Using Ultra Capacitors To Extend Battery Life

Hi Tom,

Thanks for the response.

I went ahead and ordered these to do some testing.  They look like the answer to a lot of problems in the off-grid world.

Yes it is true that inverters have capacitors in them.  My goal though was to shield the batteries from seeing high current bursts.  The caps in the inverter are there to assist it without any real regard for the battery.  The battery is still expected to come up with these massive currents.  These caps are going to be physically larger than the inverter itself and much higher quality/capacity than anything inside it.

When the fridge goes to start there is a huge high current burst that the deep cycle batteries are forced to produce.  I can tell by the voltage drop that the batteries don't really like this to happen.  In my studies of how lead acid batteries work this is one of the main things that causes short battery life. 

I know that the batteries/inverter are capable of producing this current however it comes at a huge cost in terms of charge longevity.  It is a unsustainable chemical reaction that causes a runaway between voltage/amps.  The more amps the inverter draws the more the batteries sag which causes even more amps to be needed.  This is especially true when the batteries are close to 50% charged.

http://cgi.ebay.com/Maxwell-2600-Farad-2-5V-Ultracapacitor-EV-used-Qty-12-/300561259628

Those are the capacitors I went ahead and ordered. 

The article that I referenced in the first post is my reasoning for doing this.  These capacitors came out of a hybrid bus.  They are used to provide the burst current needed for the bus to accelerate from a dead stop.

There are videos on youtube of people starting large diesel engines and car engines from these caps.

There are even videos of go carts totally powered by these caps.  I didn't even know they had caps this large until I ran across them by chance.

I am going to be doing a lot of research and I'm going to post all my finding in a scientific way.  I ordered a 500amp shunt and I'm going to be using data logging software to graph the difference in amp draw seen by the battery.

If I can "smooth" the current draw seen by the battery by 25% (or more) I should be able to achieve a lot longer run time on my battery.

Sorry for the long winded reply.  I have been really enjoying my solar system and learning about all this.

Travis

This is an interesting post, Travis, and it caught my attention.  A couple of practical factors also occurred to me.  Since the operating voltage may not exceed 2.5 volts for each capacitor, If you are trying to match them to a 6 volt battery for example, you would then need to place 3 capacitors in series before wiring them in parallel with the 6 volt battery, so that each capacitor doesn't see much over 2 volts across its terminals.  As with any series capacitor arrangement, equalizing circuitry must be used to maintain an equal voltage across each capacitor.  This is done simply by wiring a resistor of equal value across each capacitor, or by some other more complex means (the sophisticated control circuitry mentioned in your first link).  Resistors represent a power loss since they would tend to drain the capacitors (and battery) without any other load on the cells.  And, when you place three capacitors in series, the net overall capacity is reduced to one third, so the 2600 farad rating would drop to 867 farads at a rating of 6 volts......for a 48 volt system with 24 capacitors in series, the rating would drop to 108 farads of capacity. At this reduced capacity, would they actually help the batteries with surge current problems, or just add a lot of expense and complexity to your system?  I'm not trying to rain on the parade, but rather, just expressing curiosity at this concept.  Jon C.

Travis.  Any resistive equalization would represent a power drain on your system and would be counterproductive in my opinion.  You have already seen a sample of the kind of fire you are playing with when you dropped the wrench across capacitor terminals, but do you realize the worst case scenario......The 2.5 volt rating on the capacitors is an indication of the dielectric strength of the separator material in the capacitors, and if this voltage is exceeded, the capacitor may experience a separator failure which usually means instant death for the capacitor.  Capacitors do not die quietly however, and discharge all of the stored energy instantaneously along with fire and smoke.  If one capacitor in your string shorted, the next ones would see an elevated voltage and would not be far behind, setting off an avalanche effect through the string.  You would see a virtual lightning storm with explosions and the likelihood of intense fire.  I have seen lead acid batteries explode as a result of heat igniting the hydrogen gases, and they are very messy.  So, that being said, I wish you success with your project, but you may want to consult an expert about this.  My knowledge about this is very limited, but I do know about smoke and fire.  Jon C.

I cannot resist posting again.  There are DEFINITE increases in the abilities of the solar system.  I am currently sitting at 14.45 volts.  If I key up the radio on high power previously the voltage would do a nose dive all the way to 13.2'ish.  The capacitors are doing their job and during short transmissions the voltage isn't dropping.  This means a few things.

1> The battery is still charging even while I transmit

2> The solar panels are going to put out more since they are closer to their Vmp rating.  I have a PWM charge controller so this brings me a little closer to MPPT efficiency.  Actually I think I am over MPPT efficiency because the battery is still seeing nearly the same charge during short cloud cover.  Lead acid batteries seem to resist charging and capacitors don't.

3> The battery will last longer (more daily capacity).  Lite usage while the sun is out isn't going to discharge the battery.  I will be able to make transmissions without my batteries even knowing I'm doing it.  They are even still charging while I transmit.

4> It seems to make the charging more efficient.  I don't have any scientific proof however when the sun goes behind a cloud the voltage is lingering high almost staying flat during short cloud cover. The battery is still getting a charge even with dropping solar voltages.  The caps are holding the voltage higher.

Obviously a capacitor isn't going to help with steady continuous loads to this degree.  But for my application I can already see a HUGE benefit.

My radio draws 114 watts during a high power transmission.  My panel is a 110 watt panel.  To me this is the true test of how well this is all working.  Previously on transmit the voltage would fall to the batteries working voltage and I would bleed off the surface charge almost instantly.

