battery, to which you can add water. the fiberglass would hold the plates together longer than the usual batteries which have little support for the plates.

with a higher short circuit amps per amp hour, you will find that the battery should be more efficient, but the plates being thinner with less active material will be stressed proportionally more. But if you keep discharge rates constant per amp hour, the battery life should be the same.

[mab]

Have you added water to your agm batteries yet? do it slowly for sure but when you add enough water to fully saturate the fiberglass they start acting like flooded cells.

I've not tried adding water to the AGM - i'm wary of killing them off. When you say they start to act like flooded cells, I'm not sure in what way do you mean like flooded?

From memory these AGMs are rated 200A continuous drain or 500A short duration and (though don't quote me) 4.2mohms. (Yuasa endurance enl100ft).

As for wether or not buying new car batteries would be worth it - well I'm not sure yet either which is why I'm trying to figure out how well they'll last first - may just continue with the oldies for now and/or try to source some better oldies for free. I've seen 95Ah 850ccA for about £50 each IIRC, and I did see 200Ah 1400 ccA for £60 but collection only from ~200miles away - although I can't find that one now (both 12v).

it's very interesting what you say about failures relating to +ve plate breaking - and vibration - As I'm unlikely to go much over 100A and they're stationary, I wonder if my batteries will find 100A for a minute or two less stressful than starting a car (surely more like 300A+, albeit for a shorter duration.

[Flux]

My suspicion with batteries is that you get what you pay for.

AGM have some useful properties for certain applications with lack of maintenance and being capable of mounting on their side and are obviously very useful for back up systems and UPS.

For most applications flooded cells work as well and the cost per life is similar, AGM last longer but cost a lot more.

For leisure batteries and car batteries I find that what you pay for them is the main factor. A good car battery will survive quite well but will be expensive, it won't compare with a good leisure battery but may do as well as a cheap one. I have seen Caterpillar tractor batteries stand up well to RE use but they are expensive,

For many years before leisure batteries were common, I used car batteries to run PA systems, probably taken to over 50% discharge and with a life of about 4 years, the number of cycles were small and age was the decider of life.

Nearly all car batteries and most leisure batteries I have had have died from failure of the positive bars, I don't think vibration is the main factor and you will get it with stationary use. The number of shorted cell failures have been relatively small and usually comes about from leaving discharged for too long.

I am not sure that leisure batteries are much different from car batteries, the plates are still quite thin but again first cost is a big factor.

Trojans and semi traction batteries are better and for most RE are a good compromise between cost and performance. True traction batteries are the ones that really last but the first cost is high and you needs lots of input to keep them properly charged.

Just to act as a short term buffer on your system a really good car battery could work out fine but cost wise I suspect car batteries, leisure batteries and AGM will have similar costs per life although the AGM may last nearly twice as long.

I have found some makes better than others but all have good and bad, you will get the odd rogue that fails early. In reality there are very few manufacturers and most of the budget stuff is likely to be not so good stuff from the big names.

Flux

[mab]

Hi Flux,

well I must acknowledge that quality of cheap batteries is suspect.

another good thing about AGMs (vs. golfcart batteries) is that they have much lower self-discharge (calcium/tin alloy for AGM rather than the antimony used in flooded traction batteries), which when I had only a small solar array made a big difference c.f. the old traction batteries I had before.
 
In my search for datasheets on batteries I've noticed that some of the 'lesure' batteries are AGM and possibly better than general opinion of leisure batteries would suggest - but then they tend to have an AGM ish price tag too.

My thinking for my OP though is if I'm using a battery as a shallow cycle buffer then a car battery is going to be cheaper than the equivalent quality AGM/deepcycle - maybe? Also, I was thinking of keeping the AGM's out of the loop and in standby (for when the hydro goes down), rather than ageing them prematurely by asking 0.5 - 1 C discharge rates from them even if it is within their spec.

I'm kind of muddling along with my existing AGM's and experimenting with some old car batteries (which don't have any better purpose ATM), but sooner or later I'm going to need some new batteries and I'll have to decide which way to go....

[joestue]

AGMs are rated 200A continuous drain or 500A short duration and (though don't quote me) 4.2mohms. (Yuasa endurance enl100ft).

4.2mOhms works out to 3000 short circuit amps, or 30 amps per amp hour. so that's pretty typical.

yes while the positive plates do break off at the positive bars, my experience has been sudden battery failure comes from the positive battery post, the plates breaking off at that bar, not the others.. but you can find batteries where all the positive plates are in the same condition.
Perhaps keeping the water level above the bar is the proper thing to do, sometimes people fill them up to the top of the plate stack.

