Long time listener, first time caller from Dateland, AZ

Yep, everything you mentioned crossed my mind as well. I hope in future videos Mr. Prowse provides a link to a titanate battery he recommends to use with the Growatt. However, as I understand it, the lithium titanate construction and chemistry is supposed to be FAR safer than the LiFePO4. But that makes no never mind if it can't handle the current draw.

It appears this size cell may be 2.5 to 3.5 Ah.
"Combination: 13 Series 3 and"
3p13s, about 10 Ah at 48V for the pack.

3: Rated discharge current: greater than or equal to10A
4: maximum instantaneous current: 40A
5: Maximum operating current: 13A

So about 500W nominal, 2kW surge.

This one is called Lithium Ion, not Lithium Titanate. Maybe the latter is even more safe. LiFePO4 is apparently pretty good. If punctured or severely over-charged it does release flammable gas. Self ignition can happen, but usually doesn't.

I would use old lead-acid batteries to be cheap. It will supply the current, just not deep cycle. I use quality AGM for my backup system, good for a few hundred deep cycles. DIY LiFePO4 as I wrote earlier seems to be a good price for capacity and many cycles, but will cost you around $2000.

Ah, sure enough. I missed that it was lithium ion. My mistake.

So, I got into a debate with a friend who said that if you convert your 12 volt 400 amp hour battery bank to 24 volts, you'll lose half your charge capacity (i.e., 400 amp hours would be reduced to 200 because running batteries in series increases the voltage, but the amp hours stay the same). I argued that the run-time capacity of both configurations is the same because the better way to look at it is watt hours. 12v x 400ah = 4800wh. 24v x 200ah = 4800wh. In short, a 100 watt light bulb would run the same amount of time on either configuration (ignoring of course efficiencies affected by different wire gauges and inverters, etc.)

Do you agree?

I do find Lithium Titanate on eBay. Mostly as car batteries. There is a posting for 6 x 40 Ah, $380


That's $65 per cell for 40 Ah. Compare to $100 per cell for 280 Ah LiFePO4, about 1/4 the price.
Oh, LTO is 2.3V vs. 3.2V, so LiFePO4 is 1/6 the price per Wh.

Some people use forklift batteries. They can last 15 years, especially if sized for multiple days without sun. But they cost a few grand.

Yeah, I looked at forklift batteries, but ironically you need a forklift to move them!

sarossell said:

So, I got into a debate with a friend who said that if you convert your 12 volt 400 amp hour battery bank to 24 volts, you'll lose half your charge capacity (i.e., 400 amp hours would be reduced to 200 because running batteries in series increases the voltage, but the amp hours stay the same). I argued that the run-time capacity of both configurations is the same because the better way to look at it is watt hours. 12v x 400ah = 4800wh. 24v x 200ah = 4800wh. In short, a 100 watt light bulb would run the same amount of time on either configuration (ignoring of course efficiencies affected by different wire gauges and inverters, etc.)

Do you agree?

Click to expand...

Same watt-hours, same run time. Double the voltage x half the current.

With higher voltage, lower current, thinner wires, more watts from the same charge controller.

sarossell said:

Yeah, I looked at forklift batteries, but ironically you need a forklift to move them!

Click to expand...

Nah, one cell at a time.
Maybe a good enough engine hoist?

My AGM bank is 8x 6V 405 Ah, can just move each 120# battery with a dolly. You can get 2V 1200 Ah cells.

I've got a Cat D6 with an after-market backhoe (pretty wimpy) on my mountain property.
But around home I use a harbor-freight 1-ton engine hoist and the appliance dolly.

Hedges said:

Same watt-hours, same run time. Double the voltage x half the current.

With higher voltage, lower current, thinner wires, more watts from the same charge controller.

Click to expand...

