Phi 3.4 48v 2.75 years old running at 82.5% capacity

KeithBriggs
Forum Virtuoso
Posts: 76
Joined: Thu Apr 12, 2018 7:56 pm
My RE system: 16k 2 8k Radian 4 FM80 48 300w panels, 6 3.4 simpliphi Grid-tied, HBX mode connect 50.4 disconnect 51.2.

Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by KeithBriggs »

Well, that's what appears to be acting like.

Got my 6 3.4 for my 8kw radian and 9000 watts of panels. upgraded to two 8ks and added 18 more 300 watt panels for a total of 14400. Toned down the FM80's a bit but the panels are on 3 different pitches grab the sun at different times and not all at once. Also have intermatic timers on all large loads so hottub will never run its filter cycle when the hot water heater is on. After installing a curcuit and timer for the LEAF, found it was easier on those batteries too to just charge at L1 (1500 watts 120v). Added a 120v receptical off of the 240v timer. Glad I ran a neutral to the timer.

Purchased PV system in 12/07, received it in January and got 8kw system up and running with the first 12 panels (4 strings) in the middle of April 08.

I posted this on another thread and was asked to start a new topic.

Houston, we do have a problem.

Twice in the last 2 weeks, I've connected to the grid. So this time I simply clicked on the battery button on the dashboard and tallied up the net battery starting at 5PM until I connected to the grid.

Last night it worked out to 13.46 kwh before connecting. On 1/2/21 it was 13.46 kwh. I have 6 3.4s that used to give me (3.4 x 0.8 x 6) 16.32 kwh so I have lost about 17.5% capacity in less than 3 years or about 6% per year.

Just before dawn when the FM80s had just started to try to wake up and while on the grid. I flipped off all but one breaker on the batteries. 2, 3, 4, 5 and 6 off, read 1, 2 on, 1 off, read 2, 3 on, 2 off, read 3, etc. Every one of them was 50.5 per the mate. I would have preferred one of them be off/bad.

Put another way, my 3.4's are now are acting like 2.8's (3.4 x (1-0.175))

I don't sell. I run HBX mode only connecting the grid after draining my batteries to 50.4v which was rare but becoming more frequent. I reconnect at 50.8v.

I did call them today and got VM, they called back about 2 hours later. From what I can remember, they are asking me roughly to go back to v1 app note. charge controller(s) absorb 56.4v and 0.1h and inverter(s) absorb 56.0v and 0.1h. He said it was their latest app note. I asked how long before I would see results and he said to give it two weeks.

A little more background. The application note back then was followed and it was quite a bit different than the current one. I would rocket up to 100% SOC and never absorbed. I believe that this was hard on the batteries but they have consistently said just about anything goes that these things are pretty resilient. New application note was followed about a year ago. Now it seems going back to the original. In two weeks I will find a way to drain the batts and see what I see.

Also, they poopooed optics and said it could be 10% off. I'm not really buying that. I'm hopeful that this change puts more capacity back into the batts and I get more out of them! I'll produce more tests and give updates.
Last edited by KeithBriggs on Fri Jan 15, 2021 2:05 pm, edited 1 time in total.
sbrownian
Forum Czar
Posts: 538
Joined: Mon Nov 30, 2020 9:23 am
My RE system: GS8048 FM80 FNDC Mate3
4kw solar
7kw Kubota propane gen
Had 1200 Ahr FLA Locomotive batts
Just replaced with
2 x 300 Ah ReLion Lithium packs
MUCH more efficient!!
Location: Central Washington State

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by sbrownian »

I've hit up the thrift stores for used 120v heaters in the past. Usually around $5 each for the 1200 to 1500 watt types.

A quick and dirty way to burn off some power!
Information, you get not; if incorrect question, you ask..
raysun
Forum Emperor
Posts: 4581
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower Two: (2) FXR3048A, (2) FM80, MATE3s, FlexNetDC
SimpliPhi 48-3.8 (6 @ 48v)
Outback IBR3 battery enclosure
REC Alpha 440W panels - 2 arrays: each of 4 strings of 2 in series
Honda EU7000is gas fuel generator

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by raysun »

" From what I can remember, they are asking me roughly to go back to v1 app note. charge controller(s) absorb 56.4v and 0.1h and inverter(s) absorb 56.0v and 0.1h. He said it was their latest app note. "

Guess I owe you an apology for misspeaking. The 56.4 v.s. 54.4 question raises it's head here as well! I may have been indulging in some 3.8 v.s. 3.4 snobbery, and beg your forgiveness. 😉

I'm going to second Phi's response, if only that it confirms my bias about charging the LiFePO4 cells.

Lithium is somewhat unlike lead acid in that there is not the extensive chemical reaction to reverse. Its not too far wrong to think charging is mainly shuttling lithium ions from cathode to anode.

Absorb phase in a lead acid battery serves a very explicit purpose. As the battery charges, the rate of reaction needs to be moderated in order that the rising internal battery impedance doesn't result in thermal damage.

In lithium, the notion of "Absorb" is nearly meaningless. Internal impedence shows no marked increase until there is essentially nowhere to shove those little ions.

When using 56.4V as the charge termination target, "Absorb" is completely meaningless. Absorb is only mentioned in the charging specs because its the only place the charge termination voltage can be specified. The "Absorb" time of 0.1H is the minimum that can be specified as well. "Absorb" time of 0.0H can be entered, but it causes Outback controllers to skip Bulk and Absorb, and proceed directly to Float.

