Wolf's PowIRwall

[Cheap 4-life]

So yeah I meant the fans for the little heaters. Once you get warmer days to not freeze off the fingers you said you will make it so there’s less moving parts. Once little heaters are in the middle divider (removing circulator fan) it should be ok maybe even better that the little heaters fans stop working when they aren’t on

 

[ajw22]

Say anything below 10°C fans would run continually and the heating matrices would cycle on and of at 8°C.

That sounds like the ideal approach to me.
Fans use relatively little power (1~2W?) compared to heaters (20~60W?), not to mention that those 1~2W the fan uses becomes heat, too.
So you'll probably want to use the fan as much as possible to maintain a low, but consistent temperature throughout the box. And the heater only as a last resort.

Not sure if this is an issue, but Air, like everything else, likes to take the path of least resistance, so you might have hot spots on the far left, and cold spots on the far right edges. Making the gaps in the shelf different sizes (small in center, bigger at the edges) would make the airflow more uniform. Perhaps you could then safely optimize further and reduce the target min temperature to 5°C.

 

[Wolf]

Update:
The Frankenstein Battery has been in operation for 9 Months and 22 Days. Today was one of the first days we had great sunshine for a continuous period of time. During these peak sun times I am getting higher voltages every time and the pack voltages are tightening up again as they should.
At my set DOD of ≈3.5v the spread has been consistently ≈ 0.065v. Today was the first day since winter set in, that Battery voltage has been this high. I was pleasantly surprised to see the pack voltage difference tighten up to 0.01v. This is all without any external influence or balancing. I am looking forward to the Packs reaching 4.1v again and have a 100% SOC. Right now Batrium is recalculating the SOC as I get higher voltages and SOC goes past 101.5%.


Battery #2 is ready to be installed as I have now gotten the lexan side protectors that will be installed as soon as temperature permits.

Also Battery #3 is in the works. All cells have been tested and sorted using Nemos Repackr.
I found these new style cell holders that are "fireproof" Not sure if that is true but they claim they are made of fireproof material, Non-flammable, it will automatically extinguish when encountering an open flame, effectively reducing the occurrence of fire accidents
¯\_(?)_/¯
Maybe?

Anyhow they look nice and work well. I figure with all the hubbub surrounding fires it couldn't hurt and the price is comparable.
The 14 pack setup. Halfway there and completed packs waiting for cell leve fuse sheets to be spot welded.

Out of 1375 cells 1133 made the grade and with this small of a pool of cells this is the results of the repacker.

Not bad
BTW all of these cells were tested with the Megacell Charger.

All details where recorded from the first touch of the cell including the voltage discrepancy of the MCC to actual.

If you want to look at the spreadsheet with all the numbers and glaze your eyes over help yourself.

Harvest LGFBM161865.xlsm

Microsoft Excel Workbook

1drv.ms

Also the Repackr info can be found here.

Nemo Repackr 3rd Battery.xlsm

Microsoft Excel Workbook

1drv.ms

Wolf

 

[Wolf]

Another beautiful day till the clouds showed up. Made clearing the 3" of snow off of the panels last night worth it.
No matter how you look at it 9.2kW produced today on the 10 of February is a beautiful sight.
Batrium is working great the Victrons are working great all is working great.

 

[jdeadman]

Nice Wolf!!!!

 

[Wolf]

I think I have finally lost it. That's it I have gone totally anal and I belive I need to be committed. Please help me out.
After all the talk of cells catching on fire and all that sort of stuff I am now more than ever committed to building the safest and most durable battery that is possible.
Anyone that has read my build thread knows how I go about testing cells and how meticulous I am as to what cells will pass my criteria and finally find a place in my powerwall. I'm sure you all are tired of this by now nevertheless I will reiterate for the first time readers.
My first process is logging all the cells info into a spreadsheet.
It has grown to this monstrosity of collecting as much data as I can. 29 data points and 1 for notes.

One thing I have found out is that you can never have too much information.
The spread sheet allows me to sort all the cells by specific parameters. To build my packs, I chose the highest initial voltage, the lowest post test IR, the lowest voltage drop after the >25 day SD test and of course capacity.
I then take those results format the data for Nemos Excel Repackr. Cell Number, mAh, and IR.
You see the great thing about Nemos Excel Repackr is that its results maintain the cell number making it easy to build your packs by cell number rather than trying to find the mAh of the cell.