Yep, that's what I said, almost verbatim. Sadly, my friend is a bit thick sometimes. He just installed 2,160 watts of panels on his fifth wheel, hooked them all up to a single 60 amp charge controller on a 12 volt bank and he can't figure out why his batteries aren't getting fully charged. I tried to explain to him that his charge controller can only put out 800 watts and he was essentially wasting the rest of the power on his roof. I conceded that such a large array would gather more light in the early morning and late afternoon, but that he was easily ignoring half of his power potential and should rewire his bank for 24v or 48v. He refused because he insisted he would only have a quarter of his amp hours then. Oy.

I still say H2O phase change is the cheapest way to store cooling for the night.
A bit of inefficiency, but PV is cheap.

Hedges said:

I still say H2O phase change is the cheapest way to store cooling for the night.
A bit of inefficiency, but PV is cheap.

Click to expand...

I agree. You've got me thinking in those circles!

Chilling air pipes with ice will be easy. Ice melts and water supports convection.
Freezing thick ice will be the problem. Poor heat conduction.
Many jugs with air circulation would work. But don't want frost build-up from room air, so might do air-to-air heat exchanger inside the (dry air filled) freezer box.

It all sounds fascinating, but each sentence left me more and more in the dark. You lost me at "air-to-air heat exchanger..."

You'll have to forgive me. I was a technical writer in the I.T. industry. I know computers, hardware, software, networks and communication protocols. So, solar and batteries isn't much of a stretch. But heat exchangers? Ehhh, not so much. And don't even think about putting me un der the hood of a car!

I'm curious if a small efficient chest freezer with an air intake on one side, a blower on the other, and "magic coldness" in between would do the trick.

Freezers work well, except when you open the door and let in moist room air, the moisture condenses on the cooling coils, builds ice, and it doesn't work so well any more. So it needs a defrost cycle.

"air to air heat exchanger" - a way to cool your room air without putting room humidity inside the freezer. Maybe run a duct through the freezer and blow room air through the duct. Like the way a furnace separates room air from combustion air.

A chest freezer could work on a small scale, but its total BTU capacity would be must less than what it takes to cool your home.
Determine BTU or watts needed for air conditioning during the 12 hour night. That's how much will need to be done during the 12 hour day.
Calculate how many pounds of ice that will take, to determine tank/box size.

I've overhauled a motorcycle engine, engine top-end and automatic transmission of a Sable. Once took a small engine class in 9th grade and just winged it from there.

But EE is my degree and career.

Mechanical and thermal, other than some classes for my EE degree, I learned them by osmosis and by reading.

I did computers (custom ASIC microprocessors) back when we wrote our own CAD tools. Haven't kept up with the complexity of software and architecture. I try to deal with weird stuff like EMI, electromagnetics, electron "optics" now. I do some coding for analysis and ATE, Matlab among others.

So, fill a top lid chest freezer with a maze of frozen water bottles and never open the lid. Then push air through a circuitous pipe or expansion coil (?) to a blower fan on the other side. As for the nightly BTU requirement, it can't be much since I would limit it to a 100 square foot enclosed space; a cold panic room or Summer fallout shelter, if you will. I have no idea how to calculate the necessary ice and tank size though.

I'm up at night right now taking an on-line class in Central Europe time zone. "COMSOL" multi-physics simulator.
Draw a shape of metal. Current run through it, heating it up, which changes its resistance and how much it is heated. The shape deforms from heat.
Way too much complexity for me, unless it has examples (which it does) and step-by-step tutorials with pictures of what GUI buttons to use (it doesn't).

Yeah, it's after 1 am here. I'm gonna turn in.
You lost me with heat and shape of metal. My brain is mush,
Nice chatting with you!

Something like that, yes.

You can figure out BTU or watts from A/C run time and ratings. Maybe a Kill-a-Watt meter. I measured current draw of a small window A/C. Startup surge was 5x nameplate rating for 0.15 seconds. Running current was 1/2 of nameplate (but it wasn't working hard on a mild day.)

