My first experiments with a Solar Electric panel and battery
It was late December 2002 when I first started to experiment with solar electric panels, a Uni-Solar brand 5 watt model. Rated for a 12 volt battery system the output is 300 milliamps in full sunshine. Not big enough to hurt myself, very badly, but big enough to do some real work.
In order to make good use of the 5 watts we should store all the energy captured by it into a battery. I chose a 12 volt 5 amp hour sealed lead acid battery. With this we can use the captured energy on cloudy days and at night. Another consideration is the 300 milliamps output limit of the panel. A battery can supply much more than this for short periods of time, which means we can operate higher power devices with a battery than we can with the solar panel all by itself.
The Solar Panel
[ photo of Uni-Solar 5 watt panel ]
Here the panel is mounted to a frame attached to the roof, well, under the roof. I didn't want to punch holes in the roof and possibly cause water leaks into the house. It's pretty clear from this photo that I was already planning for expansion. The empty area on the left would be filled soon enough.
The Battery
[ photo of 12 volt 5 amp hour battery ]
The battery is a standard "sealed" lead acid type. Sometimes referred to as VRLA (Valve Regulated Lead Acid) or Deep Cycle. Homes normally powered by solar electricity which have a battery bank use either Wet Cell or Sealed VRLA batteries. Both of these types are always Deep Cycle never "car" batteries. Keeping in line with home solar power, I chose the Sealed Deep Cycle battery for this experiment.
Charging the battery
It was a sunny morning in December 2002 and I was holding the little 5 amp hour sealed lead acid battery that I had purchased the day before. I was anxious to see the solar panel charging my new battery. So, at first I just propped up the panel inside one of the windows on the south side of the house to let the sun shine on it through the window glass. I hooked the solar panel in series with an ammeter on the positive wire to monitor current flow. I think I saw about 130 milliamps on the meter. Great! This is so cool, getting power from sunshine to charge the battery. OK, the battery is charging, so I left it like that and walked away for a few minutes.
Oops! Wrong thing to do. When I returned, the battery was making a sound! The battery sounded like water simmering in a pan on the stove. Yikes! I disconnected the solar panel immediately. The sound stopped. The battery wasn't hot or leaking anything. I only left it hooked up to the solar panel for about 20 minutes.
Important lesson learned: Don't connect solar panel directly to battery and leave it unattended!
That single event has probably shortened the life and reduced the capacity of the battery.
OK, some sort of automatic monitoring is needed for charging the battery. I knew that and was ready with an answer, a charge controller, but I didn't have one yet. No more charging from the solar panel until a proper charge controller is attached.
Oops! Wrong thing to do. When I returned, the battery was making a sound! The battery sounded like water simmering in a pan on the stove. Yikes! I disconnected the solar panel immediately. The sound stopped. The battery wasn't hot or leaking anything. I only left it hooked up to the solar panel for about 20 minutes.
Important lesson learned: Don't connect solar panel directly to battery and leave it unattended!
That single event has probably shortened the life and reduced the capacity of the battery.
OK, some sort of automatic monitoring is needed for charging the battery. I knew that and was ready with an answer, a charge controller, but I didn't have one yet. No more charging from the solar panel until a proper charge controller is attached.
Solar Charge Controller
[ photo of Morningstar Solar charge controller ]
I chose the Morningstar ProStar-15 Charge Controller because the power handling capability is well suited to my small solar electric system, and it is regarded as highly reliable and good design from the reviews I've seen. It's capable of handling 15 amps total charge current into the battery, and 15 amps on the load side which is way more than I need right now. It will give me some room to add more solar panels and batteries to the system in the future, which is good.
Here the charge controller is fully connected to into the solar electric system: solar panel, battery, load circuit bedroom reading light. I have connected the load to the output of the charge controller instead of the battery directly. It will shut down the load if battery voltage drops below 11.4 volts, and reconnect power to the loads when the battery terminal voltage rises above 12.6 volts. A nice feature that will save the battery from being completely drained.
Here the charge controller is fully connected to into the solar electric system: solar panel, battery, load circuit bedroom reading light. I have connected the load to the output of the charge controller instead of the battery directly. It will shut down the load if battery voltage drops below 11.4 volts, and reconnect power to the loads when the battery terminal voltage rises above 12.6 volts. A nice feature that will save the battery from being completely drained.
Bedroom reading lamp
[ photo of the first, original reading lamp ]
The bedroom reading lamp shown here is pretty much the way I used it. A bare automotive 12 volt bulb soldered to some zip cord with a slide switch in-line for on off operation. I think I used a clamp or some tape to hold the lamp in position for reading. Primitive, yes, but it worked. My first foray into solar electricity.
I was excited. Kinda cool laying there in bed reading by a light that came from sunlight earlier in the day. The local power company isn't reading the power used by this lamp on their kilowatt hour meter on the outside of my house. How cool is that? Yeah, this little step into solar electricity caused many thoughts to stir in my imagination.