Now if the capacitors can stay balanced without maintenance then this is the PERFECT setup. (So far they are staying balanced)

Version 3 of my capacitor bank.  I decided I needed to beef up the connections between the caps.  I used some aluminum stock that I picked up at home depot.  I also used a sheet of plexiglass to hold the cells together.  This seems to work wonderfully.



I still haven't gotten my shunts so I still don't have any scientific testing done however at the end of the night my voltage is much higher than previously.  These are making a huge difference from what I can see so far.

I have 12 more of these caps on order and I'm going to be testing 24 also.  I am going to see what capacity creates the greatest increase.

(I hope I'm not boring anyone with my posts)

Jon, This is actually beyond my wildest expectations.  I think the relationship between the battery and the capacitors is like the relationship between concrete and steel. 

Neither the battery nor the capacitor is perfect by itself.  The capacitors have limited (compared to a large battery) power storage however can provide nearly unlimited amounts of surge amps.  The battery holds a lot of power yet cannot provide any surge amps without a nasty sharp reduction in amp hour capacity along with destroying the plates inside the battery.  Together they produce a "super battery".

That is a good call on the fusing.  I appreciate your friendly reminder.  I hadn't even considered protection and was focused on trying to minimize losses of any kind.  I never even thought about fusing this because its tied to the batteries directly.  After reading your post and giving it some thought I really should have some type of fusing installed.

I'm not sure how I would fuse cell to cell in the capacitor bank.  If I use a lead acid battery as an example there is no fusing between the individual cells inside the battery itself.  Is fusing between cells necessary?  A friend suggested that I alter the aluminum I have currently hooked up to make it thin in the center.  Then the capacitor to capacitor links would physically melt/blow and should provide some level of protection.  The question would be how thin should they be made in the centers.

I don't know what the proper value would be to select for a fuse between the capacitor bank and the battery.  This capacitor bank is able to provide insane levels of current however I never expect to draw those levels of power.

Since I am getting 100amp shunts I guess 100amps would be the proper value for my testing.  I would have to replace these fuses once I attach more batteries and loads in excess of 100amps however.  I am open to any input on proper values for the fusing.

Please everyone keep in mind that I'm very new at this.  As I stated before I'm just an average person with a screwdriver and a multimeter.  If anyone sees anything I am doing wrong please let me know.  Chances are I honestly don't know the proper way.  This is especially true if you notice anything that is a fire hazard or a personal safety concern.

---------------------------------------

Some people have questioned why I am doing this at all.  I have had lots of people tell me that simply adding another battery (or batteries) would do the same thing.  The reason I am doing this experiment comes down to something called Peukert's Law.

http://en.wikipedia.org/wiki/Peukert%27s_law

This law defines the relationship between current draw and overall capacity of the battery.  It would be impossible (impractical) to design a battery bank around the 20 hour rate that most battery manufactures use.  My battery is a 100AH battery so the example given on wikipedia works perfectly for my system.

I am attempting to violate this law and produce a practically sized battery that can be charged completely daily for maximum battery health yet can provide all the power that I need.

I was doing some math on this last night and currently I'm seeing approximately a 40% increase in battery capacity.  This cannot be due entirely to the capacity of the capacitors and must have more to do with the peukert effect. 

Keep in mind that the capacitors are constantly charging and discharging as the battery voltage changes.  The capacitors don't work like a battery and only respond to changes in voltage.  The more the voltage sags (or increases) the more current the capacitors produce or accept.  I key up the radio and the voltage sags which causes the capacitors to discharge.  When I quit transmitting the battery attempts to return to its unloaded voltage and in the process of doing so it recharges the capacitors for the next blast.  Unlike a battery these capacitors can do this millions of times at nearly 100% efficiency.

I went ahead and followed your links and the links embedded in the pages and from what I can see my initial impressions are correct. Two controllers set at  different voltages, and charge rates would take advantage of the system in the best way possible.  One for the capacitor bank and one for battery bank are needed if you are going to get the most out of the system.  If you use three or four diodes in series between the batteries and the capacitors in the system to set the capacitor bank a little higher than the battery bank then they can be used a as charging supply for the batteries. I would use a large power relay to shunt around every thing when the high demand load kicks in and then drop the diodes and the controllers back in after the high demand ends. A short run from a generator or a brief connection to the grid with a high rate charging system for just the capacitor bank at night could be used to kick the capacitor bank back up to 2 volts higher than the battery bank. It will rise that high because the shunting relay is disconnected and the diode block is back in the circuit. Then the battery charging controller would send the capacitor charge to the batteries at a much tamer rate than hundreds of amps which the capacitors could provide but is not so good for batteries.  All this stuff only seems worth it if you have a demand higher than you can normally get from your available battery supply.  I have just been using car batteries and playing around with this stuff but I got my education B.C. (Before calculators)  Back then they taught us to do the math and show the work.  You got a much better feel for a problem that way. Fork lifts have to use deep cycle batteries because it is a guarantee that they will be abused.  The caps are helping
avoid that abuse and providing an additional punch for the lift. Typically an electric fork lift will be rated at a 5000lb lift limit.  The graph on the page about the ultra caps shows the battery current but doesn't show the capacitor current.  I would guess that they made up the remainder between 720 amps and what the batteries provided.

After the lift you will notice a slow smooth decline in the battery current.  It looks like to me that when the demand is off the caps have a small resister between them and the battery so the batteries can charge the caps back up gradually.  I saw a car battery blow the hood open on a car once. Nobody was anywhere near it but a charger that had switched to a 35 milli-amp trickle had just been removed from it. It made a mess. I wonder how big a bang those forklift batteries would make.