56.4V "Absorb" is essentially "Bulk until Full". A constant current (Bulk) charge until the cell voltage rises to 100% SoC is the basis for many Lithium-specific battery chargers. It is the method to get the Outback chargers behaving in the same manner. As long as the charging current isn't exceeding the battery's rated maximum rate, it doesn't appear to stress the cells unduly. LiFePO4 cells can be charged to a higher voltage and at a higher rate, but there's no real benefit in doing so. In fact, there's detrimental effects that are easily avoided by avoiding the maximums.

The alternate charging method, constant current to a certain cell voltage, then apply a constant voltage saturation (Absorb) charge is employed as well, of course. However, it appears to cause issues in some cases, perhaps declining capacity is one manifestation.

SimpliPhi initially specified 1H Absorb Time at 54.4V. That turned out to be much too short for the 3.8 battery, and may have been as well for the 3.4. My experience charging the 3.8s showed the charge phase reached end current at 1:29, so it makes sense they amended Absorb Time to 2H in the later specs.

My theory on the two week rehab period has to due with intercalation, the chemical process of inserting the lithium ions into the anode lattice. I have no way of proving it, but presume the voltage rise between 54.4 and 56.4 does a better job shoving those last ions deeper into the structure. Having the ions aggregate more toward the surface would seem to reduce capacity. The increased "charge pressure" may help reverse the reduction.

Let us know how it works out!
Mike Curran
Forum Emperor
Posts: 2275
Joined: Sat Aug 28, 2004 6:28 pm
My RE system: Outback - Garage roof:
- 8 (2x4) Evergreen 180's into one FM80 (2007/2020 - replaced MX60)
- 6 (2×3) ALEO S79-300's into one FM80 (2017)
- 2 grid-tied VFXR3524A-01's series stacked, replaced GVFX's (2020)
- 12 Surrette/Rolls 2V x 1766Ah (2007)
- Hub10.3, Mate3s, FNDC, RTS. Tigo ES maximizers on each PV module.

Westinghouse Solar - Barn roof: (2012)
- 30 (2x15) 235W panels with Enphase M215 microinverters, grid-tied

Outback Skybox - Barn roof: (2019)
- 14 Talesun 275W in series (DC array input to SB inverter/charger)
- 3 SimpliPhi 3.8 batteries, 48V, 225Ah total
- AC coupled input from 14 Talesun 275W, Enphase M215 microinverters

All self-designed and self-installed
Location: Chagrin Falls, Ohio

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by Mike Curran »

That 6% drop corresponds exactly to how, in a Car & Driver test, Tesla batteries responded to recharging to 100% on a regular basis, also over a 3 year period. But also with frequent "supercharging", another factor. That test used Tesla's charging routine, so hard to say how voltage level may have affected the results.
https://smart.tigoenergy.com/p/pZXn7SZQyO45?date=2021-04-19
https://enlighten.enphaseenergy.com/public/systems/Hctc107221
KeithBriggs
Forum Virtuoso
Posts: 76
Joined: Thu Apr 12, 2018 7:56 pm
My RE system: 16k 2 8k Radian 4 FM80 48 300w panels, 6 3.4 simpliphi Grid-tied, HBX mode connect 50.4 disconnect 51.2.

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by KeithBriggs »

No worries raysun. Thanks for all that information. Helps me understand it much better. I just kissed 100% yesterday so no much ion packing and didn't quite connect to the grid this morning. I might already have some improvement though as my SOC was a couple % higher even though I consumed 13.22 kwh.

Its worth re-stating that way back when I first installed, I didn't wire correctly. I had different lengths to each battery (conservation of AWG2 copper wire). So for about a week I was overcharging and over draining a few of them. Dan at simpliphi corrected me and interestingly said that they would work themselves out in a couple days (not weeks). That's why it was good for me to see years later that every battery (to the 3rd digit) was reading the same voltage.

I have only 6 bats for a 16kw system so cannot use all 66 amps AC until I get a couple more. Dan a month ago said as long as the original batts are less than 5 years old, its okay to add more. And he said its okay to mix and match any 3.x again within the 5 year age difference.

The new app note nicely done and is attached.

Mike, similar to what you found, I like this article https://www.geotab.com/blog/ev-battery-health/

There is some crossover with our Lith chemistry.
raysun
Forum Emperor
Posts: 4581
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower Two: (2) FXR3048A, (2) FM80, MATE3s, FlexNetDC
SimpliPhi 48-3.8 (6 @ 48v)
Outback IBR3 battery enclosure
REC Alpha 440W panels - 2 arrays: each of 4 strings of 2 in series
Honda EU7000is gas fuel generator

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by raysun »

The great thing about adding the newer battery blocks is, while they must be derated to the capacity of the existing blocks, lithium loves partial state of charge.

The 3.8 blocks will not see 100% when the 3.4 blocks do, so will happily sit with the cells less stressed.

There's no risk of cell imbalance like with lead acid as the charge/discharge curve is so flat.

Also, with added net capacity, pouring in and draining out every drop of charge becomes less of an issue.