Improvements to the system
Well, that was the start. It didn't take long before I was thinking about how to improve the safety and efficiency of operating the solar electric system. One of the first things I did was to move the battery and charge controller into the master bathroom, closer to where the solar panel wiring enters the house. The window seen in the photo behind the solar panel at the top of this post is the master bathroom. I drilled a tiny hole into the aluminum window frame to pass wires from the solar panel into the house. Yes I know, this doesn't meet code but it's just a temporary hook up for experimenting. So onward.
I set the battery on the window sill. Attached some metal brackets to the back of the charge controller so I could attach it to the window sill with some clamps.
I set the battery on the window sill. Attached some metal brackets to the back of the charge controller so I could attach it to the window sill with some clamps.
[ photo of solar electric system on master bathroom window sill ]
The analog meter in this photo is connected in series with the solar panel so I could monitor incoming current. The charge controller can't read anything less than 200 milliamps. You can see the little plastic clamps on either side of the charge controller holding it to the window sill. The output connection to the reading lamp has a fairly long run of wires to get all the way to the headboard. It's about 15 feet from here to there.
The automotive bare bulb pulls about 600 milliamps from the battery when it's on, which is a lot for a small light. So the next improvement is to find a useful reading lamp using less power. Here's my first answer to lowering the power.
[ photo of better reading lamp ]
A 12 volt fluorescent lamp pulls about 300 milliamps, that's half the current of the automotive bare bulb. And the light from this one is softer, whiter and diffused. Much nicer for reading. Things are looking up! If you look closely at the picture you'll see clamps at either end holding the lamp secure to the head board. Not pretty, but again, it works and after all this is an experiment.
Battery monitoring
The battery and charge controller are in the master bathroom while the reading lamp is in the bedroom, can't see them from the bedroom. The charge controller has a volt meter and LED's to show battery condition but, it would be nice to see something at the reading position. I would be able to see if I could read another page or if I should turn the light off and go to sleep. Some time back I built a car battery voltage monitor using LED's. That would work here, it's 12 volts!
[ photo of 12 volt car battery monitor ]
All I had to do was attach two wires where the 12 volt power connects to the reading lamp. Cool! Now I have an indicator showing me battery voltage while in bed reading by the light. Well this is fun.
How well is it working?
I'm happy with the operation of everything connected to the system, but. Yeah, you knew there was going to be one of those. It seems there isn't much battery power at night to run the reading lamp. The "low battery" shutdown on the Charge Controller has shut off the reading lamp during use more than once. Annoying? Yes. I've seen the yellow and red LED's lighted on the Charge Controller fairly often telling me the battery is low.
Well, research into proper charging profiles for lead acid batteries has shed light on the reasons for this happening. Listed here are important steps in caring for a deep cycle lead acid battery used in a solar home application.
- A battery should never be discharged more than 10% of its marked capacity in Amp Hours.
- A load on the battery should never exceed 10% of its marked capacity.
- Charge current into the battery should never exceed 20% of its marked capacity.
- A sealed VRLA battery should never be overcharged.
When any of these are exceeded, even just briefly, the life and capacity of the battery is reduced. Normally they will live for 5 to 8 years, or even longer with great care.
So, right off I managed to do #4 when I connected it directly to the solar panel and didn't monitor the voltage. My 5 amp hour battery is probably only worth 3 amp hours now, or even less!
Let's take a close look at what we have here
Loads:
- Reading lamp, 300 ma, 1 hour per day, 0.3 ah/day
- Charge controller, 20 ma, 24/7, 0.48 ah/day
- LED voltage monitor, 30 ma, 24/7, 0.72 ah/day
Total 1.5 ah/day (0.3 + 0.48 + 0.72 = 1.5)
Charging:
- Solar panel charging, 300 ma max
- Prime Solar Window, 4 hr sunshine/day
Total 1.2 ah/day (0.3 x 4 = 1.2), ideal conditions assuming 300 ma for 4 hours, can't happen, see below.
Battery:
- Sealed lead acid, 5 amp hour capacity (probably less now), 0.5 ah capacity solar home rating
These figures show me that we've got a huge load on the little 5 amp hour battery, too much in fact.
Daytime charging cycle is not enough to put back as much as we took out. The Prime Solar Window cited above is when the sun is high enough above the horizon to supply the maximum amount of energy into the solar panels. At other locations it can be 5.5 hours or even less than 2 hours, depending on your latitude above the equator. The calculation 300 ma for 4 hours cannot happen because the battery voltage will rise and the sun changes position during this time. When the voltage gets to about 14.4 volts the charge controller goes into regulation mode which holds the voltage constant by reducing current going into the battery. The result is that 300 ma won't hold the full 4 hours. Which means our calculation above is not correct in reality, but it does give us a rough idea of the charging capacity available during the day.
