Extending Battery Life, by D.P.

Many preppers believe that batteries should play a prominent part in their preparations. For a variety of reasons, they are probably correct in that assumption. From what I have read on this blog they also generally believe that their batteries will reach end of life (or at least have a greatly diminished capacity) after 3-5 years. This is understandable but not necessarily correct. Since deep cycle batteries are not cheap its also an expensive assumption. This submission deals with how to extend the useful life of various types of rechargeable batteries.

There will be an emphasis on DIY [from very simple to complicated, so just pick your level …] and sustainability. I believe that there will not be a quick path out of the troubles before us. Whether society collapses overnight or in a more managed descent, historical time patterns suggest we’ll be lucky to regain today’s ‘normal’ (= go to store and just buy whatever you need) by 2020. In the meantime we may have to work with what we have on hand. So preppers should have an ‘I am in it for the long haul’ mentality. Along those lines: if I can build it, I can fix it! … and help others in my community along the way with my knowledge. Let’s use what little time we have left to prepare wisely.

First of all, creating your battery bank:

NOTE: This part only applies if you wire your batteries in parallel (i.e. create a 12V bank). If you wire them in series (24V or higher output) you can skip it. The best way to kill your batteries is wiring your bank incorrectly because the load will not be shared equally among batteries, leading to premature failure of the overstressed battery that will then start draining the good batteries in the bank. The correct way to wire a bank is easy to understand if you keep in mind that the full path current (inverter + to -) will take the route of least resistance. So we need to make sure that the wire length and number of connections are the same regardless of which battery the current goes through.

The following connection schemes (that I found on a UK web site which credited ‘smileypete’) achieve just that.
For two batteries:
1+ to 2+; 1- to 2-
Tap and charge bank through 1+ and 2- (or 1- and 2+) terminals

For three batteries:
Connect all + terminals to an external terminal with wires of equal size and length
Connect all – terminals to an external terminal with wires of equal size and length
Tap and charge bank through the external terminals

For fours batteries:
1+ to 2+; 1- to 2-; 3+ to 4+; 3- to 4- This effectively creates 2 blocks
1+ to 3+; 2- to 4- Tap and charge bank through 3+ and
2- terminals
or 1- to 3-; 2+ to 4+ Tap and charge bank through 2+ and
3- terminals

My 8 battery bank consists of 2 of these banks of 4 connected in parallel to the inverter through identical cables and I have noted no uneven discharge problems with my setup.

A word of caution:
With a bank of this size you will want to be very very careful when (dis)connecting cables, tightening nuts with metal wrenches, etc. A near zero resistance short will release far more energy than you care to deal with and can easily cause burns, fires and explosions. Also have proper fuses on all incoming and outgoing lines for the same reason.

The ‘battery life’ issue:
The problem with common battery life knowledge lies with what we are told about overcharging them. Overcharging is generally believed to bad thing … and it is … but not always … and so we need to qualify the term overcharging. Overcharging a battery with too much CURRENT (amps) is ALWAYS bad: it will shorten your battery’s lifespan, overheat the battery, boil off water, and can be dangerous if the battery or its surrounding area has venting problems. Overcharging with too much POTENTIAL (volts) is not necessarily a bad thing if the process is properly controlled. The good news is that this control is easy to implement.

About lead-acid batteries:
I am only talking about flooded cells here. Although I have read that gel and AGM types behave in the same way as flooded cells during (over)charging, I have no experience with them so I am not prepared to make generalizations at this point in time.

Maximum (dis)charge current:
People that have studied batteries more than me hold the opinion that limiting the current through a battery to its C20 rate is desirable since this will avoid overheating and does not shorten battery life expectancy in any way. This is true for both the charging and discharging process. A medium size deep-cell battery (T-1275 as example) is rated at 150 Ah. Its C20 current is therefore 150 / 20 = 7.5 Amps. Here we see an immediate problem because this means that we can safely draw only 12 * 7.5 = 90W from one battery. That’s okay for LED lighting, a 12V fan and charging cordless tools but little else. So we need to make a bank by connecting multiple batteries together. My 8 battery setup has a 720W capacity which runs my (corded) power tools without me having to worry about stressing the batteries at all. I don’t even worry if I connect a 1500W industrial vacuum cleaner or small welder to the inverter since I am still only discharging at C10 rates which won’t impact a battery too much if it happens only once in a while. It also means that I can charge the bank at 8 * 7.5 = 60 Amps without stressing the batteries in it.