Having a large(r), easily maintained battery is the best thing this off-gridder has done in years to improve his mental health. ;)
KeithBriggs
Forum Virtuoso
Posts: 76
Joined: Thu Apr 12, 2018 7:56 pm
My RE system: 16k 2 8k Radian 4 FM80 48 300w panels, 6 3.4 simpliphi Grid-tied, HBX mode connect 50.4 disconnect 51.2.

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by KeithBriggs »

At about 13:30 I happened to walk by laptop and Optics said voltage was 55.9 and SOC was 99% Then I looked it again and it was 53.6. I think it kissed 56v and dropped into float as two charge controllers were silent and two were floating. After checking several times over last hour its been running about 2 kw in and 1.8 kw out (mostly the car) and still 53.6. Not seeing anything above 54v for a year, this is promising. I will make a Lowes run tonight and charge for a 3 or 4 hours guaranteeing a draw on the meter but no matter as helco gives you 50 kwh per month included with the meter fee.

Update: about 15:30 SOC was still reading 99% and volts were still 53.6 so (like in another post I saw and I've done before) I forced all 4 CC (charge controllers to bulk which is about 3 buttons on FM80s and oddly you don't have to put in the code. It took only about 3 minutes for them all to be charged again. Interestingly and not sure if its because I did the bonus bulk or just natural ion packing but now at 17:30 near sunset all CC's are floating and 100% but now at 53.7v. Extra 0.1v either from packing or rounding error.

Clearly updated settings are restoring capacity. This is the second night. Like yesterday, I didn't quite connect to the grid but last night I used 0.52 kwh more but my SOC is 2% higher. I plugged the car in and my wife fired up the electric skillet and toaster so for a while, we had 3.8kw going or about a C/6 load with SOC under 30% and voltage was 50.6.

First night showed about 3x the improvement of last night. Not sure where I'm going to land but headed in the right direction with the new parameters.
KeithBriggs
Forum Virtuoso
Posts: 76
Joined: Thu Apr 12, 2018 7:56 pm
My RE system: 16k 2 8k Radian 4 FM80 48 300w panels, 6 3.4 simpliphi Grid-tied, HBX mode connect 50.4 disconnect 51.2.

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by KeithBriggs »

2 dark days, not connecting the grid but not getting even near 100%. 16.5kwh from the sun today. Yesterday got to 76% SOC and today only got back up to 59% so likely to connect to grid in the morning.

Matthew at simpliphi was poopooing my 6 batteries on a 16kw system and 14,400w even tho I explained that loads are on timers and panels are on 3 pitches. They are covering their butts to avoid any warranty replacements.

My supplier pointed me to some 100ah 48v for $2200 each. I went to alibaba and found several options. Since I have the 3 batts in CO, I'm going to haul them to HI and have a total of 9 here and my new 8k system in CO is going to have chinese Lifepo4's given my latest go-round with Phi. I wasn't even hinting that I had a warranty issue but they already have their defensive posture.

My search is far from exhaustive but if you want LFP alternatives, this is what I like so far:

https://www.lithium-battery-factory.com ... n-battery/
and
https://arklithium.com/products/

my supplier 2200 quote was for the ark.
sbrownian
Forum Czar
Posts: 538
Joined: Mon Nov 30, 2020 9:23 am
My RE system: GS8048 FM80 FNDC Mate3
4kw solar
7kw Kubota propane gen
Had 1200 Ahr FLA Locomotive batts
Just replaced with
2 x 300 Ah ReLion Lithium packs
MUCH more efficient!!
Location: Central Washington State

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by sbrownian »

Those ArkLithium batts have a BMS system that ALL the manufacturers should emulate.

Not a big fan of Bluetooth, but to be able to see what the heck is going on in the box can be invaluable.
Information, you get not; if incorrect question, you ask..
KeithBriggs
Forum Virtuoso
Posts: 76
Joined: Thu Apr 12, 2018 7:56 pm
My RE system: 16k 2 8k Radian 4 FM80 48 300w panels, 6 3.4 simpliphi Grid-tied, HBX mode connect 50.4 disconnect 51.2.

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by KeithBriggs »

I agree re: those arks esp the active balancing. Oddly, manufacturer wants 2500 per bat (vs 2200 - head shaking). I also like the 5 yr warranty with ark vs the L-B-F's 3 year and the chinese company requires 10 unit minimum. I asked them for a price for 10 units and they asked for my phone number. That's how I left it with them.
KeithBriggs
Forum Virtuoso
Posts: 76
Joined: Thu Apr 12, 2018 7:56 pm
My RE system: 16k 2 8k Radian 4 FM80 48 300w panels, 6 3.4 simpliphi Grid-tied, HBX mode connect 50.4 disconnect 51.2.

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by KeithBriggs »

After about a week with the latest parms, I've got about 1/3 of my lost capacity back. After thinking about and messing with it for much too much time. This is my summary.

My 3.4 were behaving like 2.8's. I'm up to 3.0 and hope I can get to 3.1. Put another way, I was down 3kwh, now I'm down 2kwh.

I'm not sure if ion packing is best by dropping to 20 something SOC (50.x volts) and working back to 100% or going light and camping at 100% for 1/2 the day. Certainly easier to add loads (L1 car charging is the easiest), but its harder to ask my wife to hand wash the dishes or clothes.

LFP is resilient if you stay within defined parameters.