The Charge Controller uses 20 milliamps around the clock and about 35 milliamps when it is actively charging the battery. So we have about 0.48 amp hours used up by the Charge Controller itself, leaving very little left for operating other loads (0.5 ah - 0.48 ah = 0.02 ah left for lights). Yikes!
The LED voltage monitor uses more than the Charge Controller! It's using 30 milliamps around the clock which adds up to 0.72 ah. Again, yikes!
Each one is too much load for the battery. Together they will kill the battery in a short time.
Now that I have a better understanding of how to charge and how to use a sealed lead acid battery, I see why the system I've put together is not working right. I put too much load on the battery. Since I want to operate more than just the charge controller, it's time for a system upgrade.
Daytime charging cycle is not enough to put back as much as we took out. The Prime Solar Window cited above is when the sun is high enough above the horizon to supply the maximum amount of energy into the solar panels. At other locations it can be 5.5 hours or even less than 2 hours, depending on your latitude above the equator. The calculation 300 ma for 4 hours cannot happen because the battery voltage will rise and the sun changes position during this time. When the voltage gets to about 14.4 volts the charge controller goes into regulation mode which holds the voltage constant by reducing current going into the battery. The result is that 300 ma won't hold the full 4 hours. Which means our calculation above is not correct in reality, but it does give us a rough idea of the charging capacity available during the day.
The Charge Controller uses 20 milliamps around the clock and about 35 milliamps when it is actively charging the battery. So we have about 0.48 amp hours used up by the Charge Controller itself, leaving very little left for operating other loads (0.5 ah - 0.48 ah = 0.02 ah left for lights). Yikes!
The LED voltage monitor uses more than the Charge Controller! It's using 30 milliamps around the clock which adds up to 0.72 ah. Again, yikes!
Each one is too much load for the battery. Together they will kill the battery in a short time.
Now that I have a better understanding of how to charge and how to use a sealed lead acid battery, I see why the system I've put together is not working right. I put too much load on the battery. Since I want to operate more than just the charge controller, it's time for a system upgrade.
New solar panel
[ photo of solar panel array, 2 panels ]
three months later, March 2003...
I knew designing the solar panel mount with an empty slot for another panel would come in handy, I just didn't know it would be this soon. Well, that's what experimenting is all about, learning from mistakes. The additional panel increases the current capability from 300 milliamps to 600 milliamps maximum, still below the recommended maximum 20% charge current, 1000 milliamps. Two 5 watt panels together make 10 watts total. That's going to help a lot.
[ photo of C-clamp holding solar panel mount to roof joist ]
I removed the battery voltage monitoring LED display to reduce the overload condition. With these two improvements the battery fully recharges during the day and I get about 1/2 hour use of the reading lamp at night. There doesn't seem to be much power left in the battery at night and it's pretty low by morning from powering the Charge Controller all night long. I expected this because we're still overloading the battery.
New battery
one year later, December 2003...
It is time to replace the abused 5 amp hour battery with a new one. I'll keep the original battery attached to the system because it seems to have some capacity left in it.
[ photo of new 12 amp hour battery ]
The new battery is rated at 12 amp hour capacity which is more than double the original 5 amp hour battery. One tenth the marked capacity of this battery would be 1.2 amp hours. Our loads require 0.78 amp hours to operate so we should be OK now. Because the new battery will only see a proper charging profile from the Charge Controller it should last a long time and serve me well.
Put everything into a proper enclosure
At the time the 12 amp hour battery was added to the system I also mounted everything into a NEMA steel enclosure box purchased at a local home improvement center. I'm using this box for something other than its intended purpose, adding several fuses to the top of the box for circuit protection and tie points for external circuit connections. I added a couple large analog meters to show battery voltage and solar panel input current. The large square hole in the front is where a battery monitoring device was attached but I needed the space for extra battery connections. It's a strong, heavy gauge, nicely painted metal box. Perfect!
[ photo of NEMA steel enclosure ]
The large terminal block had several high current pigtails attached, allowing me to connect external sealed lead acid batteries to the system. These batteries come from UPS units I no longer use. I just wanted to put them to use instead of just letting them sit idle, so here they are.
[ photo of TriMetric battery monitor ]
This is the Trimetric battery monitoring device I added to the system. I found out about this useful device while reading Home Power Magazine which is all about solar power use in the home. Very enlightening reading! You set up some parameters in the Trimetric describing the size of your battery (amp hours), the voltage, and the current sensing shunt you have attached to monitor incoming and outgoing current. Once this is done the device can tell you how much power has been used and how much is left. Very useful information which goes a long way to help you maintain a healthy battery bank.
The Trimetric display gives you access to a lot of information about the condition and history of your battery. The information can help you to spot trouble so you can fix it before damaging your battery. For a solar power battery charging system, this is the most important piece of equipment you can have.
