20 second UPS?

[islandbayy]

If you use the Powerwall as a power backup solution, you have to contend with the fact that it does indeed deliver backup power ... Starting about 5-10 seconds after the power goes out. So for certain things that you'd rather not have any power interruption for (like your computer, maybe your network gear), you need a UPS.

The problem with existing UPS is that their lead acid batteries last about 4 years. At which point you've got an expensive and messy upgrade procedure. And the UPS is designed to bridge far more than 20 seconds of downtime. Since lead acid batteries don't deal with rapid deep cycling week, you have to get a reasonably sized one.

So, my thought was, I wonder if there is a market for a super capacitor UPS? IF you could build it out of components that would last 20 years, you could build a fairly small and light unit since it would only need to bridge a 20 seconds power gap in a Powerwall (or any generator) household. Thoughts?

FYI, depends on the UPS for lead acid batteries. The batteries in my APC BackUPS 1500 are 11 years old, and powered my powerful desktop computer and monitor + a lamp for 13 minutes during a power outtage a week ago Friday until I could close all open apps and save the video content I was editing (That is what took so long, saving a 80 gig file..... After shutdown, still powered my LED lamp for another hour before it shut off. At that point, I ran a extension cord to my Model S, and used a 600 watt inverter on the 12v terminals under the nose cone to power my living room lamp, tv, for another couple hours to watch netflix until we went to bed.

Cheap batteries dont last long, quality batteries last a very long time. I still have my old 1991 Alfa Romeo 164's starting battery, the original factory installed one, from 1991. I dont use it in a car any more, but I use it to start a 1/4 scale diesel locomotive. still holds at 12.9v Give lead acid more credit, they are very resilient if treated properly.



Why get a super capacitor UPS? I'd say a lithium one would be good. I was thinking of testing a 12v Lithium vehicle battery in place of the lead acids when my current batteries bite the dust.
If charged to a decent SOC, so over--build the capacity in terms of amp hours, but then charge the cells to a lower SOC, they will last a insanely long time. So keep the SOC to about 50-70%, and those cells in standby would last 20+ years. Thats what I did with my first laptop. Those cells are now 17 years old, and as of 2 weeks ago, still had about 90% of the original capacity, and that was 1998 technology. Things have improved since then!

 

[wk057]

At that point, I ran a extension cord to my Model S, and used a 600 watt inverter on the 12v terminals under the nose cone

I'm sure you're aware, but I'll note for those who may not be. The nose cone 12V terminal is fused at 50A I believe. So, 600W max at 12V, before inverter efficiency losses. You don't want to pop that fuse. It isn't the simplest to get to.

 

[FlasherZ]

My lead-acid experience has been nothing like yours, islandbayy. I have 6 different UPS's (former pre-Schneider MGE's and some APC's) distributed around the house; I've tried everything from the cheap chinese batteries to the highest-rated expensive ones, and in each case the lifetime has been 3-6 years for the batteries. The UPS automatic weekly tests scream like hell when the batteries fail.

Likewise, I've never had the luck you've had with automotive batteries, either, whether Exide, Interstate, Delco, or Wal-Mart. They all typically fail just after their warranty expires, except my 1965 Pontiac, which tends to eat them on an annual basis.

 

[wk057]

except my 1965 Pontiac, which tends to eat them on an annual basis.

Weak alternator? Maybe do an EV conversion.

 

[jerry33]

My lead-acid experience has been nothing like yours, islandbayy. I have 6 different UPS's (former pre-Schneider MGE's and some APC's) distributed around the house; I've tried everything from the cheap chinese batteries to the highest-rated expensive ones, and in each case the lifetime has been 3-6 years for the batteries. The UPS automatic weekly tests scream like hell when the batteries fail.

On the small UPS I have, I get about 5-6 years like you do. The big one (6 KVA PowerWare) has batteries that last about eight years.

 

[wdolson]

I have several APC UPSs around the house and the batteries in those tend to die about 2-3 years, even when I get the best quality replacements I can get.

I did get 6 1/2 years out of my original battery in my Buick. The alternator went before the battery. But that is the best I've seen for a car battery. Though lead acid batteries can last a while under the right conditions. The primary reason people are researching better battery technologies is the weight.

A lead acid battery can take more extreme charging profiles than any other battery. You can run them down to zero, the charge them up to 100% and leave them and it doesn't damage them. All other rechargeable batteries have some sort of caveat you have to be aware of if you want a long life.

I wouldn't try putting Li-ion batteries in a UPS designed for lead acid batteries unless the charging firmware was changed too. If you try to maintain Li-ion batteries like lead acid batteries, they will likely have a short life.

 

[FlasherZ]

Weak alternator? Maybe do an EV conversion.