LFP is NOT reslient if you venture outside those defined parameters.

Simpliphi does not have your back. They will error on blaming you for your errors.

According to arklithium, LFP loses 1% capacity per year staying within defined parms.

I'm happy to eat humble pie for:
-- having inconsistent battery cable lengths when I first installed for about 10 days.
-- rarely loading the batteries with more than C/2
-- once when off island, the batteries dropped below 50v and the breakers tripped.

From the arklithium product page, I now think of the cells as cups. Illustratively, overcharging or over-draining them is like to spilling or squeezing the cup. It might spill back in and the cup might bounce back near its original shape. To some degree, loss is not reversible.

My understanding is that breakers protect wiring from over current. The breaker on the Phi's should trip if overcharging or over-draining and not just voltage. I don't understand the battery breakers and I don't trust them.

3 more 3.x Phis will put me at 22.5 (hopefully 23.5) kwh but more importantly, I can dial my inverter back to 16kw and my charge controllers back to 80a. Note that I've never overcharged given the 3 different pitch angles of my panels.

I hope to give a final update in about a week.
raysun
Forum Emperor
Posts: 4581
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower Two: (2) FXR3048A, (2) FM80, MATE3s, FlexNetDC
SimpliPhi 48-3.8 (6 @ 48v)
Outback IBR3 battery enclosure
REC Alpha 440W panels - 2 arrays: each of 4 strings of 2 in series
Honda EU7000is gas fuel generator

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by raysun »

"I'm not sure if ion packing is best by dropping to 20 something SOC (50.x volts) and working back to 100% or going light and camping at 100% for 1/2 the day. Certainly easier to add loads (L1 car charging is the easiest), but its harder to ask my wife to hand wash the dishes or clothes."

Drawing concensus from the articles and research papers I've read over the past 7 years, here's what stuck in my head:

Lithium charging is more akin to shuttling ions between the electrodes than a fundamental chemical state change. The notion of the cells as simple containers is not too far wrong.

The charge/discharge process degrades all active cell components: cathode, anode, electrolyte, and separators. The rate of degradation is influenced by depth of discharge, "height" of charge, and rate of charge/discharge. Temperature is also a factor, and the ideal is 50-80°F.

Lithium cells are degraded least when not completely discharged, and not fully charged. There seems to be an ideal longevity between 80% / 20% SoC. The lower the rate of charge and rate of discharge, the longer the cells will last as well.

Charging to 100% is hard on the cells. The Lithium ions pack imperfectly in the anode, leading to spiky lumps called "dendrites", which can grow to the point they pierce the separator between anode and cathode, causing short circuits and capacity loss.

Holding the cells at 100% SoC and adding further charge can lead to accelerated cell failure. Float charging is not recommended. Lithium cells in purely float service have very short service life.

In the case of cell "rehab" the two competing conditions must be balanced. Like Equalization charging in Lead Acid batteries, the process of bringing the specific gravity in line in each cell, simultaneously degrades the cell structure. Similarly, "repacking" the depleted anode in a lithium battery both increases capacity and degrades the anode structure. Don't try to force too much, too fast. Don't expect full restoration of capacity.

Lithium cells are limited both in the rate of charge they will accept, and the rate of discharge they can safely deliver. All battery cells are, actually, but Lithium is far more restricted than equivalent lead acid in terms of rate v.s. cell life. Rapid charging is possible, but shortens cell life materially vis-a-vis slow charging. Rapid discharge is also possible, but again, reduces service life.

There are two approaches to Lithium battery charging: Constant Current to a certain state of charge, then constant voltage until full; and constant current until full, then terminate charge. Its not clear (to me, at any rate) that one is superior to the other. It may depend on available charging current. My system seems to naturally self-regulate to around 60A into a battery that is rated at 225A maximum charge. With this "slow" nominal rate of charge, I use the "Bulk until full" method. The recommended termination voltage is below the cell maximum voltage, so I feel confident the method is not creating undue stress (but begs the question, what is the ideal "due stress"). Time will tell if it is the right choice.

Another issue that has been debated for years is the concept of cell balancing. It is one in which I take no sides, but do have an opinion. The goal of balancing is two fold: to maximize capacity while minimizing the risk of degradation. Both worthy pursuits. Lithium batteries are expensive, getting the most "mileage" per charge cycle is an important concern for mobile applications, though less so for fixed applications. Keeping cells from voltage extremes of over-charged, and over-discharged, is likewise important for health. In a multi-cell battery (of any chemistry) unbalanced cells can lead to some failing prematurely due to the added stress caused by out-of-band voltage conditions.

In lithium batteries, the notion of "top balancing", getting every cell to the same state of full; and "bottom balancing", keeping every cell from going past a minimum voltage; are the two primary strategies to maximizing capacity while staying out of the "danger zone". How effective they are depends on one's objectives. If the objective is to squeeze every drop of charge possible out of the collection of containers, then balancing is a good strategy. If the objective is long service life, then squeezing the maximum amount of energy into and out of the cells is counterproductive.

Again, I take no side in this debate, but in the DIY world, one of the early pioneers of lithium battery use started out as the strongest proponent of balancing, and eventually became one of the strongest opponents.