How is it working now?
Very well. The battery stays charged and in good condition. Once or twice during long, cold, cloudy winter days the battery didn't reach fully charged for several days. The charge controller's low battery shut down turned off the power to save the battery from being discharged too deeply. Those were some long, cold winter days where we had a thermal inversion: freezing fog on the ground in the valley where no sunshine gets through to the solar panels. Cold, but pretty, seeing frost collect on the trees, fences, telephone wires and solar panels!
Knowledge is power
All this time spent learning about battery charging and solar panels has taught me a great deal about the proper ways of designing and building a complete system.
- Find the right size battery for the loads you intend to operate.
The battery for my system has to operate a nighttime reading lamp for one hour per night and the Charge Controller 24/7, that's about 0.78 amp hours. Both batteries add up to 17 amp hours available for use but probably less due to the abused 5 ah battery. I purposely choose a battery too big so I could add some more loads for experimenting. Testing to see how much load is to little, and how much is too much.
- Find the right size solar panel array to keep the batteries charged.
The solar panels need to recharge the batteries with about one-and-one-half times the amount that was taken out of them. Two 5 watt panels will put out about 600 milliamps in full sunshine. 600 ma times 4 hours is 1.2 amp hours, which is pretty close to what we need. I won't let the battery get that low. Taking two days to bring the batteries up to a full charge isn't ideal, but it's not too harmful either. Just don't design a system where two days is the norm.
- Use a high quality Charge Controller for charging the batteries.
The Morningstar Pro-Star 15 is indeed a high quality, well designed Charge Controller. It is capable of handling much more power than my little system is using, allowing additional panels and batteries to be added at a later time.
Charge Controller battery charging profile
[ chart Sealed Lead Acid battery voltage charge profile ]
This chart describes a normal "solar day" for the batteries, showing how battery terminal voltage changes during the day. Sunlight energizes the solar panels, in turn causing the Solar Charge Controller to charge the batteries in a very specific way. This curve represents an "ideal" charging profile on a sunny day. Some charge controllers do it very well. The Morningstar Pro-Star 15 in this system is one of those.
Night
This is typically when you use the batteries to provide light around the house, TV, radio and other uses.
Bulk Charge
When the sun comes up over the horizon in the morning the charging cycle begins. Direct current (DC) from the solar panels wakes up the Charge Controller which sends current into the batteries. Battery voltage starts rising slowly as the sun rises higher in the sky increasing the current. At this time the solar panels are producing as much power as they can.
Absorb Charge
When a set point voltage is reached, about 14.4 volts, the current begins to decrease which stops the voltage from rising any further. During this time the batteries are absorbing, or accepting a full charge. This process removes sulfate from the lead plates and returns it into the battery acid solution.
Float Charge
When the current reaches a specific minimum value the battery terminal voltage is reduced. At this point the batteries are completely charged. While there is still sunshine available the Charge Controller will hold the batteries at an elevated voltage to simply keep them full. This voltage is about 13.7 volts at 70 F degrees. When the sun goes down the cycle starts all over again at night, discharging the batteries.
[ photo of 33 amp hour battery ]
2 1/2 years later, February 2007...
I'm adding another 2 solar panels to the array to charge the batteries a little quicker. The solar panels are mounted in a new location. Stronger mounts, aesthetically pleasing to the neighbors (and me too), and it will allow future expansion. The first mounting for the two panels was just attached with C-clamps, this one uses 3/16 wood screws with pilot holes. Yeah, lots more professional.
[ photo of 4 solar panel array ]
I used B-Line angle steel to fabricate the mounting brackets for the solar panel array. That stuff is wonderful for building things. It's the adult sized Erector-Set, remember those?
[ photo of solar panel array mounting brackets ]
As I suspected the 1:5 ratio was a little too far to go. It kept the batteries charged pretty well but it was a struggle at times. Cloudy days would keep them from being fully charged. Several times in the 2 1/2 years it would take two days for the batteries to rise to the absorb voltage, only making it into the 13 volt range on some days.
The additional solar panels will bring the system up to 20 watts total, 1.2 amps maximum charge current. This puts the ratio at 1:2.5 with 20 watts and 50 amp hours. A lower ratio like this will charge the batteries a little faster. On cloudy days it will bring the batteries to a higher charge level too.
At this point the system ratio is below my gut feeling ratio of 1:3 being close to optimum. To adjust this a little I'm going to add four 7 amp hour batteries to the system for a total capacity of 78 amp hours. That moves the ratio back to almost 1:4, where I lived with it that way for about three years from about 2003 to 2007. I was pretty much satisfied with the performance back then so adding these unused batteries to the system should be OK.
78 amp hours of battery capacity is some serious power storage! I'm going to be able to use some of this power on my electronics workbench to do some real work. This is getting funner. Is that a real word?
almost 2 years later, October 2008...