Anyhow, AGM cells aren't full, they are 90% saturated, giving the gasses produced on overcharge sufficient space to migrate over to the other plate and react with the active material to absorb a constant overcharge. --this also increases the resistance of the electrolyte.
When you fill an agm battery full the bubbles have no where to go and eventually float to the surface. if you overcharge the battery fast enough the gasses will push the electrolyte out of the battery.

if you can find antimony traction cells, they don't corrode nearly as quickly as the calcium cells.. which don't have any tin in them anymore, and tin didn't do anything measurable.. allegedly.

[mab]

Hi Js,

Ah OK, though I'm still a little ware about messing with the AGMs.

I doubt I'll find any traction cells anytime soon - the old ones I had were someone elses cast-offs and in rather a poor state - in the end I weighed them in as they took too much charging for my tiny solar array at that time and the best one only had 25Ah of useable capacity anyway (110Ah nominal).

[joestue]

i guess i'm annoyed at the mechanical layout of the positive grid plates. i've never found a structurally efficient dendrite type tree structure. i've only found diamond grid structures in car batteries, and square grid structure for AGM cells:
http://thefraserdomain.typepad.com/photos/uncategorized/power_currentcollectorcomparison.jpg

most of the car battery plates i've inspected look like this:
http://patentimages.storage.googleapis.com/EP0902491B1/imgf0002.png
-its actually expanded metal and starts out as a much smaller plate.

i would expect a grid that looks like this to last twice as long
http://www.sae.org/dlymagazineimages/web/516/5322_5022.JPG

[mab]

Hmm, yes, I have seen a battery brochure showing grids like that third link - don't remember where now and I suspect it was for a more pricey battery (the cheap batteries don't offer any documentation).

Having looked at the prices I fear this idea may be dead in the water - If I have to buy expensive car batteries then I'm paying much the same price as for some good quality 2nd hand 'standby use' AGMs (sort of what you both where saying in your 1st posts - oh well, I got there in the end  ).  Another set of AGMs will beef up my existing AGM bank's 'CCA' (leaving aside arguments about new & old batteries together) and still allow deep cycling when needed too. From what I've seen new golf cart batteries (Trojan T105 IIRC) are a similar Ah/£ too, but wouldn't mix well with the AGMs I don't think.

It may be more scientific to actually buy some cheap car batteries and try them and see if they do last in this application, but I'm reluctant to spend the money to prove the theory unless I can be sure they'll last TBH.

I think I'll still keep my old car batteries paralleled for now as they're sharing the heavier loads with the AGM's (and take >50% of the load at higher discharge rates) and as long as the hydro's coming in they should be happy - and possibly add some more if I can get them for free...

[mab]

just resurrecting this post in the hope of asking Joestue another AGM watering question (if you read this):-

I've acquired a bunch of rather tired AGM's which I was hoping to experiment with adding water (and maybe improving).

I was wondering how you gauge how much water to add to a cell (160Ah @ 10Hrs in this case)?

Also, on the rolls/surrette website I read about a 'recovery charge' of 2.6-2.7Vpc to desulphate the battery; If I add water to an AGM is there any reason not to try this on an AGM cell?

I also ran across this paper about improving used AGMs when digging around:-

http://www.battcon.com/PapersFinal2002/DeMarPaper2002.pdf

Without asking you to read through the article - just if you happen to understand the terms:- it talks about adding water & 'catalyst' with reference to 'cat-vents' and 'maximisers'; Cat-vent is self explanatory - a vent cap with H2/O2 recombinant catalyst in it, but wasn't able to figure out Maximiser, or if the reference to catalyst might mean adding something to the actual electrolyte or just fitting the Cat-vents (i'm hoping the latter).

I don't propose to fork out for cat vents if I can achieve the same results by just topping up regularly. 

m

Having read your replys to Greenkarson's post I'm guessing maybe adding 160-190ml/cell (don't have any deionised water yet  so it'll be a day or several before I can try).

[mab]

Actually, I'm now thinking the maximisers are similar to the cat-vents and the reference to catalyst is to the vents.

which just leaves the questions as to how to judge how much water to add? and if they will take a recovery charge?

[joestue]

Well don't use tap water but you can distill your own if you really want to.

With a fully charged battery, electrolysis consumes water at a rate of 0.336 cm per ampere-hour overcharge.

so that paper you linked is primarily interested in backup batteries which are generally discharged in less than 3 hours. during these conditions, the impedance of the battery is of primary importance. low water levels = high impedance. 

i would add water until the fiberglass is saturated. you may have to drill a hole in the case to see inside.

[mab]

cheers JS, I just didn't want to start adding water without an idea of when to stop.

I've carefully removed a vent cap from the 1st cell I'm going to try, which gives me a 1/4" hole I can just about see through. I suspect this cell is sulphated though which is why I was going to try a recovery charge.

I wasn't planning on distilling my own water TBH, though it's probably fairly doable. I did wonder if there's a difference between distilled and deionised?

thx

m