Looking for a home UPS that is smart

[adder]

Yes thats the plan, if I where to use the batteries in the night generously. I would save about, 2k a month(with limited geyser running), more so in summer.
Once I get a electric scooter, I will save even more. Will also add a solar water heater tank only (no collector) and use solar electricity to heat water instead of solar thermal. That way all can use solar hot water. Presently only one or two person can use stored hot water in the morning( no solar generation in morning, with one or two standard 25l water heater, which would have heated the water by 4pm, using solar electricity).

Also in summer the top floor ceiling temperature is 34c to 36c even at 10pm, outside roof temperature during day hits 55c to 58c. By 10pm the outside ambient temperature is 24c, yet the room is very hot because of the radiating heat from the ceiling. Now when I add solar panels the roof will get shade(not all of the roof) and could also run the AC during day to neutralize any heat the ceiling may have gained by the sun.

Is it worth it, well in my case I had to buy new higher capacity batteries any way and needed another inverter for the down floor.

Sure the solar inverter increase the cost by another 50%, but any other normal inverter needed 8 batteries. Only these solar inverters have such a high capacity and they still work with just 4 batteries. The power factor of this is close to 1 much higher then others.

Again all the calculatuons of saving and uses are all theory, I will only know truely how much I will save after installing them.

In order to disconnect bescom, I would need 400ah x4 of lead acid battery or 220 to 240ah x4 of LFP or 200ah of LTO battery.
btw a 1kw panel would never give the full 1kw in indian temperature. At best you will get about 700 to 750w.

 

[NotMyRealName]

This is a really interesting and unique solar panel system:

https://blog.pickmysolar.com/smartflower-solar-comprehensive-review
https://www.sunlux.com/blog/smartflower-solar/

Looks costly though, as expected...

 

[blr_p]

These guys have the best videos on the subject of costs.

Lifestyle reasons

I'm amazed with the inverter (notice the lack of any VA crap, only watts) and 20kWh+ battery bank how careful they still have to be.

Batteries they use. SimpliFi

$2900 for 3.5kWh. They have 7 so $20k or 14 lakhs just for batteries

Comes with a 10yr warranty. 10,000 cycles at 80% DOD. Lithium phosphate

 

[adder]

That's a lot of money for inverter $2200 for only 4.4kw, plus the cost of those batteries, now the Tesla 13.5kw pack with 5kw inverter powerwall sounds cheaper compared to those.

 

[blr_p]

WiFi configuration is before the OTP which is the last step, so yeah that completed successfully.

At app. first start post install, it scans for the Zelio (which I guess is in AP mode), router WiFi configuration is entered into the app, the app validates the WiFi configuration, pushes it out to the Zelio, then disconnects from the Zelio WiFi AP and connects to the router WiFi. Next, a QR code on the invoice is to be scanned and validated which also completes. The next step is to enter a mobile number and wait for an SMS OTP from Luminous that is terribly, slow/unreliable in getting across. Once the OTP is received, logon from the app. using the OTP and wait for the app. to attempt to connect to the Zelio. That's where I am stuck at with a "communication failure" error.

The manual says

A month later have you got anywhere ?

 

[blr_p]

I am not sure how luminous can make sensor work, the batteries vent holes have various diameter and thread pitch, infact they deliberately change the design, so that once the vent cap/floats break you are forced to buy a new battery.

My guess is it is only designed to work with luminous model, mostly for non tall tubular batteries which need filling in 2 to 3 months. Since the tall tubular batteries needs to filled once in 6 to 10 months depending on the brand and model.

It's doing a simple load test

 

[blr_p]

That's a lot of money for inverter $2200 for only 4.4kw, plus the cost of those batteries, now the Tesla 13.5kw pack with 5kw inverter powerwall sounds cheaper compared to those.

Inverter can handle a surge of 8000W. I suppose the price goes up depending on the quality of the transformers ?

This next video clarifies things better. Since everything was on one circuit they had to turn things off. Nothing to do with the setup

 

[adder]

It's doing a simple load test

I guess it's similar to what APC does, which has a battery runtime calibration which is nothing but a load test. Apc, Luminous now owned by Schneider.

Inverter can handle a surge of 8000W. I suppose the price goes up depending on the quality of the transformers ?

My inverter can handle 10000w for 10sec, the powerwall does 7000w peak for 5 sec.

There two types of inverter one that uses a lot of IGBT which is transformer-less and all the bulky ones which use a transformer(more expensive). In my inverter and all other compact wall mount based ones are all transformer-less.( they still have small tiny transformers like inside your smps) just not a big one like you see in those regular inverter/ups.

Also, all this while I assumed the Tesla powerwall 2 could run directly of solar panels, but it actually is just a inverter with li-battery. So you need additional $1000 to $2500 for a On grid solar inverter and $6000 for powerwall and additional cost for the panels. So it makes sense now why people who have Tesla powerwall sold in their countries still choose to go with custom installations.