At this point I have a large 12 volt system powering my home. The system is teaching me about wire size and voltage drop across terminals and connections. It's pretty amazing how easy it is for 12.7 volts at the battery to become 11 volts at the load when the circuit includes a few connections in between. The idea here is to start with a higher battery voltage. Then by the time the electricity arrives at the load the voltage drop will be a smaller percentage of the original battery terminal voltage, you get more power to the load.
The system is working very well. In addition to using the 12 volt system for room lighting, I've been using it at my workbench to power electronic circuits I'm working on, charging 1.5 volt AA and AAA batteries, and all sorts of other things. Having a larger battery bank seems to keep a lot of power in reserve. Which brings me to the next adventure...
It's time to try a 48 volt system. I've been thinking about energy storage and power. The relationship of voltage and power is interesting. All of the calculations for solar panel watts to battery amp hours that I have spoken of here are based on 12 volts. Those calculations don't work for 24 or 48 volt systems. So I want to see what higher voltage will do besides have less voltage drop on a long run of wire.
[ photo of 12 solar panel array ]
Solar panels for the 48 volt system are just an extension of the four 12 volt system panels you saw in the photo just above this one. Eight solar panels are used for the 48 volt system making the whole array twelve panels long. They are all the same Uni-Solar brand 5 watt, 12 volt 300 milliamp output.
[ photo of 12 solar panel mounting brackets ]
Four of these solar panels wired in series connection will give you 48 volts output at 300 milliamps. I want to build this system using four 33 amp hour batteries like the single one I added to the 12 volt system in 2003. It's a nice size and I have the perfect size shelf space just waiting for them. I need to make sure there is enough solar power to charge this battery bank. So, what have we got here...
We're right inside my gut feeling ratio at 1:3.3. The solar panels and battery choices have put us right in the middle of the sweet spot for building this system. I currently don't have any big plans for a load on this system, so I'll just take it slow and see what develops.
An interesting application for the high voltage of this system presented itself. I have a collection of antique radios, some of which are battery powered. They usually take a low voltage high current A-battery to operate the filaments in the vacuum tubes, and a high voltage low current B-battery to operate the plates in the tubes. So... I connect a 1.5 volt AA cell for the filaments, and the 48 volts from the solar electric system for the plates. Yeah, I did it and it works great! Now this isn't really a "load" on the 33 amp hour batteries because these radios only draw 0.01 amps, that's 10 milliamps! One of these radios would never deplete the battery bank in this system. But it's kind of fun running the old vacuum tube radio with the 48 volt system.
Here is the 48 volt Morningstar Charge Controller mounted in a NEMA metal electrical box just like I did with the 12 volt system.
[ photo of 48 volt Charge Controller NEMA box ]
The LCD display in this photo is showing 52.6 volts and you can kind of see the same on the lower meter on the right. That's because the normal terminal voltage of a 12 volt battery is 12.7 volts when fully charged, usually higher than that when it's freshly charged. So the 52.6 divided by 4 is 13.15 volts. It's this high because when I took this picture the sun was just setting over the horizon after being held in float mode at 13.7 volts (54.8 volts) in the afternoon.
Charge current into a sealed lead acid battery seems to work best when it is about 1/40th to 1/60th that of the marked battery capacity. That doesn't appear logical or intuitive. It just seems too low to be any good. But that's what my experiments seem to bear out.
I have lived with the Step 5 setup now for 6 years. It works very well.
Step 6
The 48 volt system. 40 watt solar panel, 33 amp hour battery bank. I initially thought 1:3.3 ratio was right in the middle of the sweet spot. After 4 years of use, 2008 to 2012, I can tell you that I was right! This little system puts a good charge into the batteries even after I've used a fair amount of power out of them. Today I have a good size load for this system which is used daily. A 48 volt to 12 volt DC to DC converter which I use to power the DSL Router connected to my computer. The converter pulls about 270 milliamps at 48 volts while the Router pulls 400 milliamps at 12 volts from the converter. And, yes, I'm still powering the old vacuum tube radio on 48 volts too.
12 volt system
Since I removed the eight 5 watt solar panels from use on the 48 volt system to use the broken solar panel in their place, they would be sitting idle. I rewired them for 12 volts and added them to the original 4 panels on the 12 volt system. Now we have twelve 5 watt panels feeding the 12 volt system, 60 watts! With 60 watts of solar panels and 78 amp hours of batteries, that brings the ratio up to 1:1.
We lost the original 5 amp hour battery in November 2012, it finally died. I noticed an unusual charging profile on the 12 volt system where the voltage would vary during the bulk charge phase. One-by-one I disconnected the batteries attached to the system. That's when I noticed the 5 amp hour battery felt a little warm. Warm! Yes, and looking at the case I could see a small bulge on one side. Yikes! As a result of these observations I removed it from service. The voltage measured about 10.5 volts. It's done. Gone. I put it in a safe place in case it decides to "GO" or leak or something. It will be recycled on my next trip to the battery store.