Yeah, as it turns out I just had the alternator rebuilt, it had two bad diode packs in it.

The car is original, still riding on 7.75x14 red-line bias-ply tires. No EV conversions for it. There's something about that 389 with the triple-deuce.

 

[Ampster]

.........

I wouldn't try putting Li-ion batteries in a UPS designed for lead acid batteries unless the charging firmware was changed too. If you try to maintain Li-ion batteries like lead acid batteries, they will likely have a short life.

On the other hand don't leave a lead acid battery a half charged for long, whereas a Lithium battery can maintain that state for years. They each have their pluses and minuses. If you like gardening you may enjoy watering your lead acid batteries as well.LOL

 

[hill]

Some ups outfits utilize a flywheel. Once the RPM's slightly drop the UPS kicks in - preventing any dip in power.

 

[swaltner]

Yep, flywheel UPS units are really cool. We have one powering half of the server room at work (other half by an older lead-acid UPS). I think both UPS are rated at 150 kVA. At 85% of it's rated capacity, the flywheel unit can run for 11 seconds. It takes something like 6 seconds for the generator to start and the automatic transfer switch to switch the input power feed from (now dead) utility power to the generator. It stays running on the generator until utility power is stable.

MUCH less maintenance on the flywheel UPS. You have to replace bearings every 5 years, but that costs about 25% of the cost of replacing batteries in a battery based UPS.

 

[bwa]

Yep, flywheel UPS units are really cool. We have one powering half of the server room at work (other half by an older lead-acid UPS). I think both UPS are rated at 150 kVA. At 85% of it's rated capacity, the flywheel unit can run for 11 seconds. It takes something like 6 seconds for the generator to start and the automatic transfer switch to switch the input power feed from (now dead) utility power to the generator. It stays running on the generator until utility power is stable.

MUCH less maintenance on the flywheel UPS. You have to replace bearings every 5 years, but that costs about 25% of the cost of replacing batteries in a battery based UPS.

I've heard this technique has been around in more simple (and less efficient) formats for even longer: a military base had this type of flywheel going. Big magnets kept on by utility power kept a clutch from the diesel generator separated from the flywheel. When utility power ran out, the magnets released, spring-loaded clutch slams into the flywheel, suddenly rotating the diesel generator into 100% speed. Zero downtime on the user side --- just maybe a little wavy waveform during the transitions. I can't imagine how bad that is on that diesel without the oil in the right place -- I could think of ways to improve on this system. But the diesel would just start running instantly. Everything was fed from the output generator side of the flywheel to the military base. Lots of inefficiencies and hell on the generator's longevity, but it's another example of flywheels being simple, old-fashioned, and supplying 100% of the rated power 100% of the planned-for time with little or no unplanned downtime. Flywheels are one of the old standards to meet and beat, as far as I'm concerned. Any complete system that can't meet the standards of a flywheel are just cheap(er), in my opinion, and have made enormous sacrifices.

These days we can be a LOT smarter than the military base flywheel system I described. I just wanted to point out they've been around in service supplying superb uptime guarantees for longer than most people have been alive, with zero cutout in unplanned outages.

I am left to wonder what happens in brownouts that don't fully release the magnets. Obviously any system needs extensive debugging and scenereo testing. I think one of the worst mistakes made by UPS implementors is half-assing it; UPS's are meant to be whole-ass, whole-solution, all-worked-up items, otherwise, what's the point? Proper engineering, testing, and maintenance, all the way, is what makes a UPS worth it, as far as I'm concerned. I've read about too many UPS's causing more outages than they've solved. In bad cases, they actually CAUSE fires, and due to the half-assing policy of their installations, the fire response systems were on the electrical system protected by the UPS, which of course went out when the UPS caught fire, so everything burned and no one knew about it. Proper design, everyone. Proper design.

 

[Ulmo]

If you use the Powerwall as a power backup solution, you have to contend with the fact that it does indeed deliver backup power ... Starting about 5-10 seconds after the power goes out. So for certain things that you'd rather not have any power interruption for (like your computer, maybe your network gear), you need a UPS.

The problem with existing UPS is that their lead acid batteries last about 4 years. At which point you've got an expensive and messy upgrade procedure. And the UPS is designed to bridge far more than 20 seconds of downtime. Since lead acid batteries don't deal with rapid deep cycling week, you have to get a reasonably sized one.

So, my thought was, I wonder if there is a market for a super capacitor UPS? IF you could build it out of components that would last 20 years, you could build a fairly small and light unit since it would only need to bridge a 20 seconds power gap in a Powerwall (or any generator) household. Thoughts?

Is that PowerWall version 1?

I have PowerWall version 2 which shouldn't do that.