On commercial charge controllers:
Let’s say you bought a 30 Amp charge controller to protect your battery and have it hooked up to a 150W solar panel and one T-1275 battery on a sunny day. The solar panel will put out about 10 Amps. This is within the 30 amp limit of the controller but above the battery’s C20 rate (7.5 Amp), so you’re happily reducing your battery’s lifespan and the money you spent on the charge controller was a total waste of resources. Why a total waste? What about stopping the charging process when the battery is full?

The important voltages for lead-acid cells are as follows:
(A 12V battery has 6 of these cells in series, so multiply the numbers by 6)
1.75V empty
2.01V 50% charge
2.06V 75% charge
2.12V-2.15V full when resting (= at least 1 hour no charge/discharge applied)
2.4V full when charging
2.6V cell balancing voltage

On charging voltages:
If you connect a solar panel directly to a battery, the battery will clamp down the voltage of the solar panel to about 13-14V(max) and absorb all the solar energy in the process. If the battery’s plates are fully charged, the additional energy will go into a process generally referred to as boiling. Is boiling a bad thing? Not necessarily and certainly not in stationary deep cycle batteries. You will need a certain amount of boiling to keep the electrolyte from settling. Your car battery doesn’t have that issue if you drive through the odd pothole or across other bumps but for stationary batteries it is a real problem.

Secondly the boiling that occurs from potential (over voltage) is different than the boiling that occurs from high current. It sounds different (small bubbles instead of big bubbles) and doesn’t boil off
the water. I am not sure what is being released but a marine battery that I bought at Wal-Mart (three years ago for stress testing) has been through many[short duration] boils and I have yet to add a drop of water to it as its cell’s water levels are still as high as when it was new.

As you can see from the table above a 12V battery is fully charged (max capacity) at 6 * 2.4 = 14.4V. But there is one entry after that for cell balancing. This happens at 6 * 2.6 = 15.6V. In short cell balancing is fixing a bad cell by over potentializing it. Generally speaking if your battery’s capacity drops, its because 1 cell has gone bad and drains the others. For a more detailed description you can google the term “cell balancing”.

Cell balancing process:
Simply connecting a solar panel directly to a battery seems to accomplish this cell balancing (= restoring the battery’s capacity) under the following conditions:
– battery is in decent shape = resting voltage reads 12.3V or higher.
– battery is not discharged during the process (i.e. you cannot use the battery)
– the process takes time; at least a few weeks if most days are sunny.

I told you that it was easy to maintain your batteries!

I ‘bumped into’ this process last winter when it was too cold to work in the yard. 1 bank of 4 T-1275 batteries was sitting at about 12.35V so I connected them to a 60W solar panel to avoid discharging them further and walked away. Six weeks later as temperatures started to rise I opened the battery box and found all batteries softly boiling. My volt meter showed 14.4V. I unplugged the solar panel and the next morning the resting voltage was 12.80V! Using the batteries this spring I noticed their capacity is much higher than it was last summer: no more instant collapse from 12.6V to 12.3V. What’s most special is that I got the batteries used. They had spent the first three years of their life powering golf carts around a local golf course and were replaced because they couldn’t get the job done any longer.

So lets do some math. At my elevation a 60W panel delivers about 3.5 Amps for a few hours on a bright sunny day in the middle of the summer and also on a sunny winter day with a fresh layer of snow on the ground. Spread over 4 batteries that is .9A per 150Ah battery. Which is barely a trickle charge for them and roughly 12% of their C20 capacity making it highly unlikely I would overcharge them even if left unattended. I think its most likely that the batteries were fixed by the high voltage generated by the solar panel. It is possible that this method works better in colder climates because my solar panel voltage is de-rated at -.5%/degree Celsius. This means that on a cold winter day it puts out 17% higher voltage than its rated capacity. For my panel that translates to about 20V in a closed circuit.

Coming back to charge controllers; it seems to me that as long as you keep your charging current below your battery’s C20 rate by matching panel to battery, you cannot destroy (but only improve) your battery by applying the solar panel’s full voltage to it. No need for a charge controller that cuts out at 14.4V, thereby eliminating the possibility to equalize your cells.

Under ‘cell balancing process’ I mentioned that the resting voltage of the battery should be 12.3V or higher. The reason for this is that batteries below that voltage cannot be restored to full capacity by just connecting them to a solar panel. Although beneficial, the voltage applied by the solar panel cannot reverse the process of battery plate deterioration called sulfating. So should we get rid of these batteries? Nope, at least not if you are a handyman. Sulfated batteries can be restored by a pulse charger, unless heavy bridging between the plates has taken place. If the [sulfation] bridges are too strong to break by shaking the battery, your best bet is to leave it there and find a replacement battery.