That little 5 amp hour (abused) battery has been in service from December 2002 to November 2012, that's 10 years! I guess the sizzling sound from overcharging it in December 2002 reduced its capacity but that's a lot of years to be doing work without any complaints.
Next up is the whole house solar electric system. I have learned a lot from the experiments carried out on this small system. I feel confident that I can design and build one to power my home. I plan to integrate electrical power from it one-step-at-a-time. In other words I'm not going to just plug in an arc welder, the dish washer and anything else electrical and just throw the switch. It too will be an experiment. More learning!
Night
This is typically when you use the batteries to provide light around the house, TV, radio and other uses.
Bulk Charge
When the sun comes up over the horizon in the morning the charging cycle begins. Direct current (DC) from the solar panels wakes up the Charge Controller which sends current into the batteries. Battery voltage starts rising slowly as the sun rises higher in the sky increasing the current. At this time the solar panels are producing as much power as they can.
Absorb Charge
When a set point voltage is reached, about 14.4 volts, the current begins to decrease which stops the voltage from rising any further. During this time the batteries are absorbing, or accepting a full charge. This process removes sulfate from the lead plates and returns it into the battery acid solution.
Float Charge
When the current reaches a specific minimum value the battery terminal voltage is reduced. At this point the batteries are completely charged. While there is still sunshine available the Charge Controller will hold the batteries at an elevated voltage to simply keep them full. This voltage is about 13.7 volts at 70 F degrees. When the sun goes down the cycle starts all over again at night, discharging the batteries.
Ongoing improvements
8 months later, August 2004...
The 12 amp hour battery is working very well for me but I want to increase the size of the system once again for some serious power capabilities. I want to operate 12 volt lights in the computer/workshop room at night.
The new battery has 33 amp hour capacity. Adding it to the already existing 12 amp hour and 5 amp hour batteries makes the total 50 amp hour capacity. Do you see the difference here? I'm sizing the battery to the load, the preferred method of designing a system.
I'm purposely not increasing the solar panel array size just yet. I want to see how this new ratio of solar panel size works with this size of battery bank. 10 watts of solar panel, and now 50 amp hours of battery capacity. This gives the system a 1:5 ratio, that's a wider spread than I originally figured would work. We'll see, it's still an experiment after all.
The new battery has 33 amp hour capacity. Adding it to the already existing 12 amp hour and 5 amp hour batteries makes the total 50 amp hour capacity. Do you see the difference here? I'm sizing the battery to the load, the preferred method of designing a system.
I'm purposely not increasing the solar panel array size just yet. I want to see how this new ratio of solar panel size works with this size of battery bank. 10 watts of solar panel, and now 50 amp hours of battery capacity. This gives the system a 1:5 ratio, that's a wider spread than I originally figured would work. We'll see, it's still an experiment after all.
[ photo of 33 amp hour battery ]
2 1/2 years later, February 2007...
I'm adding another 2 solar panels to the array to charge the batteries a little quicker. The solar panels are mounted in a new location. Stronger mounts, aesthetically pleasing to the neighbors (and me too), and it will allow future expansion. The first mounting for the two panels was just attached with C-clamps, this one uses 3/16 wood screws with pilot holes. Yeah, lots more professional.
[ photo of 4 solar panel array ]
I used B-Line angle steel to fabricate the mounting brackets for the solar panel array. That stuff is wonderful for building things. It's the adult sized Erector-Set, remember those?
[ photo of solar panel array mounting brackets ]
As I suspected the 1:5 ratio was a little too far to go. It kept the batteries charged pretty well but it was a struggle at times. Cloudy days would keep them from being fully charged. Several times in the 2 1/2 years it would take two days for the batteries to rise to the absorb voltage, only making it into the 13 volt range on some days.
The additional solar panels will bring the system up to 20 watts total, 1.2 amps maximum charge current. This puts the ratio at 1:2.5 with 20 watts and 50 amp hours. A lower ratio like this will charge the batteries a little faster. On cloudy days it will bring the batteries to a higher charge level too.
At this point the system ratio is below my gut feeling ratio of 1:3 being close to optimum. To adjust this a little I'm going to add four 7 amp hour batteries to the system for a total capacity of 78 amp hours. That moves the ratio back to almost 1:4, where I lived with it that way for about three years from about 2003 to 2007. I was pretty much satisfied with the performance back then so adding these unused batteries to the system should be OK.
78 amp hours of battery capacity is some serious power storage! I'm going to be able to use some of this power on my electronics workbench to do some real work. This is getting funner. Is that a real word?
almost 2 years later, October 2008...