So if one chooses not to walk the straight and narrow path in balance, what is one to do? Simply avoid the edges. In my case, the "cheat" is to have a battery that is larger than my needs. I am committed to daily cycling. I have solar PV, and the sun comes up every day, so the commitment is natural. The capacity above 20% SoC is more than enough to meet our most demanding consumption. In fact, its a struggle to dive as deep as 40%. The days the charging drives the battery to 100% are frequent enough to keep my system monitors in sync. The days peak charge is less is just fine. I consider hitting 80% SoC as a Red Letter Day. IMO, should the cells be a bit out of balance is of little consequence in this scenario.
Last edited by raysun on Fri Jan 22, 2021 9:52 am, edited 1 time in total.
provo
Forum Czar
Posts: 720
Joined: Thu Jul 27, 2006 4:34 pm
My RE system: Sixteen Evergreen EC-120
(4 strings, total 1920W)
Eight Rolls S-550 (2 strings, total ~800Ah @ 24V)
One FM80
One VFXR3524A
Hub 10
Mate3s
FNDC and Trimetric
Honda EU3000is generator
Location: Sierra foothills

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by provo »

Thanks, raysun...that whole post is now permanently ensconced in my Batteries folder :grin:
raysun
Forum Emperor
Posts: 4581
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower Two: (2) FXR3048A, (2) FM80, MATE3s, FlexNetDC
SimpliPhi 48-3.8 (6 @ 48v)
Outback IBR3 battery enclosure
REC Alpha 440W panels - 2 arrays: each of 4 strings of 2 in series
Honda EU7000is gas fuel generator

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by raysun »

provo wrote: Thu Jan 21, 2021 5:45 pm Thanks, raysun...that whole post is now permanently ensconced in my Batteries folder :grin:
I hope I know what I'm talking about. 🤣🤣
sodamo
Forum Emperor
Posts: 1140
Joined: Mon Apr 03, 2006 1:55 am
My RE system: Totally off grid - Hamakua side, Big Island, Hi

36 Trina 280 & 16 Phono 250 in 48 volt array (fixed) 14080w
Radians GG8048/GS4048 w GSLCs
4 FM 80 charge controllers
Hub 10.3
FNDC w/3 shunts
Mate 3s
OpticsRE - MMKL - Ninole Hi
12 SimpliPhi 3.8
14Kw Kohler 14RESA Propane

Charging Only
2FX3048T
Hub
FNDC. - shared shunt feeds from Radian FNDC
Mate 3s
10Kw MEP 803a Military gen

Honda EU7000i

Davis VantagePro2 Wx Station
On line at: https://www.wunderground.com/dashboard/pws/KHININOL2
Location: Ninole, Hi

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by sodamo »

Ray, don’t know if you do or if you don’t, but it was a good read 🤙🤙🤙
David
Please visit http://vacation.ninolehawaii.com
raysun
Forum Emperor
Posts: 4581
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower Two: (2) FXR3048A, (2) FM80, MATE3s, FlexNetDC
SimpliPhi 48-3.8 (6 @ 48v)
Outback IBR3 battery enclosure
REC Alpha 440W panels - 2 arrays: each of 4 strings of 2 in series
Honda EU7000is gas fuel generator

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by raysun »

sodamo wrote: Thu Jan 21, 2021 11:10 pm Ray, don’t know if you do or if you don’t, but it was a good read 🤙🤙🤙
TY
sodamo
Forum Emperor
Posts: 1140
Joined: Mon Apr 03, 2006 1:55 am
My RE system: Totally off grid - Hamakua side, Big Island, Hi

36 Trina 280 & 16 Phono 250 in 48 volt array (fixed) 14080w
Radians GG8048/GS4048 w GSLCs
4 FM 80 charge controllers
Hub 10.3
FNDC w/3 shunts
Mate 3s
OpticsRE - MMKL - Ninole Hi
12 SimpliPhi 3.8
14Kw Kohler 14RESA Propane

Charging Only
2FX3048T
Hub
FNDC. - shared shunt feeds from Radian FNDC
Mate 3s
10Kw MEP 803a Military gen

Honda EU7000i

Davis VantagePro2 Wx Station
On line at: https://www.wunderground.com/dashboard/pws/KHININOL2
Location: Ninole, Hi

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by sodamo »

Now that the pain of purchase is passed I take a bit of comfort overriding 2 installers recommending 10 blocks and insisting on 12. On days I touch 100% I drop into the 60’s overnight.
David
Please visit http://vacation.ninolehawaii.com
raysun
Forum Emperor
Posts: 4581
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower Two: (2) FXR3048A, (2) FM80, MATE3s, FlexNetDC
SimpliPhi 48-3.8 (6 @ 48v)
Outback IBR3 battery enclosure
REC Alpha 440W panels - 2 arrays: each of 4 strings of 2 in series
Honda EU7000is gas fuel generator

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by raysun »

sodamo wrote: Thu Jan 21, 2021 11:36 pm Now that the pain of purchase is passed I take a bit of comfort overriding 2 installers recommending 10 blocks and insisting on 12. On days I touch 100% I drop into the 60’s overnight.
After walking a tightrope with a VRLA battery that would barely make it to 8 in the morning, I luxuriate in the fact that, on the most demanding span of days I've ever had with this battery, I still had 25% to burn before I even got near the lower limit.

I've always opined that installing an overly large lead acid battery was a mistake due to the demands of proper maintenance. With lithium, it seems, bigger is better. If I could justify the expense, I'd have twice as large a battery as i have now.

