Home UPS/Inverter calculation question 5

[edison123]

I am planning buy a home UPS/Inverter. The total load is 1.2 KW / 1.5 KVA, a mix of a large fridge, LED tube lights, ceiling fans, TV's, laptops and modems. But all the on-line calculators (from the suppliers) recommend a 3.5 KVA inverter. Any particular reason for this more than the twice rating of connected load?

Muthu

http://www.edison.co.in

 

[itsmoked]

They typically worry about in-rush loads. See if their calculators drop the recommendations without a refrigerator. A 200W refer can draw 15A starting for a considerable period if they fail to start on the first try. That would drop out or trip off the inverter if all the other loads were present causing them phone traffic.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[davidbeach]

To get the specified load/duration point? You're interested in both a kW number and a kWHr number and you want to hit both. If the system only has published ratings for one of those values it may need to "over-achieve" in order to hit the other number.

 

[edison123]

Thanks Keith. If I drop the fridge (500 W), they show 1.5 KVA. That is still 2 times the rest of the load of 700 W (sans the fridge). I was thinking 2 KVA inverter for the original load of 1.5 KVA (with the fridge) to account for the inrush loads. Too optimistic?

David - Isn't KWHR number more for the batteries AH than for the inverter itself?

Muthu

http://www.edison.co.in

 

[itsmoked]

Muthu; There's a lot to getting this right - solid. Do you realize the hoards of batteries supporting all that stuff will require? I'm putting one in myself but decided to run two separate outlets in the house next to some original outlets. They'll get to have cool red covers! One at our... "data center" to keep the cable modem, router, and desk un-interrupted and one in the bedroom for a CPAP machine.

I was going to run an outlet to the refer too but decided that was a bad plan because the other two functions were more urgent. If the power fails our over-packed refer would stay cool all night without the door being fanned anyway. I would pull it out a foot and feed it an extension cord from the data center if I decided things were getting urgent. This way I believe I can husband my battery stores successfully. I happen to be using two new 8Ds I came across that are solar charged (see other thread).

Anyway, unless you're actually building a power-wall type thing that has 10kWhrs of storage, including things like lossy fans, etc will get you dark again pretty fast.

If, on the other hand, you're trying to survive frequent short blinks and you're willing to go dark on real power outages that's a whole different situation.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[edison123]

Thanks Keith. I talked to one of the suppliers about overrating the inverter. David was right. The KWHrs does matter since they do DC input of 24 V (2x220 AH batteries in series) for rating up to 1.5 KVA for 4 hours only. Apparently 220 AH battery max is only available commercially around here.

Since I wanted 8 hours standby, the VAH doubles to nearly 12000, they need 4x220 AH batteries in series to meet my requirement. At 48 V DC input, the minimum size of inverter available is 3.5 KVA and hence their recommendation. The inverter cost from 2 KVA to 3.5 KVA upgrade is only 40% more, but the batteries are doubling the cost.




Muthu

http://www.edison.co.in

 

[itsmoked]

Good to know. Hope to hear how it all pans out, what worked well, what didn't.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Parchie]

Muthu,
Please remember that your inverter would be delivering the needed power to charge its batteries plus supply its load after each long utility black out! That's one of the reasons we double the inverter capacity when buying one.

 

[DanEE]

I've been running at home for years a slightly smaller unit than your requirement, a BEST FERRUPS 1.4kVA/1.0kW unit, factory configured for external batteries, 15amp internal charger option and cord and plug connection. This unit is spec'd to handle a 150% surge (no duration given) and 125% for 10 minutes if online and 110% if running on inverter. Needless to say on a hard surge (for example the inrush from starting a motor too large for the unit) shutdown is extremely fast, I believe faster than typical magnetic breakers which open very quickly on overload.. Another major reason for installing this unit is surge protection, spec'd at 2000:1 with the ferro magnetic transformer based design.

As noted the big costs are the batteries, and I'm pretty sure I have a noticable hit on the power bill as this older technology unit is only in the mid 80% efficiency range, with constant load on it.

What helped with my costs for many years was working in the commercial UPS/DC power plant business. The UPS unit came from a customer site that needed to upgrade and good used batteries (for many years) came from a few customer sites that due to extreme availability requirements meant switching out 10 year design life batteries every 3 years. The rather old pic below shows 6 6v batteries strapped for the 12v battery bank operation. We've had week long power outages and by dropping load back, it had sufficient batteries at the time to carry the load for 3 days; This is a nice system to have even if you have a generator.. With the flexibility it added, we would only run the generator around 2 hours in the morning, and then in the evening for (all electric) well water, showers, hot water, cooking and battery charging.. except for the trusty wood stove for heat.