At this point I have a large 12 volt system powering my home. The system is teaching me about wire size and voltage drop across terminals and connections. It's pretty amazing how easy it is for 12.7 volts at the battery to become 11 volts at the load when the circuit includes a few connections in between. The idea here is to start with a higher battery voltage. Then by the time the electricity arrives at the load the voltage drop will be a smaller percentage of the original battery terminal voltage, you get more power to the load.
[ photo of solar panels powering my home ]
The system is working very well. In addition to using the 12 volt system for room lighting, I've been using it at my workbench to power electronic circuits I'm working on, charging 1.5 volt AA and AAA batteries, and all sorts of other things. Having a larger battery bank seems to keep a lot of power in reserve. Which brings me to the next adventure...
Danger High Voltage
It's time to try a 48 volt system. I've been thinking about energy storage and power. The relationship of voltage and power is interesting. All of the calculations for solar panel watts to battery amp hours that I have spoken of here are based on 12 volts. Those calculations don't work for 24 or 48 volt systems. So I want to see what higher voltage will do besides have less voltage drop on a long run of wire.
[ photo of 12 solar panel array ]
Solar panels for the 48 volt system are just an extension of the four 12 volt system panels you saw in the photo just above this one. Eight solar panels are used for the 48 volt system making the whole array twelve panels long. They are all the same Uni-Solar brand 5 watt, 12 volt 300 milliamp output.
[ photo of 12 solar panel mounting brackets ]
Four of these solar panels wired in series connection will give you 48 volts output at 300 milliamps. I want to build this system using four 33 amp hour batteries like the single one I added to the 12 volt system in 2003. It's a nice size and I have the perfect size shelf space just waiting for them. I need to make sure there is enough solar power to charge this battery bank. So, what have we got here...
- Solar panels: 8 times 5 watt solar panels is 40 watts, 300 milliamps times 2 is 600 milliamps.
- Batteries: 4 times 12 volt 33 amp hour wired in series, that's 48 volts at 33 amp hours. But this must be calculated in 12 volt terms, so that's 4 times 33 amp hours for a total 132 amp hours.
- Charge rate: 132 amp hours divided by 40 (C/40), that's 3.3 or a ratio of 1:3.3. Great!
We're right inside my gut feeling ratio at 1:3.3. The solar panels and battery choices have put us right in the middle of the sweet spot for building this system. I currently don't have any big plans for a load on this system, so I'll just take it slow and see what develops.
An interesting application for the high voltage of this system presented itself. I have a collection of antique radios, some of which are battery powered. They usually take a low voltage high current A-battery to operate the filaments in the vacuum tubes, and a high voltage low current B-battery to operate the plates in the tubes. So... I connect a 1.5 volt AA cell for the filaments, and the 48 volts from the solar electric system for the plates. Yeah, I did it and it works great! Now this isn't really a "load" on the 33 amp hour batteries because these radios only draw 0.01 amps, that's 10 milliamps! One of these radios would never deplete the battery bank in this system. But it's kind of fun running the old vacuum tube radio with the 48 volt system.
Here is the 48 volt Morningstar Charge Controller mounted in a NEMA metal electrical box just like I did with the 12 volt system.
[ photo of 48 volt Charge Controller NEMA box ]
The LCD display in this photo is showing 52.6 volts and you can kind of see the same on the lower meter on the right. That's because the normal terminal voltage of a 12 volt battery is 12.7 volts when fully charged, usually higher than that when it's freshly charged. So the 52.6 divided by 4 is 13.15 volts. It's this high because when I took this picture the sun was just setting over the horizon after being held in float mode at 13.7 volts (54.8 volts) in the afternoon.
Summary of things learned
Each one of the systems described above was a methodical step in discovering how many watts of solar panel is required to charge a battery of a specific capacity. That is why I "lived with" each system for a period of time. I wanted to know how each design would work through the worst part of the year: Winter. When a system performed well in the Winter, it's a good design. On the other hand if it performed poorly in the Winter, something needs to change. Either reduce the load or increase solar panel size.
Step 1
5 watt solar panel, 5 amp hour battery. This is what I call a one-to-one ratio. This was my starting point. Unfortunately I "toasted" the battery right from the start, so this experiment didn't go as planned. My attempt to salvage this experiment led to the next step.
Step 2
10 watt solar panel, 5 amp hour battery. This is a two-to-one ratio. If 1:1 can't keep up with the loads then we'll just put more power into the battery with a larger solar panel array. What I took away from this setup is my gut feeling ratio that 2:1 of solar to battery is probably too high. The battery gets recharged OK but there isn't enough battery power to satisfy the load yet. This inspired the next step, bigger battery.
Step 3
10 watt solar panel, 12 amp hour battery. We're back to a one-to-one ratio. 12 amp hour battery is about two times the capacity I started out with. This system worked very well through the Winter season. So well in fact that it led me to the next experiment.