The biggest benefit may just be the reduced stress of maintenance. I guess lithium is an effective mood stabilizer whether taken internally or applied externally.

I know @K and @G are wrestling with declining capacity issues right now so my rosy evaluation probably rings a bit hollow. I've been through battery warranty issues myself, to no avail, so that part doesn't seem to improve with the change in chemistry. A small bright spot, however, is the existing battery can be supplemented with additional blocks if need be.
gtarolli
Forum Guru
Posts: 419
Joined: Thu Dec 21, 2017 10:39 am
My RE system: Outback: off-grid
- 1 GS8048A, Mate3, FlexNetDC, Hub10.3
- 2 FLEXmax 80
- 1 FLEXmax 60
- 30 (10x3) 300w panels (9000w total)
- 6 SimpliPhi 3.4 kWh LI batteries (400ah , 20kWh)
- 1 Honda EM5000 generator + one EU3000
Location: Wainiha, Kauai, HI

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by gtarolli »

I believe I am the other (2nd?) person having the approx 16% loss in capacity after a few years. This seemed to be noticeable in June/July a few months after I went with the slow-and-steady charging profile, where I set absorb to 55v for 1 hr (54.8 seemed too low). I didn't realize it might take a few weeks or month to perhaps lower the capacity. I think I did set my absorb back to 56v for a few days, but earlier in this thread I read it could take 2 weeks to see the change. So today I am setting it back to 56.4v for .1 hr. Because voltage varies so much with PV charge amps, I much prefer this method. As noted many times in forums you can reach 55v with 4-8kw of PV charging at a low SOC % , e.g. 80%. The "float mode" battery protection can prematurely reset SOC to 100% and if you have auto-charge-termination enable, you will go into float mode. I have it disabled for this reason - I think with LiFP it seems safer to just reach 56.4v and stop. The 6 minutes it stays there should not be an issue (I hope!).

Note temp. compensation can vary your voltages by .4v pretty easily, so I think its wise to disabled it (I use a limited temp comp hack to do this).

Keith - who different were your wire lengths? I have 2-3' of very thick wire going from the radian to a 2-way split and then thinner gauge running to two splitter blocks that branch each wire into about 1' of thinner gauge to each of 3 batteries. Some of these wires could be 1-2" shorter than others, probably out of 6' total length. let's say perhaps a 2% difference. is that enough to cause issues?

My other guess as to what happened, as my capacity drop seemed sudden, was one of the battery connections in the 3-way splitter block could have gone bad. I should have replaced these with an 8 connection (1 in , 6 out) block, but they were out-of-stock and my installer had the smaller 4 connection (1 in, 3 out) blocks available.

So... I am changing to 56.4v today and will wait a few weeks to see how things go. My way of measuring things is when the SOC line crosses the voltage line (about 52%) in the early morning with low loads, I look at the voltage for say 50, 52, 54% SOC. It used to be 52.3 to 52.5 and didn't drop below 52v until mid-40% SOC. Now it seems to be about .4v lower.

Hoping for the best ....
raysun
Forum Emperor
Posts: 4581
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower Two: (2) FXR3048A, (2) FM80, MATE3s, FlexNetDC
SimpliPhi 48-3.8 (6 @ 48v)
Outback IBR3 battery enclosure
REC Alpha 440W panels - 2 arrays: each of 4 strings of 2 in series
Honda EU7000is gas fuel generator

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by raysun »

gtarolli wrote: Fri Jan 22, 2021 10:27 am I believe I am the other (2nd?) person having the approx 16% loss in capacity after a few years. This seemed to be noticeable in June/July a few months after I went with the slow-and-steady charging profile, where I set absorb to 55v for 1 hr (54.8 seemed too low). I didn't realize it might take a few weeks or month to perhaps lower the capacity. I think I did set my absorb back to 56v for a few days, but earlier in this thread I read it could take 2 weeks to see the change. So today I am setting it back to 56.4v for .1 hr. Because voltage varies so much with PV charge amps, I much prefer this method. As noted many times in forums you can reach 55v with 4-8kw of PV charging at a low SOC % , e.g. 80%. The "float mode" battery protection can prematurely reset SOC to 100% and if you have auto-charge-termination enable, you will go into float mode. I have it disabled for this reason - I think with LiFP it seems safer to just reach 56.4v and stop. The 6 minutes it stays there should not be an issue (I hope!).

Note temp. compensation can vary your voltages by .4v pretty easily, so I think its wise to disabled it (I use a limited temp comp hack to do this).

Keith - who different were your wire lengths? I have 2-3' of very thick wire going from the radian to a 2-way split and then thinner gauge running to two splitter blocks that branch each wire into about 1' of thinner gauge to each of 3 batteries. Some of these wires could be 1-2" shorter than others, probably out of 6' total length. let's say perhaps a 2% difference. is that enough to cause issues?

My other guess as to what happened, as my capacity drop seemed sudden, was one of the battery connections in the 3-way splitter block could have gone bad. I should have replaced these with an 8 connection (1 in , 6 out) block, but they were out-of-stock and my installer had the smaller 4 connection (1 in, 3 out) blocks available.