You can then sort by cell number and it gives you the pack number the cell should go into.

Now here comes the crazy part after all the cells have been distributed into their corresponding packs I perform a last and final voltage and IR check. At this check I compare the present IR reading to the >25 day IR check and compare the voltage reading to the value of when the cell came out of the charger/tester. At this point if there is too much of a deviation the cell gets replaced with another. So far this has not been necessary as all the cells have been within stringent parameters,. During this process I actually sort the cells again and record the row and slot they go into. So you see I am going insane with my anality.
On the other hand if I ever have a mishap with one of my batteries I at least will know after forensic analysis which cell and what number may have caused the issue. Nuts right?

Alright on a more sane note Battery #2 is almost ready to install into the Battery Box. The Lexan sides have been installed and the Logmons have been placed less 1 which still has to have the holders glued on. I also ran out of nylon nuts and washers which have been ordered. My math was off by 6 or I misplaced them.
Some pictures of the process.
Of course the Lexan sides have been cut, edges rounded, holes and access relief drilled.
Small 1/16 holes drilled into the cell holders to accommodate the #4 x 5/8 screws to hold the logmons in place.
I am actually drilling into a void and it gets no where near the cell for those that would be worried.


Battery #3 (the gray cells) is underneath and I have just finished pac #7 with spot welding the fuse sheet. I do love the Kweld it is just so predictable and consistent.


That's it for now more to come soon.
So should I be committed? If so please don't tell them where I live.

Wolf

 

[completelycharged]

You have already been committed, just to a different cell.

Monty python summed it up as well "There are a great many people in the country today, who through no fault of their own, are sane."

 

[Wolf]

Little update on the Battery Box environmental controls.
Well I am very happy with the way the heating system is now working.
After splitting the Box into a warm and cool side the convection with fan assistance is working very well.
Box and cell temperature has stabilized even through the coldest of evenings.
The green trace on the 90 day chart is the shed temp in °C and the red trace is the Battery temp. As you can see after the box modification temps have stabilized very nicely. Most of the power comes from my AGM 12V Shed Battery. I do supplement with grid when necessary and I have calculated the usage to ≈2.5kW on sub 0°C per day. When it's above 0°C its less. I can make that 2.5kWh up very quickly especially on a sunny day.
All in all I think the heating of the cells has been a successful experiment and knowing that the cells are in the ≈5°C range I have no fear of max amp charge during the cold and sunny days.
Wolf

 

[Wolf]

So this and last weekend were very productive. Last weekend as you know I was able to procure the rest of the Lexan side covers and get the purple pack ready for installation. This has been done and I couldn't be more happy.
Batrium all wired up and it has recognized all the longmons.. Battery 1 and Battery 2 Voltage equalized and all is functional. I have now doubled my capacity to ≈20kW.
Soon to be tripled. and finally quadrupled. Oh boy then what do I do?
The 1 inch thick purple insulation sheets will be replaced by lexan when it get s a bit warmer. Putting this battery in this weekend already put my frail little body into danger of hypothermia..
Wolf

 

[Wolf]

Small update:
The 3rd battery is all sorted rechecked and spot welded ready for the bus bars.
Once done the capacity testing begins let's see what we get this time. Im guessing ≈207Ah per pack tested at 4.2V to 3V.
Wolf

 

[OffGridInTheCity]

14 -> 28 "blue bars" on Batrium. Soon to be 42 - congratulations!
Looks like you have room on the shelf for 5 batteries - is that the plan?
Looks NICE!

 

[Wolf]

14 -> 28 "blue bars" on Batrium. Soon to be 42 - congratulations!
Looks like you have room on the shelf for 5 batteries - is that the plan?
Looks NICE!

