UPS Supercapacitor Conversion

[joeqsmith]

Only what you have seen posted in this thread.   People seemed more interested in discussing UPSs and battery brands than discussing the actual project.   Longer term, I plan to put something together or possibly buy one.   

[BradC]

Only what you have seen posted in this thread.   People seemed more interested in discussing UPSs and battery brands than discussing the actual project.

Unless I'm blind, you've shown some outcomes, but nothing about the unit itself. I can't even seem to find what UPS you used, let alone which caps or how they're configured. I do know it's double conversion but you've put it in conventional mode.

The outcomes look good, but I have no idea how you got them.

[joeqsmith]

I do know it's double conversion but you've put it in conventional mode.

https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2016721/#msg2016721
https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2207082/#msg2207082
https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2494173/#msg2494173
https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg3101123/#msg3101123

[BradC]

I do know it's double conversion but you've put it in conventional mode.

https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2016721/#msg2016721
https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2207082/#msg2207082
https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2494173/#msg2494173
https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg3101123/#msg3101123

Sorry, mis-read that then.

[richard.cs]

Longer outages are from overhead lines and trees.  As I have mentioned in this thread, these events are fairly rare.

Four a year would be considered more than rare here. I have been in my current house 5 years and have had one. I probably had 5-10 in the house I lived in previously during the 25 years I was there. Short dips, enough to reset a PC or similar maybe 2-3 times per year.

[joeqsmith]

.... let alone which caps or how they're configured.

I didn't see much of a point of posting which caps I procured as I doubt very much people would be able to buy the same ones.  My guess is they are rejects from a larger company that are rebranded.   I was going for cheap to get a feel what is the worse performance I could expect.   

Looking at datasheets, obviously the rated voltage of an SC is much too low to be of use and they will need to be placed in series.  I doubt that is what you are asking.

[joeqsmith]

Pretty much what I am seeing, the leakage current continues to decrease along with the capacitance.   

https://passive-components.eu/wp-content/uploads/2017/10/3.3.-BUT-Kuparowitz-Supercap.pdf
https://www.osti.gov/etdeweb/servlets/purl/20671677

[joeqsmith]

We are quickly approaching three years of 24/7 operation.    The UPS continues to provide more than enough uptime to ride out the short dropouts we see.   

Once I have some time, I will pull it back apart and take some measurements.   

[joeqsmith]

The modified UPS has now been in operation for close to 4 years.   It's been almost 2 years since I looked at the hold up time. 

I checked the balance of the caps and nothing unusual appears to be happening. 

The cheap supercaps continue to degrade.  Using the 300W resistive load, the holdup time has dropped to about 12 seconds.   Still plenty of life left and no house fire to report.   

[madires]

Thanks for the updates! I've retired my classic UPS after the SLA battery pack went bad because the mains became more stable over the last few years. Meanwhile you can also get a UPS with a Li-Ion battery pack. The claimed lifetime is about 7 years which is a bit more than the typical 5 years of standard SLAs, but still not as long as special SLAs optimized for UPS usage.

[joeqsmith]

The amount the caps are degrading does not appear linear.   We started at roughly 17 seconds holdup.
https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg1635044/#msg1635044
It's been over a year since the last time I checked it and we lost about a half second or so was all.  It seems to be settling in.

[flyrod]

Thank you for sharing your project and continued updates!  I live in a rural area with plenty of power problems, and I also got sick of replacing lead batteries all the time.  I switched a couple of our UPS's to a123 LiFePo batteries and these have held up quite well.  With initial testing I used a small 8-cell pack to make coffee, and my espresso machine was pulling around 55A from the pack without issue.  I got the cells used (I think they were made in 2009) and I've had them on the UPS for around 5 years now.  They still have around 85% of the datasheet capacity.  This pack is around 15AH with a 1200w space heater as a test load:



I have two more smaller UPS's that now need replacement 12v batteries, and I'm looking at the capacitor route.  I may have missed it, but what do you use for balancing the pack/array?  Also, how much are you de-rating the capacitors?  Have you tried higher loads on the UPS?

[joeqsmith]

I started this experiment on 6-28-2018.  Its been almost five years now and the house has yet to burn down as a result.  I had some free time and decided to make some measurements a couple of months early.   

...
 I may have missed it, but what do you use for balancing the pack/array?  Also, how much are you de-rating the capacitors?  Have you tried higher loads on the UPS?

Sorry for the year delay but I don't check this thread often.   

I did not care about efficiency and balanced them by measuring the leakage when they were new and increased it using a resistor in parallel with each.   When I measured the voltage drop today they are all still well balanced.   

I had measured the UPS charge and dropout voltage to determine the capacitance required to achieve the hold up time.   It looks like I doubled the capacitance required. 

I have ran it at higher loads but continue to collect the data a 300W to give some idea how the parts are degrading.   I wasn't expecting the hold time to be as good as it is and had to repeat the test.   Looks like about 11 seconds.   Consider we started out with about 17 second holdup, we have lost a fair amount.   Because you asked, I increased the load to a bit over 700W and the holdup time dropped to about 3.7 seconds.   I let it recharge the bank and tried it again with roughly the same result.

As we still have these short dropouts,  I plan to let it continue to run.