So... I am changing to 56.4v today and will wait a few weeks to see how things go. My way of measuring things is when the SOC line crosses the voltage line (about 52%) in the early morning with low loads, I look at the voltage for say 50, 52, 54% SOC. It used to be 52.3 to 52.5 and didn't drop below 52v until mid-40% SOC. Now it seems to be about .4v lower.

Hoping for the best ....
Yes, @G is U.

One thing I noticed with the 54.4V constant voltage (Absorb) charging profile is the 1 Hour Absorb Time originally specified by SimpliPhi was too short. They seem to agree, and the latest integration guide specifies 2 Hour. I found that to be too long, and like lobbing mortar rounds, the mid-point between the two specs was just right. I ran the 54.4V Absorb phase for a week, and found the Charged Return Amps (2% of C20) was reached at 1:29 each time.

While I don't trust the FNDC to accurately track State of Charge (among other things, Peukerts Constant is wrong for lithium), I trust it explicitly for charge termination. Using the 56.4V "Bulk until full" profile, or the 54.4 Bulk/Absorb profile, charge termination can be dialed in and execute quite consistently. I use the 56.4 profile, and can watch the charge terminate right on the recommended ending amps (2%), 60 seconds after the termination voltage (56.4) is reached. It took a little observation and tweaking, as the charge current declines precipitously when the battery is full. (I use the web app in the Mate for observing, as it updates rapidly.)

My Battery Charge parameters are below. Notice the Charge Return Amps is quite high. At the end of Charged Time (60 Seconds) the charging amps hits 9A (2%). All this concides with reaching 56.4V, and the FM80 charge controllers report 1 Minute of Absorb Time at charge termination. (Absorb Time is set at 0.1H, but the other 5 minutes never execute.)
Screenshot_20210122-084504_Chrome.jpg
gtarolli
Forum Guru
Posts: 419
Joined: Thu Dec 21, 2017 10:39 am
My RE system: Outback: off-grid
- 1 GS8048A, Mate3, FlexNetDC, Hub10.3
- 2 FLEXmax 80
- 1 FLEXmax 60
- 30 (10x3) 300w panels (9000w total)
- 6 SimpliPhi 3.4 kWh LI batteries (400ah , 20kWh)
- 1 Honda EM5000 generator + one EU3000
Location: Wainiha, Kauai, HI

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by gtarolli »

I sometimes raise Charged Amps to 32 or 45, as it doesn't seem to trigger. I still am not convinced it measures net amps. I have my efficiency set down at 94% as that seems make it so the voltage rises rapidly when I hit 98% SOC. So I trust FNDC more than most.

My only doubt is that after .1hr at 56v absorb, I did start Bulk again and the voltage immediately go back to 56v and then amps go down to just my loads. So I was "full" at 56v. But maybe full at 56.4v is say 10% more amps (in my case 40 amps, I have 5*3.4 and 1*3.5 batteries).
User avatar
EA6LE-ONE
Forum Guru
Posts: 497
Joined: Tue May 05, 2020 7:51 am
My RE system: System 1:
MATE3s, 3 x Radian GS8048A,
6 x FLEXmax 100, HUB10,
2 x EnergyCell 48V 2700RE,
1 x FLEXnet DC, 6 x FLEXware ICS Plus,
78 x PANASONIC VBHN325SA17 325W

System 2:
MATE3s, Radian GS8048A,
2 x FLEXmax 100, HUB10
1 x EnergyCell 48V 2700RE,
1 x FLEXnet DC, 2 x FLEXware ICS
21 x PANASONIC VBHN325SA17 325W
Location: Providenciales, Turks & Caicos Isl.

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by EA6LE-ONE »

You might want to experiment using one shunt operation for the FNDC and have the inverters and the CCs connected before the shunt to the battery. maybe the numbers will look better.
raysun
Forum Emperor
Posts: 4581
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower Two: (2) FXR3048A, (2) FM80, MATE3s, FlexNetDC
SimpliPhi 48-3.8 (6 @ 48v)
Outback IBR3 battery enclosure
REC Alpha 440W panels - 2 arrays: each of 4 strings of 2 in series
Honda EU7000is gas fuel generator

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by raysun »

"My only doubt is that after .1hr at 56v absorb, I did start Bulk again and the voltage immediately go back to 56v and then amps go down to just my loads. So I was "full" at 56v. But maybe full at 56.4v is say 10% more amps (in my case 40 amps, I have 5*3.4 and 1*3.5 batteries)."

Above 56V more charge current may go into the battery, but I don't think it results in more charge. It may drive intercalation (probably good), and heating the electrolyte (probably not so good.)

However, once a cell is full, it doesn't normally accept more charge current.
JRHill
Forum Emperor
Posts: 1519
Joined: Thu Feb 17, 2011 11:12 am
My RE system: VFXR3648, FM80, 3k panels, Mate3s, FNDC, 4ea 3.8 PHIs. EU7000is AGS.
Location: PNW
Contact:

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by JRHill »

raysun wrote: Fri Jan 22, 2021 10:51 am
gtarolli wrote: Fri Jan 22, 2021 10:27 am I believe I am the other (2nd?) person having the approx 16% loss in capacity after a few years. This seemed to be noticeable in June/July a few months after I went with the slow-and-steady charging profile, where I set absorb to 55v for 1 hr (54.8 seemed too low). I didn't realize it might take a few weeks or month to perhaps lower the capacity. I think I did set my absorb back to 56v for a few days, but earlier in this thread I read it could take 2 weeks to see the change. So today I am setting it back to 56.4v for .1 hr. Because voltage varies so much with PV charge amps, I much prefer this method. As noted many times in forums you can reach 55v with 4-8kw of PV charging at a low SOC % , e.g. 80%. The "float mode" battery protection can prematurely reset SOC to 100% and if you have auto-charge-termination enable, you will go into float mode. I have it disabled for this reason - I think with LiFP it seems safer to just reach 56.4v and stop. The 6 minutes it stays there should not be an issue (I hope!).