Pulse charger:
So what’s a pulse charger? Essentially its an air-core magnetic coil that is pulsed with DC voltage. As the current through the coil is turned off, its magnetic field collapses and releases a short high voltage spike that will get the desulfating job done if you can capture it and send it into the battery. The size of the voltage spike is related to the size of the coil and the amount of power delivered by its power source. There is actually quite a bit of science involved if you want to optimize the design, but any coil/power source configuration will do something albeit at lower efficiency. Keep in mind that small coils cannot handle large batteries: they will create a surface voltage, but your battery has no capacity when you start using it.

So for large batteries (car batteries of larger) your pulse charger will need to be able to handle a decent amount of power. My current pulse charger’s coil is made of a pound or so of magnet wire (10x 90′ strands of 24GA magnet wire wound in parallel [low internal resistance] on a 4″ high form). A smaller coil would not have the mass of copper required to generate pulses with enough energy content.

When I attached the charger made with this coil to a 30W solar panel (1.75 Amps) it worked just fine. When I connected it to a 60W panel (3.5 Amps) it never worked because the charger’s switches were instantly zapped (power MOSFETs rated at 400V). Those switches have been replaced by 1000V parts and now everything works fine. The charger even managed to bring deep cell batteries measuring 11.4V and a 12V car battery indicating 4.5V back to life. If batteries get that bad, the initial charge takes several sunny days and a 60W panel to achieve and several charge/discharge cycles are required to get back to a reasonable capacity. For the 11.4V deep cell batteries I used 2 60W panels: one connected to the pulse charger and one connected to the battery. You need voltage levels to reach 13.8V – 14V in order to get battery capacity above 50%.

Charging a battery with 500-600V is dangerous indeed if you apply continuous current. However the coil’s magnetic field collapses in less than 10 nanoseconds. So @ 12 kHz I am charging the battery for 12000 * 10 * 10-9 = .12 msec/second; giving it plenty of time to absorb/disperse the energy.

For the technically inclined handy man:
You can build your own pulse charger for $50-$100 in materials, depending what you have on hand. Following are its crucial parts:
– 5000uF capacitor to store energy from solar panel
– diode(s) between capacitor and coil input to force voltage spikes into battery (600V 30A ultrafast)
– diode(s) between coil output and battery (pos. terminal) to tap voltage spikes (600V 30A ultrafast)
– a wire connecting the capacitor’s positive terminal with the battery’s negative terminal (don’t forget!)
– switch(es) between coil output and common ground (800V+ power MOSFETs, shorter fall time is better – I am using four switches to spread the load.
Stressed and hot semi-conductors and longevity do NOT go hand-in-hand.)
– heat sink for switch(es) – I salvaged one from an old desktop computer power supply
– MOSFET driver (UC2950 works for me)
– 555 timer or microcontroller to turn switches on/off @ 12 kHz ~50% duty cycle
(if you know how to write a simple BASIC program a microcontroller is the better option – picaxe 08M2 SoC’s can be purchased for about $2/piece [www.techsupplies.co.uk] and programmed through a laptop’s serial port using free-to-download software)

Will transistors work instead of MOSFETs? Yes, but not as well. Their fall times are usually measured in microseconds as opposed to nanoseconds for MOSFETs. The faster you can cut the current through the coil, the higher your voltage spikes will be.

BTW I did not come up with this design myself. Its adapted from postings in various alternative energy forums, mostly based on the work of someone who goes by the moniker Jetijs. Too bad a lot of people in those forums get hung up on chasing over-unity effects within their contraptions, which is next to impossible due to the small size of their builds. But we can still use their technical insights for other purposes.

Why use a microcontroller:
On my system I use a microcontroller for two reasons:
– When I make a mistake in the design its easier to fix a piece of software than to de-solder some components.
– This is still a work in progress: from time to time I get an idea based on what I see on my volt meter and I want to test that. Again its easier to adapt the software than to built a new circuit board. And if the idea doesn’t work its simpler to delete the code than to try to reclaim parts from a now obsolete board.

As a result of the cumulative ideas, I have now a much more versatile charger than if I had to build it with a simple timer chip. For instance: On start-up the microcontroller tests its power source and loads an initial set of parameters based on the test results. If it realizes during operation that it picked the wrong set, it can fix that mistake. In order to optimally use the available power, the micro controller can vary the charger’s duty cycle from 5-65% and it’s frequency from 4-40 kHz as it tries to keep input voltage close to 17V when connected to a 60W panel, which seems to be the sweet spot for this combination. The idea is to try to create an optimal spike not just when the sun shines brightly but also under less favorable conditions or with different size panels. The charger just creates a different number of and/or smaller spikes per second.
Again, this was no grand design; its simply what the project evolved into to date.