Thanks I am happy about that as the sun is finally coming around the bend so to speak.
Both Victron mppts and the added panels are making quite a difference as I had a max 49A charge so far today.
Can't wait till march and then summer.
The 5th shelf is for the box's environmental electronics and heaters so no room there.
Wolf

 

[jdeadman]

Very Nice Wolf

 

[Wolf]

March 8th 2021 the first day of the year all my packs reached 4.00V yippee! and it's only 14:20 in the afternoon.
Looks like I have another 2 hrs of absorption. Maybe I can get to 100% charge.
So far no balancing required.
Wolf

 

[OffGridInTheCity]

Southern OR USA - All time record last year was on... June 21st on a clear/sunny day. Had to enable the dryer to use up the extra energy so I didn't waste it. This time of year (March) I'm maxing at 78% of that.

 

[Solardad]

Southern OR USA - All time record last year was on... June 21st on a clear/sunny day. Had to enable the dryer to use up the extra energy so I didn't waste it. This time of year (March) I'm maxing at 78% of that.

Right..?!? I had three record days (for the earliest days to break 35kwh) over the last week. It’s like some threw a light switch and production has almost doubled... almost a month early. today I ran some mini splits to use up my excess power to heat up our 3 season porch.

 

[Wolf]

................ today I ran some mini splits to use up my excess power to heat up our 3 season porch...........

Isn't it funny that we look for things to use our excess power generation on.
Just the other day (the NE has been good for sun) I was at 100% by about 1pm and I said to the housemate isn't there something we can do to use the extra. No she said I've done all the laundry, vacuumed, and the dishwasher isn't full enough. So I thought a bit and decided it would be a good idea to balance charge the lead acids. ( I have a 48V flooded cell bank for a backup to the backup) Checked all the cells for proper fluid levels They were all OK as I anticipated as I don't really use them much. (Twice this winter when the power went out due to the weather.) I fired up the 48V NOCO Genius charger set it for repair/equalization and "yea" pulling 300W of free charge energy.
The Victron Battery balancers also came to life and did their job. The NOCO Genius when done with the equalization charge goes back to maintenance keeping the lead acids happy with a full charge.

On another note The PowIRwall is doing just great. So far absolutely no issues with any of the packs. As the batteries are being exercised they are responding very well. Battery#1 (The Frankenstein) is doing quite well. With the addition of Battery#2 I now have been able to run the house all day and all night. Cell voltage deviation has also been minimal between packs with a max of ≈0.04V for the last 24hrs as reported by Batrium.
Batrium claims a voltage measurement accuracy of ±0.05 so take that for what it's worth. I am now also tracing voltage deviation between the highest and lowest packs in Grafana. It is the purple trace. You will also notice yesterday afternoon I was actually in float mode on the Victron MPPTs. Another indication that I need to get my butt in gear and get Battery#3 ready to go.

Nevertheless all 28 packs are really close to each other.
I did have a bit of top balancing occur on 3 or 4 packs from the new battery that were slightly high as they approached 4.09V out of an allowed 4.08V. That seems to have helped as today the max Voltage is at 4.08 on the offending packs. The Victron MPPTs have throttled back and are in float mode again. 2:30PM and I have another 3hrs of sun being throttled. Got to get Battery#3 finished.

I also am working on figuring out battery to inverter efficiency/inefficiency.
Preliminary calculations show ≈85% efficient or a loss of ≈15% between battery wattage output and inverter wattage output.
More to come with calculations and pictures. I know you all like pictures.
That's it for now.
Stay tuned.

Wolf

 

[OffGridInTheCity]

Fantastic to hear that things are going well!

I also am working on figuring out battery to inverter efficiency/inefficiency.
Preliminary calculations show ≈85% efficient or a loss of ≈15% between battery wattage output and inverter wattage output.
More to come with calculations and pictures. I know you all like pictures.

I use my Midnite Classic 150s daily "PV Input" totals for incoming and cheap Amazon meters on the "AIMS Inverter Outputs" as my consumption number. Here's a snap of what my AIMS have output so far for 2021 to show what I mean by 'cheap meters'.

Dividing the sum of Inverter Output (15,013kwh) by Midnite Classic Input (18,030kwh) for all of 2020 - it came in at 83.3% efficiency for 2020 using Midnite + AIMS. AIMS advertise as 88% peak efficiency but more generally seem to be 85% efficient under good circumstances.

One question I have - the numbers are only as good as the metering devices that collect the numbers. So maybe I have 81% or 85% ? who knows how accurate Midnite and cheap meters are... and 18650 is maybe 1% loss... but I'm presuming that I'm in range of expectations based on the equipment I have and don't need to take any drastic action.