This next video clarifies things better. Since everything was on one circuit they had to turn things off. Nothing to do with the setup

Those off the grid people need to size the batteries and inverter capacity correctly, since where they choose to live, like in the woods they don't have the option of Grid electricity.
Also most people in the west still use incandesent bulbs and other inefficient devices. They also get hot the water for every tap, so they need a larger water heater and whole house hvac.

In my case the moment the load is more then what the inverter can give with solar and battery combined, it would rivet to utility.

Also in the past few weeks, I was isolating the house wiring for non essential heavy load stuff to utility only, they will not be powered by the inverter.

 

[blr_p]

Thinking of a way to record battery performance and the idea i had was to attach an electricity meter to the output of the inverter. Electricity meters are cheap, measure RMS and easily availble. Downside is these meters cannot be reset to zero for the next battery. So i found something else.

This meter has a reset button and keeps the energy used KWh in memory. It will retain that value if the meter is powered off the other values are real time only. Says it can be used for sine wave only and anything else will end up causing it to catch fire. So to use with pure sine wave inverters only.

The idea is when a battery is new to measure how many units are delivered throughout its life. Along with load tests we get an idea of how the battery is performing over time. The meter should only come on when the inverter is working and remain off when mains power is present.

Take a 100Ah battery, should put out 100x12 or 1.2kW for a full cycle. This is at C20 discharge rates. Faster discharges like C5 will deliver less total power.

If its life is estimated to be 500 cycles at 80% DOD then we can expect a theoretical maximum service life of 0.96*500 or 480 units.

There will be many partial cycles that in aggregate will add up

The idea of having a meter count units throughout the life of the battery gives an idea of its actual throughput up to the point of replacement. The additional info can be useful to gauge what load is placed on the inverter at any time.

 

[adder]

So for that meter to work, you have to do a load test but the problem with that is capacity varies depending on the load. Also how will the meter know if the sinewave coming out of the inverter is via On battery or Mains Bypass. You will have to reset it each time there is a power cut to start the meter.

 

[blr_p]

So for that meter to work, you have to do a load test but the problem with that is capacity varies depending on the load.

Could we approximate partial cycles into one cycle.

Also how will the meter know if the sinewave coming out of the inverter is via On battery or Mains Bypass. You will have to reset it each time there is a power cut to start the meter.

This is a harder problem to figure out. Both input to & output from the inverter will be live :/

I don't want to be doing any resetting each time there is a power cut.

 

[adder]

Could we approximate partial cycles into one cycle.


This is a harder problem to figure out. Both input to & output from the inverter will be live :/

I don't want to be doing any resetting each time there is a power cut.

AC meter just won't work practically, which is why I am looking at battery monitoring which can measure DC, both during charging and discharging. Even these are only accurate for discharging not for charging because even if you put say 10amps of charge to the battery for 1hr, the actual capacity the Lead acid battery would have got is anywhere between 6ah to 8.5ah depending on how efficient the lead acid battery is. Lead acid battery are more efficient (85%) when its in depleted state during charge cycle and are of poor efficiency(60%) when its capacity is close to full.

But measuring discharging is what matters for knowing the capacity. But those victron energy battery monitor that I posted in the previous page, do have a formula to compensate for the lead acid efficiency.

 

[blr_p]

But measuring discharging is what matters for knowing the capacity. But those victron energy battery monitor that I posted in the previous page, do have a formula to compensate for the lead acid efficiency.

The idea was to have a simple units counter like a miles counter on a car. How many units has the battery delivered in its lifetime.

I guess unless a way can be figured out to distinguish between mains power and battery output all that meter will do is

1) measure load on the circuit in watts since these smart inverters don't provide that value instead give just a %. It would be a secondary check in that sense. This would also act as a monitor during load tests.
2) what the mains voltage is at and how it varies if one cares to look
3) the power factor on the load. A low PF will need more power and yield less run time than expected. Clamp meters don't show PF. The Mastek I have does not.
4) how much the UPS draws by itself with no load

Interesting data but not quite the whole package i was hoping for. Think I might still get it. It will only be on the lighting circuit. So anything on the heating circuit will be missed.

Maybe i need another for that for the whole house. Will need the 100A variant. One linked earlier only goes to 10A and does not include a current transformer (CT).

Whether I have to look for another model that shows active power as that's what we get billed for and what the inverter has to supply instead of real power.

 

[blr_p]

I think Regal as a LFP battery but doesn't have the cycle life of modern LFP cells like those found in chinese EV vehicles.

Was trying to find the video ad where i saw the regalia lithium chemistry mentioned and found it.

There. That video came out in Jan 2017. Luminous never again mentions the chemistry in their brochures or subsequent ads. I suppose three modular batteries refers to the battery sizes of the three variants

So the next question is whether such a battery can be found in the capacities Luminous mentions and for a suitable price. I find such a candidate on alibaba. So the price would fit in if bought in bulk.

What is interesting is how closely the controls on this battery match the regalia.