For the not technically inclined:
A company called Energenx sells a charger called the rejuvenator. The underlying technology is close enough to what I described above that I expect them to work, though I haven’t tried them. They are quite expensive and use a 110V outlet, but if they double the life of your battery bank it might be a worthwhile investment.

Dry-cell batteries:
So far I have only dealt with lead-acid batteries. However pulse chargers produce the same effects in dry-cell rechargeable batteries. Some claim success with alkaline batteries too, but I haven’t seen that myself: increased voltage, yes, real sustained capacity, no. The technology seems to work with lithium cells too if you are careful with regards to voltage, but I have no lithium cells so I cannot speak from experience. Do not expect to recharge a laptop battery with it: you will probably zap the embedded electronics rendering the battery useless.

I am now using a pulse charger exclusively for my NiMH and NiCd batteries and it works very well. I should qualify that statement: for good quality batteries. Cheap Chinese batteries have about a 50% failure rate after a few cycles due to membrane rupture. On the other hand I have some NiMH from 2001 that are still in use. I was about to throw them away by the time I built my first pulse charger because they powered my cordless mouse for only about one day before dying. Then I put them in my pulse charger and now they run the same mouse for 4 to 6 weeks before they run out. I also found some Radio Shack NiCd batteries from the 1980s that are now doing duty again in garden lights with better results than some of the batteries shipped with new lights. Solar light batteries receive some pulse charger time during the winter months and some are now into their 5th season and still keep the LED going through the night, though you can’t save them all.

Most notable is that the batteries stay cold during the charging process which helps to improve their life expectancy because heat is the biggest killer of small rechargeable batteries.

I am charging AA and AAA cells in sets of 4 to around 6V. On a nice clear day you can achieve this with a 1W solar panel if you charge one set at a time. For charging multiple sets simultaneously, use 3-5W panels as a minimum power source. 9V batteries should be charged to around 10.5V to reach full capacity. If you want to use your charger with larger panels it should monitor these voltage levels because it is relatively easy to zap dry-cell membranes if you put too high a voltage across them. A 1W panel has a hard time reaching 6V under the best of circumstances so no worries there.

A pulse charger for these batteries has the same parts and layout as the one described above but with much smaller/cheaper parts. The coil is a single layer of 24GA magnet wire about 4″ high that uses a piece of 3″ ABS pipe as coil form. An empty Coke bottle works great as coil form too, but avoid PVC as its too dense and impedes the magnetic field noticeably. The capacitor can be 100 uF, the diodes 100V 1A ultrafast or Schottky, the switch needs no driver or heat sink and can be something like an IRF510 (100V, 5A) if you use a 555 timer to drive it. With a microcontroller you should use an IRF520N or similar low input voltage MOSFET.

Will pulse chargers run from power supplies other than solar panels? Yes, I have run them from 12V and 24V batteries as well as laptop power supplies without problems. Pretty much any DC power supply works well since the large input capacitor stabilizes the supply if needed.

Would I normally consider buying a solar panel just to charge a battery? No way, still too expensive per kWh. But I expect supply chain problems to arrive before panel pricing gets much better than it is now and I want to avoid the darkest of the ages. When that day comes I need generating capacity at home, not at a distant vendor’s place. For my location solar works better than other alternatives and I decided I might as well start using the panels now and know what to expect from them when it counts.

For the skeptics that feel the urge to write in about how and why all of this won’t work: Please don’t. I am fully aware that what I wrote goes against conventional wisdom. Which is why you need to replace your batteries every few years, so I can pick them up on the cheap (sometimes even for free) and restore and use them again. Many thanks for the opportunity. Especially when using a pulse charger you are using a totally different process when charging your battery than with a conventional charger as evidenced by a very different charging voltage curve and battery temperatures. I have built and tested all the setups myself and am simply reporting the results I have seen. This posting is meant to get word out to the preparedness community, hoping to help them a bit with their decision making and preparations.

For the rest of you: Take the worst battery (lowest resting voltage) from your bank and connect it to a 15-20W solar panel for a few weeks. [You can use a larger panel too as long as its output is less than your battery’s C20 rate.] Then exchange it with the second worst battery in the bank. Keep repeating until you have rotated through the entire bank. Alternatively you can use spare batteries for the rotation. Then start the entire cycle again with the first battery if you want to keep your batteries in good shape. You will be pleasantly surprised by the results.