Nickel Iron vs. Lead Acid - Off Grid battery debate

[ursus_maritimus]

Several things come to mind immediately:
1. Are either of you using any kind of magic float oil on top of the electrolyte?
2. Is the ring below, above or straddling the space between the lowest electrolyte level and the highest? And is it darkest at the top?
If so, it could result from the electrolyte clinging to the plastic as the level boils down and drying on the side.
3. Or the result of a reaction between the electrolyte and the plastic in the presence of high oxygen concentration just above the electrolyte surface? The hydrogen will rise and be dissipated, while the oxygen will just replace the nitrogen-oxygen mix which is normal air.
4. Possibly some impurity in the electrolyte or something dissolved off the plates at the initial fill is just rising to the surface of the electrolyte and being deposited on the plastic as the level goes up and down?
5. Next???

Can you safely reach into the cell through the fill hole with a non-conductive alkali-resistant probe and rub the surface of the case to see whether it is an insoluble deposit that will rub or scrape off or a change to the plastic itself. (Either one would be worrisome to me.)

In my case, this ring was in some of the cells when I took them out of the crate dry. I know they fill the cells and must do some sort of testing before emptying and shipping (there is a residual amount of fluid in each of the cells when shipped; it would be wise to assume it is caustic). Assuming the cases and cells were new, these deposits must be created very early during the operation of a cell.

The Changhong manual I had (via internet) didn't mention float oil, and it also seems unlikely they would put this in the cell for the short time needed to test them. I do have some of the float oil but didn't get it from Changhong. (I might use it in some of the cells this spring) According to some of the threads on various sites the past year it is a special grade that was made by Chevron, but no longer in production. In one of the help columns in Homepower magazine Richard Perez advised putting a layer of mineral oil USP in cells after changing the electrolyte. In some of the earlier years of that publication, they went through a period of enthusiasm for NiCads, but that seems to have vanished. They must have gained some experience dealing with float oil and carbonate formation. (In his earlier "Battery Book" R Perez did have a chapter on NiFe, but was not very enthusiastic about them at the time.) It might be worth contacting them for their experience with electrolyte chemical quality, carbonation and float oil.

ftp://sunsite.unc.edu/pub/academic/e...s/29/29p44.txt

One interesting excerpt from link above

"Testing for electrolyte carbonate level is possible via titration.
For the specific procedure see HP #15, page 23. It is more expensive
to test a
cell's carbonate concentration than it is to replace the electrolyte
in a cell. So the test procedure is usually skipped if the cell
is over ten years old. "

[Mike90250]

Mike,

If I understand your question correctly now, you would like to define 100% Power as a C30 drain to 1.0v - then you would like to know what % of that power is discharged via a C30 drain to 1.1v. Have I got that part right?

Well, yeah, that's about it, lets call the battery temperature 70 or 72F, or whatever the spec is, that the charts are published to. And then is the issue of, what is "Full" ? (setting empty to 1.0V seems safe)
Best to go by an amount of dwell time at voltage (which voltage spec), or
when absorb (fixed voltage) amps drop to some % of C.
Or is there a 3 method that's valid ?

I don't have float oil, just relying on the check ball valve in the flip caps. (useless flip caps I might add, the port in them is too narrow to add fluid, I have to unscrew the cap to add water.

Original fill was distilled water with the chemical mix, in the ratio supplied.

Not able to sample the "bath tub ring" the fill port is too long and narrow to get to the edge of the cell.

I have a titration kit, and will try it when the weather warms up. If I have to replace electrolyte after just 2 years, it will be a sad day indeed. The bio-diesel grades are not too pricey, but un-carbonated lab grade, looks to be real expensive.

[ursus_maritimus]

"Testing for electrolyte carbonate level is possible via titration.
For the specific procedure see HP #15, page 23. It is more expensive
to test a
cell's carbonate concentration than it is to replace the electrolyte
in a cell. So the test procedure is usually skipped if the cell
is over ten years old. "

Reconsideration: The NiCads he would have been playing with were probably much smaller volume than the cells people are using for serious off-grid installations (I'm not serious, just curious with mine, not nearly large enough to convert to CO2 scrubbers. ), so I doubt the test is more expensive than new electrolyte in that case?