Wikipedia will gives properties of water, like latent heat of freezing. Probably in units of Joules, which are simply watt-seconds. Ice to water transition at 0 degrees C is about the same as heating from 0 to 100, which is about the same as going from liquid to gas.

Things expand when they get hot. Railroad rails could buckle. My steel conduit is OK, but plastic conduit in the sun turns into an "S" shape.

Sleep well! I was going to do Midnight to 8:00 AM but couldn't make it past 4:00 AM. Getting too old. I used to be able to pull 36 hour days doing rote work (slinging rectangles in IC or board layout).

sarossell said:

Then I read the specs and hit another wall in my understanding. It says it can handle from 150 to 250 VDC, but only 18 amps from the solar panels. Surely that's a typo, right? But page 44 of the manual says "Max. PV Input Current” 18 Amps. So, okay 4000 / 18 amps = 222 volts; within the stated maximum input voltage. But how the heck do I combine 270 watt panels with a 38.5 volt in-circuit voltage at 7.6 amps each? The best I can figure is six parallel groups of two panels in series reaching 231 volts at 15.2 amps for a total of 3,511 watts. I can’t parallel more than two panels. That would exceed the 18 amp limit.

Click to expand...

I don’t think this point got enough attention. Quite besides the specifics of the panels you have I’m not sure its safely possible to get the full 4kW solar throughput.

So the PV Array MPPT Voltage range is listed 120Vdc ~ 250Vdc
Max PV Array Open Circuit Voltage is listed 250Vdc
Max PV Input Current is 18A

4000W/18A = 222V
so you would need your MPPT to track a working voltage above 222V to get the full power.
That’s not going to be safely possibly if the Voc has to stay under 250V. You would need panels with Voc less than 12.5% higher than Vmp. I don’t think that’s ever going to happen, panels usually have a Voc 24% higher than Vmp.
And that doesn’t even account for safety for cold weather voltage rises.

So either the MPPT can’t get the full 4KW without extreme levels of over panelling or it can handle Voc > 250V gracefully and it’s not documented.
Speculation: I notice there is both a warning code 13 for “Solar charger stops due to high PV voltage” and a fault code 61 for “PV voltage is too high”. There’s also the point that the MPPT start up voltage is 150 +- 10 Vdc which I suspect means the MPPT won’t start searching at a point over 250Vdc though there may be other reasons to point that out? If this is the case I suspect that the Open Circuit Voltage Limit of 250V only applies at first connection at start up, ie. don’t plug it in with the panels in the sun.
But that’s all very wishful thinking and I wouldn’t act on that unless you’re willing to blow your Growatt up for science or have a trusted source confirming this.

The other option to my mind is overpanelling. (There is no fault code for PV power or current, only a warning code for overpower). If you were to over panel by 50% the MPPT may track a voltage in between Vmp and Voc. Looking at some spec sheets it looks like panels often produce over 66% of their power at a voltage halfway between Vmp and Voc. Which gets you to 12% difference between Voc and output voltage. You’d still be sitting at almost spot on 250Voc and wouldn’t have any margin for cold weather, but you could have your full 4kW in mid day sun, and still get a more broadened power curve as it tracks a voltage closer Vmp in the shoulders making up for the lower current.

The much more likely scenario here is that the marketing is for very optimistic numbers and you shouldn’t expect more than 3000W out of it if you’re operating it safely with cold weather margins. Then maybe in cold weather if you’re over paneled you’ll be able to get a little more out of it for free.

While I agree the numbers don't appear to match up well for my situation, I did receive confirmation from Ian Roux with the company...

I have no doubt you will be just fine at 5S3P on each unit.

Ian Roux
Watts247.com
(801)403-0336
Utah, USA

Click to expand...

The device is smart enough to switch between power sources and snychronously parallel with additional units, so I suspect the device is smart enough to deal with higher available currents from the panels and perhaps has fail safes built in for slight over voltage. I admit, I am skeptical, but he did say "no doubt", so we'll see. I'll just have to be careful I guess.