I'll be interested to see how you are gathering your info as I know that you're a very precise 'kind of person' and how you interpret it

 

[Wolf]

I'll be interested to see how you are gathering your info as I know that you're a very precise 'kind of person' and how you interpret it

Well that statement puts the pressure on.
So here is my theory and process. By the way the more I learn about grafana the more I love it.
The concept is really simple divide watts out by watts in and you get efficiency or lack thereof. As in 50 watts out 100 watts in hence 50/100 =0.5 50% efficiency or 90/100 =0.9 or 90% efficient.

How to accomplish this with the sensors and equipment we have at hand.
One piece of equipment we have that is supposed to be "calibrated" is the Batrium Shunt. It in essence measures current flow and voltage. We know we can calculate wattage with the equation V*A=W hence 50V *10A =500 Watts.
The only problem is that the batrium shunt is bidirectional in escence giving us positive and negative readings. Also it is not really a good indicator of actual battery usage as the only time it is giving us an accurate amperage draw is when there is no solar input which is at night.
Additionally at that time it is a negative value and division becomes problematic without going to extensive mathematical formulas.
The other piece of equipment is the IoTaWatt which has CTs on both L1 and L2 giving me the Wattage of my inverters output.
As I subscribe to the kiss principle (well most of the time) I installed a 100A/75mv calibrated shunt that measures the actual amperage draw by the inverters through an ESP32. Just some code and math. If that reading correlates with the negative amp reading of the batrium shunt within a reasonable amount I should be on track , and it does!
So grafana to the rescue.
First things to plot all Watts in and out. Inverters, Grid, Batrium/Batteries, and Solar.

Great OK that gives me a good overall view.
Now the thing we are interested in is basically Inverter watts in and inverter watts out.
Once we get those figures we can then determine efficiency.
To accomplish this I used the grafana transformations feature to do some math with all the sensor input readings.
I came up with this graph which is the information we need watts into the inverters and out of the inverters with positive numbers.
You can see the blue trace (Inverter Output Watts) is just a bit under the red trace (Inverter Input Watts) as expected.

Now it was just a matter of calculation watts out/watts in.
Grafana again to the rescue. It provides relatively instantaneous varying results of 90.64% at this moment and with a graph over time we can get an average which in my case is 90.1937% efficiency from my inverters.
This is very encouraging as the GTIL2 inverters I have claim to have a Peak Inverter Efficiency of 90%.
Kudos for a chinese company actually telling the truth.

And another Calculation.
Now the next challenge is to put Solar Watts into the equation to determine the Theory of Everything which would unify all the fundamental interactions of nature: gravitation, the strong interaction, the weak interaction, and electromagnetism. ... Oh wrong theory.

Nevertheless adding solar watts to the whole mix will be interesting to determine whole system efficiency.

Wolf

 

[Wolf]

So I couldn't let it go I was on a roll mathematically speaking.
Here is my method of calculating the Solar to battery efficiency.
First of Victrons MPPT controllers are awesome giving me the numbers to work with.
So first calculate the actual watts the panels produce going into the Victrons which gives me total watts produced by just the panels
In this case 2864.12W.
Then calculate total Victron output.
In this case 2790.69W
Which gives us an efficiency through the Victrons of ≈0.974 or 97.4%

OK next we take the Victron total wattage output, subtract the inverter wattage draw and we get what is left over to charge the batteries.
In this case total Victron Watts is 2805.47 minus Inverter Watts 365.22 equals 2440.25 Watts that should be going to the Batrium shunt.
The Batrium shunt is claiming 2371.28 Watts so we calculate again and the results are 0.971 or 97.1%




Once the calculations are condensed the results are as such on an
ever changing basis but the numbers hover in the 96% to 98% range.

So if my math and methods are correct. from solar panels to the household plug my losses are ≈13.24%
1.71% Solar/Victron + Victron/Battery 2.17% + Inverter 9.36% = 13.24%
We shall round it out to 15% losses in which case the system is ≈85% efficient as previous preliminary calculations seemed to indicate.
I'm happy with that!
Wolf