Over the weekend I'll try to find the cells with the dark rings and see if the geometry of the smaller cells permits some sort of coiled swab to access the inside of the case. But it's been dry for over a year, and depending on what affinity this mystery deposit has for the plastic, it may be thoroughly "soaked in".

The plastic on the cells I have is not clear "transparent". It is something like "frosted glass". The plates are only visible as shadows. Sort of like the 70s Tupperware containers.

If these cells are designed primarily for "industrial" and government use, where the users would almost certainly have standard life-cycle plans where fussing over electrolyte to squeeze extra life out of them would not be a consideration, would the designers be that interested in the condition of the plastic beyond 20 years or whatever the official lifespan is? I had read somewhere the factory systems that make these were a German design (Varta) so perhaps that would be a better place to look for answers than the Chinese manufacturers?

[Robert1234]

what is "Full" ?

Defining full is the crap shoot when you are working with the Ni-Fe batteries in the field. My experience has been that once at float voltage you really don't add much (if any) more power into the cell even if you go into float with the cell only half charged. Perhaps what might be more consistent would be to refer to the number of amp-hrs remaining in the cells to a level of 1.10v once you hit the 1.0v mark at a specified draw? Even that moves around though as the cells like to "recover" back to a higher voltage when the drain is released or decreased. I fear there is no clean answer to your query.

Didn't realize you are on year 2. Do you feel as though you have lost capacity? or is this the first winter of full, "hard" use?

[inetdog]

Defining full is the crap shoot when you are working with the Ni-Fe batteries in the field. My experience has been that once at float voltage you really don't add much (if any) more power into the cell even if you go into float with the cell only half charged.

How about something really unambiguous and conservative just for testing: When you have delivered energy equal to three times the cell's nominal capacity (that should allow for more than worst case charging inefficiency) the cell is FULL. Defining full during normal cycle operation is definitely more problematic.

[Robert1234]

Pretty close to what I already am doing with regards to charge loading for lab tests. On the 150 amp-hr cells I charge at 30 amp overnight. Figuring 15 hr on avg charge, that's 450 amp-hr or right on 3x loading of a C5 charge rate. Would be interesting to compare the percent discharge curve & charge / discharge efficiency as we start dropping the charge loading too (ie short charging the cell to < 100% capacity). Haven't done many full discharges at constant drains as low as C30 since my cells are so small and I don't expect to use them that mildly, but we'll see what it looks like just for grins. Gonna take some time though as I'll only be able to drop out maybe 3 tests a week due to the drain lengths.

[inetdog]

Gonna take some time though as I'll only be able to drop out maybe 3 tests a week due to the drain lengths.

You haven't figured out how to do an "accelerated" C30 discharge, eh? And I suppose that you have to discharge at what you guess will be the C30 rate, then correct if the cell lasted longer or quit sooner and try the whole thing again from the start?

[Robert1234]

Purdy much spot on My 150 amphr is based on a 24 amp drain to 1.05 (my current inverter cutoff), followed by a 24 amp drain to 0.90 (the targeted Magnum inverter cutoff), followed by a 5 amp drain to 0.90 (just to see what else might be left). So... C30 will be an arbitrary notation for a 5 amp draw till I otherwise change my definition of the capacity of these regen cells.

I'm pulling another C5 drain from the cell today while I otherwise work. Will full charge up again Monday night and start the first full low power drain on Tuesday. May give a 4 hr short charge some Monday & drain just to see what kind of % ROI one can expect in the lower charge state of the cell.

Not really sure if this will help Mike much though as "it is what it is" for his setup. Perhaps this may suggest he can improve efficiency a bit by part charging? If so, he'll need to repeat all of this on his own setup as the results can be fairly battery specific. Guess let's just see what it says and we'll go from there.