I came here because my Mac shut down between 3:26 and 3:27 this morning when my PowerWall 2 system switched from using battery to using grid. Something about the inverters failed or made dirty power in such a way my Mac could not handle. This is about the 3rd or 4th time this has happened, but I think the first time it happened during a normal grid switchover at low power. Does anybody else have experience with that? We only had between 500 Watts and 600 Watts usage at the time.

I'll fix my grapher and post a better graph tomorrow.

Update 1: The other finicky sensitive device in the house that doesn't like very momentary drops, the gas-powered stove with electric controls, also is showing that a power interruption occured.

My server computer usually makes it through these momentary outages; it must have more capacitors than my Mac.

Keywords: momentary power loss, power failure.

 

[davecolene0606]

Is that PowerWall version 1?

I have PowerWall version 2 which shouldn't do that.

I came here because my Mac shut down between 3:26 and 3:27 this morning when my PowerWall 2 system switched from using battery to using grid. Something about the inverters failed or made dirty power in such a way my Mac could not handle. This is about the 3rd or 4th time this has happened, but I think the first time it happened during a normal grid switchover at low power. Does anybody else have experience with that? We only had between 500 Watts and 600 Watts usage at the time.

I'll fix my grapher and post a better graph tomorrow.

Update 1: The other finicky sensitive device in the house that doesn't like very momentary drops, the gas-powered stove with electric controls, also is showing that a power interruption occured.

My server computer usually makes it through these momentary outages; it must have more capacitors than my Mac.

Keywords: momentary power loss, power failure.

Hey There;

If I shut off my main breaker the load transfer happens faster than any clock or computer in the house can recognize. Having said that, a while back we had a pretty wicked Tstrm and a very rough loss of power that had a surge followed by loss of power. In this case I had the oven clock go to power fail mode and the 2 UPS units tripped for about 10 seconds or so.

My guess is that the "smoothness" of the reason for loss definitely is factor. I kept the UPS units on the computers/routers/entertain system for this reason.

Cheer!

PS 2 PW2 with Sunnyboy Inverters

 

[eml2]

I have 2 iMac's and never experience any problem when house load switches from PW to grid or vice versa.

 

[miimura]

If the Powerwalls are in Standby then they will take longer to kick in and you will probably have a momentary outage. However, if you are running Self Powered all the time, then they are either charging or discharging with the grid not doing much, so you will not have an interruption. I am still running 3 UPS units. One on my server, a tiny one on my DVR, and another big one for my structured wiring cabinet (cable modem, ethernet switch, etc) and my home office.

 

[strangely]

I've switched a few time from grid to powerwall (2.0) using the main breaker, and the cutover is pretty quick. The only devices that I have that seem sensitive to this are a few z-wave smart bulbs that default to on after power is removed. I did this yesterday once the powerwall was already discharging in self consumption mode and it didn't even effect the lights when I flipped the breaker.

 

[cwied]

If the Powerwalls are in Standby then they will take longer to kick in and you will probably have a momentary outage. However, if you are running Self Powered all the time, then they are either charging or discharging with the grid not doing much, so you will not have an interruption.

I have one data point contradicting this. In the one real power outage I've experienced, my Powerwalls were discharging. Despite this, all my equipment reset (including my routers, which generally are fairly resilient to power issues). I think that how quickly the Powerwalls kick in is only part of the equation. How cleanly the grid goes down probably also matters. It's apparently not always as clean as flipping a breaker.

Note that my solar inverters shut down several times in the days prior to the outage, so the power was already abnormal before the outage (but not enough so to cause the Powerwalls to take over).

Edit: one other addition: the small in- and outflows when in "standby" when in self-consumption or time-based control modes lead me to believe that the Powerwall inverters are always running during those times. I haven't watched backup mode as much, so I guess it's possible that the inverters do turn off in that mode.

 

[mspohr]

Powerwalls can act as a UPS but it doesn't seem that you can rely on them for "instantaneous" switchover.
If you have sensitive equipment such as computers and network hubs, it would probably be best to install it behind a UPS which is designed for rapid switchover.

 

[BJReplay]

My experience has been if the PW is discharging into load, it is seamless. If it is charging from PV or on standby (so the inverter isn't running), it takes longer, and some devices may notice the momentary loss of power.

 

[cwied]

My experience has been if the PW is discharging into load, it is seamless. If it is charging from PV or on standby (so the inverter isn't running), it takes longer, and some devices may notice the momentary loss of power.

Was that when testing by switching a breaker, or was that with a real power outage due to downed power line or such? My Powerwalls were discharging and I still had an interruption. Unfortunately I wasn't home to see exactly what happened, but my routers rebooted and many clocks were reset. I suspect that the power was glitchy before finally going out and the Powerwalls didn't take over until the power was completely out.