Note temp. compensation can vary your voltages by .4v pretty easily, so I think its wise to disabled it (I use a limited temp comp hack to do this).

Keith - who different were your wire lengths? I have 2-3' of very thick wire going from the radian to a 2-way split and then thinner gauge running to two splitter blocks that branch each wire into about 1' of thinner gauge to each of 3 batteries. Some of these wires could be 1-2" shorter than others, probably out of 6' total length. let's say perhaps a 2% difference. is that enough to cause issues?

My other guess as to what happened, as my capacity drop seemed sudden, was one of the battery connections in the 3-way splitter block could have gone bad. I should have replaced these with an 8 connection (1 in , 6 out) block, but they were out-of-stock and my installer had the smaller 4 connection (1 in, 3 out) blocks available.

So... I am changing to 56.4v today and will wait a few weeks to see how things go. My way of measuring things is when the SOC line crosses the voltage line (about 52%) in the early morning with low loads, I look at the voltage for say 50, 52, 54% SOC. It used to be 52.3 to 52.5 and didn't drop below 52v until mid-40% SOC. Now it seems to be about .4v lower.

Hoping for the best ....
Yes, @G is U.

One thing I noticed with the 54.4V constant voltage (Absorb) charging profile is the 1 Hour Absorb Time originally specified by SimpliPhi was too short. They seem to agree, and the latest integration guide specifies 2 Hour. I found that to be too long, and like lobbing mortar rounds, the mid-point between the two specs was just right. I ran the 54.4V Absorb phase for a week, and found the Charged Return Amps (2% of C20) was reached at 1:29 each time.

While I don't trust the FNDC to accurately track State of Charge (among other things, Peukerts Constant is wrong for lithium), I trust it explicitly for charge termination. Using the 56.4V "Bulk until full" profile, or the 54.4 Bulk/Absorb profile, charge termination can be dialed in and execute quite consistently. I use the 56.4 profile, and can watch the charge terminate right on the recommended ending amps (2%), 60 seconds after the termination voltage (56.4) is reached. It took a little observation and tweaking, as the charge current declines precipitously when the battery is full. (I use the web app in the Mate for observing, as it updates rapidly.)

My Battery Charge parameters are below. Notice the Charge Return Amps is quite high. At the end of Charged Time (60 Seconds) the charging amps hits 9A (2%). All this concides with reaching 56.4V, and the FM80 charge controllers report 1 Minute of Absorb Time at charge termination. (Absorb Time is set at 0.1H, but the other 5 minutes never execute.)

Screenshot_20210122-084504_Chrome.jpg
Just to confirm: you were using the 56.4 profile with charged voltage at 55.6v? If the FNDC meets the charged parameters at or passing 55.6v how does that affect the FNDC SOC calculation if you are running up to 56.4 with charge termination disabled? Does the FNDC tracking stop at 55.6 or does it continue up to 56.4vdc? Thinking of it as counting in and out it would seem that it tracks to the top and counts down from there. And it would follow the SOC is reset at that upper threshold.

As for the FNDC sub-forum question, how the history plays into this is still another curiosity. The answer to that is better left there.

Whether this thread or the post I just put into the FNDC sub-forum, I'm trying to get an understanding of some curiosities.
raysun
Forum Emperor
Posts: 4581
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower Two: (2) FXR3048A, (2) FM80, MATE3s, FlexNetDC
SimpliPhi 48-3.8 (6 @ 48v)
Outback IBR3 battery enclosure
REC Alpha 440W panels - 2 arrays: each of 4 strings of 2 in series
Honda EU7000is gas fuel generator

Re: Phi 3.4 48v 2.75 years old running at 82.5% capacity

Post by raysun »

The App Note on FNDC Battery Charge says to set Charged Voltage 0.4V lower than the lowest Absorb Voltage on the chargers. On my system, the FM chargers are set to 56.4, and the FXR chargers to 56.0.

Watching the charge cycle, once the voltage rises to 55.6, it continues rapidly into the 56+V realm.

I may reset the FXR chargers to 56.4V as well, as I'm not so concerned about the FMs having priority for the 60 seconds it takes to transit from Charged Voltage to charge termination. In that case, I'd reset the Charged Voltage to 56V.

I do use Charge Termination, and the evolution from charging at the Battery Charged voltage to 56.4V @ 2% of C20 charging current occurs in 60 seconds on my system.

Depending on the number of days since Charged Parameters Met, the FNDC may have drifted off the true SoC. My system always reaches 100% before Charged Parameters Met is set, and I simply ignore it, as charging continues past that point to the true 100% SoC at charge termination.
Last edited by raysun on Fri Jan 22, 2021 3:00 pm, edited 1 time in total.
Post Reply