[Mike90250]

This is my first hard use of the batteries, and till we get sun in the spring time, I'm not able to experiment much with them, because I'd have to charge with generator, and I don't want to do that yet.

[plataoplomo]

till we get sun in the spring time

May I have your impressions so far? Worth it? Not worth it?

Thank you

[Mike90250]

May I have your impressions so far? Worth it? Not worth it?

Thank you

I've only been pushing them for a month.

they take a lot to recharge, 2 sunny days (21 KWh harvest) and still not recharged after several cloudy days. Had a 13 day run of clouds, recharging with genset kept system running safely, and with less fuel than having to get a Lead/Acid bank up to 80%.

regular maintainance will include electrolyte change outs, which is going to be a several day job,
in my case, 100 gallons of new juice to mix, and 100 gallons to dispose of. (Maybe I'll make a monster batch of bio-diesel with it)

So, currently, undecided. With luck, if it takes me 5 years to decide, then it may not have been a bad choice. If I decide in less than 5 years, that means I've made a poor choice

[plataoplomo]

if it takes me 5 years to decide, then it may not have been a bad choice. If I decide in less than 5 years, that means I've made a poor choice

Thank you very much.

[Robert1234]

Mike, You've got a lot of storage capacity there. You think a 21 kWh harvest peaks out the capacity in your bank? I'd think you'd probably have "more room in the inn" so to speak (if I may use a seasonal analogy )

[Edit on reread - my bad - you plainly say they are definately not recharged]


My personal hesitation with the Ni-Fe isn't the longevity nor the electrolyte changes nor even the high investment cost. It's what is going to be my kWh out vs kWh in. Doing some out vs in ratio studies at low draw rates involving various states of charge will be interesting and perhaps beneficial. My panel locations at our cabin will not be optimal, so being able to store and retrieve as much power as possible from the array is really important to me. I may need to rethink some of my early work as perhaps I've been too hard on my little bank making my decisions at C5 draw rates?

I'm charging up the cell now and will start the full C30 discharge tomorrow am when I get back to the lab. It will be Wednesday am before I have the first data points to share - assuming my network administrator doesn't intrude on the test.

[Robert1234]

Test finished with a total drain of 296 Amp-Hr. Very different shape to this curve versus the high draw tests I have run. The majority of the power was drawn out of the battery before the volts fell to 1.10 (see data below). It's also interesting to note that the battery open circuit voltage was back up to 1.26 volts this am once the draw was halted and the battery was allowed to rest.

So based on that, at very small draw rates there is very little power left in the cell between the 1.0 and 1.1 volt interval. Will be interesting to see how this repeats as I raise the discharge rate at 5 amp intervals. But first, I think I will do some short charge - discharge efficiency measures as I work my way back up the discharge rate curve.



Draw Rate - 5 amp
AmpHr to 1.10v - 286
AmpHr to 1.00v - 295
AmpHr to 0.90v - 296


P.S. For those that might want to do storage cost comparisons, at this amount of storage my Ni-Fe bank cost me $96 per KWh. I have Trojan L16RE-B's priced at $128 per KWh (based on a 4.1 amp draw). Using my 25 amp draw data, my Ni-Fe are $190 per KWh with the Tojans at $142 at a 18.5 amp draw. Costs are VERY similar, but the Edisons beat the snot out of the LA on longevity (the "youngest" Edison's in my bank are 50 years old). But.... like I say, the efficiency question still has me leaning towards the Trojans for our NC install.

[Robert1234]

Mike,

Other data that might be useful to you since you are only partially charging from an almost dead battery set...

Short Charge Data 1
30 Amp @ 0.62 Hr = 18 Amp-Hr

Draw Rate - 5 amp
AmpHr to 1.10v - 12.35
AmpHr to 1.00v - 13.66
AmpHr to 0.90v - 14.22


~ 9% of Power Left Between 1.0v and 1.1v
~ 77% Efficient Delivering Power Used to Charge