Measuring Stratification in 2V Cells

[Outtasight]

Desp, I guess you could prove most things with an electron microscope 

Justme, being a bit of a masochist I ended up writing a 14 page report but the SEM images of the separators were La Pièce de Résistance (no pun intended) 

Thanks to everyone who contributed and helped as I would probably have given up without them.

That'll learn 'em to mess with us here .  Glad to hear you got them in the end.  Nice pics too... It's interesting that the rubber doesn't mix with the polyethylene.  It's just sorta sitting on/in it like bits of gravel.

[Eleanor]

Outtasight, they are good pictures and it was very serendipitous to have access to a £1 million SEM and someone who knew what they were doing  

I found a presentation with some SEM pics which could be compared with the ones we had (I wasn't Googling "Rubber as Entertainment, honest)   :

http://www.authorstream.com/Presentation/Laurie-31503-bci2005-new-CellForce-OUTLINE-MILESTONES-CellForceA-Natural-Extension-70-Years-Experience-Rubbe-as-Entertainment-ppt-powerpoint/

The first SEM pics are of polyethylene and CellForce which is another polyethylene-rubber separator. However, from the thickness (0.33mm) and colour I decided the newer cells had Daramic HD separators. The second pic shows lead and antimony deposits on the surface of FlexSil which is a rubber separator - the rubber additive seems to be able to cause the lead and antimony which have been shed from the positive electrode to form crystalline structures and this prevents the antimony plating out on the negative electrode.

Once I knew what the separators were made from the graph you posted earlier became a major part of the argument - Daramic Industrial CL is a polyethylene separator. For good measure I threw in a few choice quotes from the Court document I posted eg "The performance of an industrial cell is often directly defined by the separator", "Misapplying the battery separators would affect the life and performance of the cell" ......

It's a relief as the Flaky 5 are starting to struggle  

[stephendv]

....so given that this is all sorted now, what charge procedures are you using for your new forklift batts?  how often are you EQing?

[Eleanor]

Ah, thought you'd never ask! There was another slight twist in the tale. The Manfacturer decided that they wouldn't supply us with new cells and they refunded the Supplier. They ordered new cells from another Manufacturer and they will be here imminently. We've ordered the electrolyte circulation system and the bank will arrive with all the tubes already installed. So, any procedures are hypothetical but lots of PV with at least one Outback controller and if necessary a weekly full charge with the Victron. I'll keep an eye on all the SGs weekly and decide monthly if an equalisation is needed. The best laid plans ... How are you getting on with yours?

[stephendv]

Electrolyte circulation: LUXURY! 

After the solar panel install the batteries are doing very well and am really pleased with the power from the panels - have not had to use the generator at all since we moved into the house over Christmas.
The morningstar mppt has a normal absorb setting which I've set to 2 hours at 2.42V and an extended absorb in case the battery voltage drops below a certain value the night before, which I've set to 2.5 hours.  So the day after an overcast day, the batts will absorb for 4.5 hours at 2.42V.
But I noticed that even with this setting they never seem to reach the full SG value of 1.29.  Even after a few days of full sun, they hover around 1.27 - 1.275, and only reach 1.29 after an EQ.

There seems to be two schools of thought for charging:
- The yanks who use Rolls and Trojan batts say that the batts should ideally be fully charged every day (but then they rarely have batts that last longer than 10 years).  And they don't have PzS cells over there.
- The victron whitepaper (http://www.victronenergy.com/upload/documents/Book-EN-EnergyUnlimited.pdf) where the author says that it's not necessary to fully charge every day, just keep the batt above 80% charge and then EQ once a month.  This reduces corrosion of the positive plate.  Incidentally, he also advices lower float voltages to also reduce corrosion.

The added complication is that my cells are very tall at about 1m, so I think an EQ more often will help sort out the stratification even though they're typically only discharged 10% every day.

So right now I'm not sure how often and how long to EQ for:
- Do a very short EQ at 2.6V, say 2 hours every week to prevent stratification and rely on the normal absorb charge to do the full charging.
- Do a longer EQ at 2.6V for 3 hours every 14 days
- Do a proper EQ at 2.6V for 4 hours every month.

And I don't know which will be best for the batts because grid corrosion at the more frequent EQ will only show up after many years of use.

[Justme]

I think your absorb & EQ times are a little short unless they have been calculated by watching the SG whilst charging (ab & eq) for those lengths of time.

[stephendv]

I think your absorb & EQ times are a little short unless they have been calculated by watching the SG whilst charging (ab & eq) for those lengths of time.

Yeeeessss, but that's the thing - as I understand it I will never be able to get the absorb and EQ times 100% correct because the charge controller just uses a fixed amount of time.  So I either have to decide to:
A) overcharge them slightly every day
B) undercharge them slightly every day, but then do an EQ more frequently to ensure that they're properly charged.

A) is good to avoid sulfation but it's bad for positive plate corrosion.
B) is OK for sulfation and should also avoid positive plate corrosion because of less frequent over charging.

... which is why I'm going for option B 

[stephendv]

In an attempt to prevent stratification I've tried a number of things.  Firstly, I did a 2.7V, 3 hour EQ about 10 days ago, and after that the cells were between 1.285-1.29.   Good.
Then I set absorb to 2.45V and 3 hours.  The day directly following the EQ, the cells were well charged, the next day too... until about 5 days after the EQ- none of them reach 1.29.  The highest being 1.285 and the lowest 1.28.  So I tried a 6 hour absorb at 2.45V, but no change, still 1.28 - 1.285. 

So the next experiment will be to do a normal absorb at 2.4V, and then an EQ and see how long it takes for the EQ to bring the cells to 1.29.  Then program the controller to do a mini-EQ every week.  Hopefully that will see all cells at 1.29 at least once a week and not stress the plates too much by being above 2.4V for too long.

[Eleanor]

It does seem to be necessary to raise the voltage to get the last bit of charge in. It will be interesting to see if it's the same with the circulation system. I know it's possible to charge at a higher current. There was a delay with our new cells being delivered - not sure why - but they will be here on Weds 
Unfortunately they want the old ones back 

[stephendv]

Just an update, I've left my batteries to absorb at 2.4V for 2.5 hours daily.  They reach 1.285 SG by end of the day (full charge is 1.29).
So today, while they were floating I did a quick EQ at 2.65V for 5 minutes.  Checked SG again, and just under 1.29 

It looks like even quick EQ's are beneficial for electrolyte mixing.
Charge controller is currently configured for a 10 minute EQ once a week, will be watching it for a while to see how it behaves (if I can get myself to stop tinkering with the damn thing long enough).