Looks like a dead ringer. So i started looking at the company's website and there is more info there

That's what it looks like inside. Uses prismatic cells. What about battery performance

I guess this is as good as it gets as Luminous isn't going to release this info. Not quite as good as the graph i posted earlier shown below but we're getting 2000 cycles by the time it gets to 80%

Am assuming the discharge is at 100% DOD as the preceding graphs show discharge ending at 2.2V. That is just a hair above the minimum allowed charge of 2.0V

So the 7 -10 yrs they're claiming is some what plausible. Too bad Luminous only offers a 5 yr warranty on the battery instead of longer. The phosphate chemistry means its the safest of the lithium chemistries to use indoors.

 

[vivek.krishnan]

@blr_p As usual, someone in Luminous was tasked to design something, and he just took a vacation to China and brought their design

 

[adder]

^ Actually pretty much every company selling batteries in India, just rebadge battery designs from china. At best they do a custom enclosure with all the guts from chinese battery pack.
There are plenty of no name cell suppliers in India, price is Rs18/watt hr+ gst. I believe its 4.2v cell li-on, to get watt hr=volt*Ah.

Those with charge cycle of 2000 or more will start about 25k+18% gst for 12v 100ah equivalent LFP battery.

As far as alibaba suppiers, without a shipping agent it's impossible to buy. Once main stream manufactures like exide and amaron start making batteries,( buying cells from china and making them into pack) the prices will fall because at the scale at which they buy. But then again the prices I mentioned are from small import suppliers not from dealers. With exide after dealer margin, the price will be high.

 

[blr_p]

There are plenty of no name cell suppliers in India, price is Rs18/watt hr+ gst. I believe its 4.2v cell li-on, to get watt hr=volt*Ah.

Those with charge cycle of 2000 or more will start about 25k+18% gst for 12v 100ah equivalent LFP battery.

This is the part that was troubling me. 4.2V cell implies traditional Li-ion, could be manganese oxide (LMO) or nickel manganese cobalt (NMC) which have fewer cycles and are not as safe as phosphate. LMO & NMC are lighter and give more energy per kg than LFP. They would be preferable in electric vehicles. Stationary apps like UPS aren't concerned about weight which LFP provides less energy per kg.

The phosphates run at a lower voltage of 3.65V at full charge but voltage alone isn't enough to identify whether it is LFP

One of the comments by Luminous for that video said the following

It charges up to 54 V and when batteries are fully charged maintained at 52 V.

Now, could you guess what chemistry is used just by that statement alone ?

At full charge of 3.6V times 15 gives 54V for fifteen cells in series

I don't know what maintained at 52V means. Nominal voltage should be 48V not 52

If there were 16 cells in series then as the Chinese manufacturer indicates in the specs, nominal is 51.2 with a max of 58.2V

 

[blr_p]

how will the meter know if the sinewave coming out of the inverter is via On battery or Mains Bypass. You will have to reset it each time there is a power cut to start the meter.

Here is one tentative solution. Use two power meters. One at the point where the mains enters the house and another connected to the inverter.

Electricity meters don't run backwards when the inverter comes on. Neither do they add units to the electricty bill when the inverter does come on.

But the meter hooked up to the inverter will count units when mains is on and when battery comes on.

So there should be two different readings at the mains entry point and at the ups point.

The difference in units indicates battery units ?

This is more involved than i want but maybe its workable.

 

[adder]

That 52v is done for prolonging the life of the battery which for a LFP pack is 90% SOC. Say when you charge that regalia lithium pack at 20Amps or 1c the moment the pack voltage reaches 54v, at that point the capacity may be around 80% charged, the charge controller then slowly starts to ramp down the amps in order to maintain the 54v. But if you want to increase the cycle life of lithium battery be it 4.2v or a 3.6v chemistry you never charge it to 100% repeatedly, so you stop it at 90 or even 80% SOC. In this SOC the rest voltage will drop to about 52v.

After me spending some $300 or more on LFP batteries over the past decade I can tell you that never charge them to full 100% SOC, unless you are absolutely sure that you are going to discharge it very soon. my first two prismatic 4 cell packs puffed abd bulged to 5x size in 1 year and but my 5 cell LFP pack are also now bulged to 5x the original size in 3 year time despite me storing them at storage voltage, the quality of those packs cannot be compared to mainstream use battery manufactures now, but my 2 Lifepo4 cylindrical cells are still working. As are all my 18650 based Li-on batteries.

 

[adder]

Here is one tentative solution. Use two power meters. One at the point where the mains enters the house and another connected to the UPS.

Electricity meters don't run backwards when the UPS comes on. Neither do they add units to the meter when it does come on.

But the meter hooked up to the UPS will count units when mains is on and when battery comes on.

So there should be two readings at the mains entry point and at the ups point.

The difference in units indicates battery units ?

This is more work than i want to put in but maybe its workable.

The difference reading could work, but cumbersome to go take reading each time.