Not pictured (and getting off topic) is another battery bank and outboard charging system to add more capacity (have had up to fourteen 12V 70AH batteries in the system). In addition to the long run time these were used to do some long term battery charging methodology studies over a 15 year period using charging methodology similar to as described in this Eaton video

https://www.youtube.com/watch?v=Zxn0-50TWwk

and most critical, a relatively constant temperature environment no higher than 68 deg. It yielded a 10-14 year battery life out of most of the 10 year design life batteries in the study. Have dealt with a number of larger UPS systems with integrated battery banks get barely 3 -5 years life out of 10 year design life batteries due to the UPS heat cooking the batteries.

Most of the in house electronics (large screen TV, network equipment, computers, DECT phones, etc), a subset of cord and plug connected lighting (sufficient for power outages) and a rather sizeable security system are powered from this unit. It powers an Article 702 (US NEC) - Optional Standby System branch ciruit with marked UPS outlets at multiple locations, installed next to selected original house outlets as itsmoked did. Nice to be able to use the computer to log on to the power company website and report a power outage!

One challenge I ran into when I originally installed this system was trying to find reasonably priced, appropriately rated (both ampacity and UL listed) hardware for the battery banks. The commencial market stuff is just too big and incredibly expensive so I went with good marine application low voltage, suitable ampacity UL listed stuff. Note the marine battery disconnect switch. I also wanted more interbank battery fusing than we even had on commercial sites. It makes nice disconnects for testing and swapping out batteries.

 

[itsmoked]

Wow Dan. Nice! Is this a full time conversion style? Wait. Looking at the green/grey picture no, it's a standby.

Your batteries are enclosed? No worries about gas buildup?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[edison123]

DanEE

Thanks for that detailed post. The most guaranteed battery life in India is 3 year replacement +1 year service warranty. Bummer.

Now the next task is how to convert the existing 3-phase load for the single phase supply from the inverter. I was thinking of a 4 pole (switched neutral) manual changeover switch that will short the 3 line terminals of the load + N for the inverter in one position and keep the 3-phase + N for the utility supply in the other position. How do you guys think I can possibly screw that up?

Parchie

Good point about the battery recharge capacity and time. I will check with the supplier.

Muthu

http://www.edison.co.in

 

[DanEE]

Itsmoked,
Regarding conversion methodology, ferro-resonant based UPS systems are a somewhat unique UPS design.. When it is online, it operates pretty much as a normal transformer, but with the addition of two additional windings (actually has more for powering control/electronics) on the transformer, first being the winding connected to an external high voltage capactor to form a 50 or 60Hz resonant tank circuit depending on model and country of application, this being common to all ferro resonant transformer designs. (Pretty good write up of ferro transformers here

http://www.electroncoil.com/ferroresonant_transformers.php

) This is operation is somewhat analogous to a flywheel based mechanical MG set..

The tank circuit has sufficient energy storage to ride through short dropouts within a cycle and also has part of the transformer core in saturation that prevents spikes on the incoming primary from getting through to the secondary powering the load. That's where the 2000:1 (2 thousand volt primary surge will pass 1 volt onto the secondary load. Also the energy storage in the resonant LC circuit has sufficient duration to allow the electronics to detect loss of input power and get the inverter online to pick up the load without passing any dropout to the load. The inverter does not drive the load directly but powers a winding on the ferro transformer with the inverter supplying an essentially square wave input, that the tank circuit shapes into an almost pure sine wave envelope. The peak of this sine wave is a bit broader than normal but otherwise pure enough to not have significant harmonic content.

These systems originally designed and built by BEST products out of Necidah, Wisconson and now a part of Eaton Powerware out of Raleigh, NC USA

http://powerquality.eaton.com/About-Us/BestPower.asp?cx=3

were bullet proof in reliabilty and performed well in high stress environments. BEST built model variants meeting additional mil-spec requirements

http://pqlit.eaton.com/ll_download_bylitcode.asp?doc_id=4761

for US Navy using these systems aboard ships to protect and provide back up power for critical electronic systems. While poking around the internet for some US Navy & BEST application articles found in the past, I found this study; a picture in the study shows a BEST UPS which one of two units tested was the 1.4kVA unit I have.

https://www.google.com/url?sa=t&rct...565vbAhUwxVkKHY96ACwQFjAHegQIARBI&url=http://

http://www.dtic.mil/get-tr-doc/pdf?AD=ADA411197&usg=AOvVaw3FAvowLvIV9_YtfYPCNsJR

The down sides of the systems were they are very heavy due to the scale of the magnetics inside and also not the most efficient (88%) compared to newer technologies.

Regarding the battery enclosure, the plastic sheets are set out from the steel frame (doesn't show up well in the photo) to leave gaps arround all sides for sufficent air circulation. I did use flooded cells in the early days (no covers on the rack then), but since have used sealed AGMs and added the covers later.