Step 4
10 watt solar panel, 33 amp hour battery added to the existing 12 amp hour gives us a total 45 amp hours. Now the ratio is one-to-four, 10 watts by 40 amp hours (rounded off). I lived with this system through 3 Winters and found the performance to be remarkably good. The battery capacity pulled me through some long cloudy spells without fully charging. My impression is the ratio of solar panel to amp hours needs to be a little closer than 1:4, maybe 1:3.
Step 5
20 watt solar panel, 12 ah + 33 ah + 7 ah + 7 ah + 7 ah + 7 ah = 73 amp hours. Here I added in the batteries from the UPS units. Now the ratio is about 1:3 (20 watts by 60 amp hours), 73 is half way between 1:3 and 1:4.
10 watt solar panel, 33 amp hour battery added to the existing 12 amp hour gives us a total 45 amp hours. Now the ratio is one-to-four, 10 watts by 40 amp hours (rounded off). I lived with this system through 3 Winters and found the performance to be remarkably good. The battery capacity pulled me through some long cloudy spells without fully charging. My impression is the ratio of solar panel to amp hours needs to be a little closer than 1:4, maybe 1:3.
Step 5
20 watt solar panel, 12 ah + 33 ah + 7 ah + 7 ah + 7 ah + 7 ah = 73 amp hours. Here I added in the batteries from the UPS units. Now the ratio is about 1:3 (20 watts by 60 amp hours), 73 is half way between 1:3 and 1:4.
Charge current into a sealed lead acid battery seems to work best when it is about 1/40th to 1/60th that of the marked battery capacity. That doesn't appear logical or intuitive. It just seems too low to be any good. But that's what my experiments seem to bear out.
I have lived with the Step 5 setup now for 6 years. It works very well.
Step 6
The 48 volt system. 40 watt solar panel, 33 amp hour battery bank. I initially thought 1:3.3 ratio was right in the middle of the sweet spot. After 4 years of use, 2008 to 2012, I can tell you that I was right! This little system puts a good charge into the batteries even after I've used a fair amount of power out of them. Today I have a good size load for this system which is used daily. A 48 volt to 12 volt DC to DC converter which I use to power the DSL Router connected to my computer. The converter pulls about 270 milliamps at 48 volts while the Router pulls 400 milliamps at 12 volts from the converter. And, yes, I'm still powering the old vacuum tube radio on 48 volts too.
2012 updates for both systems
I am happy with both systems described here, but some new developments have presented themselves and I decided to make changes to both the 12 volt and 48 volt systems. Here's what happened.
48 volt system
January 2012 my boss and I buy a pallet of 200 watt 24 volt solar panels for a fraction of the price either of us paid for our originals several years ago. Long story short, I ended up with a broken solar panel from that purchase, completely shattered glass. In another blog post here you can read about how I repaired it, and how I rewired it for 48 volts. That broken solar panel now powers the 48 volt system. At about 125 watts, due to the damage, and 132 amp hours of batteries, that brings the ratio up to 1:1.
January 2012 my boss and I buy a pallet of 200 watt 24 volt solar panels for a fraction of the price either of us paid for our originals several years ago. Long story short, I ended up with a broken solar panel from that purchase, completely shattered glass. In another blog post here you can read about how I repaired it, and how I rewired it for 48 volts. That broken solar panel now powers the 48 volt system. At about 125 watts, due to the damage, and 132 amp hours of batteries, that brings the ratio up to 1:1.
12 volt system
Since I removed the eight 5 watt solar panels from use on the 48 volt system to use the broken solar panel in their place, they would be sitting idle. I rewired them for 12 volts and added them to the original 4 panels on the 12 volt system. Now we have twelve 5 watt panels feeding the 12 volt system, 60 watts! With 60 watts of solar panels and 78 amp hours of batteries, that brings the ratio up to 1:1.
We lost the original 5 amp hour battery in November 2012, it finally died. I noticed an unusual charging profile on the 12 volt system where the voltage would vary during the bulk charge phase. One-by-one I disconnected the batteries attached to the system. That's when I noticed the 5 amp hour battery felt a little warm. Warm! Yes, and looking at the case I could see a small bulge on one side. Yikes! As a result of these observations I removed it from service. The voltage measured about 10.5 volts. It's done. Gone. I put it in a safe place in case it decides to "GO" or leak or something. It will be recycled on my next trip to the battery store.
That little 5 amp hour (abused) battery has been in service from December 2002 to November 2012, that's 10 years! I guess the sizzling sound from overcharging it in December 2002 reduced its capacity but that's a lot of years to be doing work without any complaints.
Next up is the whole house solar electric system. I have learned a lot from the experiments carried out on this small system. I feel confident that I can design and build one to power my home. I plan to integrate electrical power from it one-step-at-a-time. In other words I'm not going to just plug in an arc welder, the dish washer and anything else electrical and just throw the switch. It too will be an experiment. More learning!
Photo Gallery
Gregg Scholfield 6-7-2013
No comments:
Post a Comment