Tuesday, May 25, 2010

Low Voltage Beep

I ran the air conditioner for a short while yesterday, along with most of today.  Around 3:30pm, I got a low voltage warning from the inverter.  The voltage reading on the inverter was at 20V (vs 25V where it starts at).

For 5/24, I ran the air conditioner for 46minutes and pulled 0.41Kwh.  For today, I ran it 5hrs, 40 minutes and pulled 1.74Kwh before I got the beep.  That totals to 6hrs, 26minutes of run time and 2.15KWhrs.  In my last post, I had estimated my battery bank of being capable of 8.4Kwhrs (Each battery being a 2.1KWhr battery, times 4 batteries).  So I seem to be getting about the run time of one battery instead of 4 (or even 2).  

Now, to be 100% fair.. once I turned off the air conditioner, I continued to run the fan for another 1.5 hours.  Granted, the power usage of the fan is less than the output of the solar panel, so it could potentially run indefinitely.

Given the current setup, I definitely will not be able to use the air conditioner for 3 days on one charge.  At this point, I'm not exactly sure why I'm only getting 25% of my capacity out of the  system.  Perhaps I need to redo the wiring inside the battery box to more evenly draw across all 4 batteries.

As a stopgag, I can try a few other options.  The easiest would be to bring the generator back down and only run it once the batteries have ran out.  I also have a 24V charger, so I could be charging the batteries off the generator at the same time.  I could also upgrade my battery array to 6 or 8 batteries.

Another option I've been considering is one of the evaporative/swamp coolers.  Generally, these are a fan that blows over water.  They work on two concepts: 1) water has thermal mass and stays cooler longer, and 2) small drops of water will cool you down and then evaporate away.  They have a few downsides: 1) they don't actually cool the air, just make you feel cooler, and 2) they increase the humidity.  If you let them run long enough in an enclosed space, the humidity increases and the cooling effect stops.  Also, if the humidity is too high, sweat stops evaporating and you warm up either further.  Another disadvantage is that you have to refill the water from time to time.

Back to the drawing board..

Monday, May 3, 2010

Electrical setup details.

Today I thought I'd take some time and explain my electronics setup in greater detail.  A lot of this system is all about crunching numbers.  The math isn't complicated - it's all multiplication and division, elementary school type word problems.

A friend of mine likes to claim that I should be able to calculate all of this out before purchasing the first battery.  And you sort of can.  You can determine how big your system will be and how long it will operate under different loads.  But what he doesn't seem to get is that you can't always determine your load.  The bulk of my setup is in place in order to operate an air conditioner.  How much I use this air conditioner and how hard the air conditioner works changes like the weather. The compressor kicks on and off as needed and this is a variable you can only learn by testing over a period of time.  Manufacturers may have this data, but rarely is it published.  Plus, a number of variables such as humidity and thermal change make this data unreliable.  More on this later.

First, we have the battery pack.  These are 4x 12V deep-cycle batteries with a reserve capacity of 175Amp-hours each.  What is an amp-hour?  Simply put, it measures the amount of Amps a battery can produce for 1 hour before it is discharged.  So a 175Ah 12V battery can (theoretically) produce 175 Amps for 1 hour before becoming depleted.  For our purposes, we really wan the "watt-hours" of the battery.  This of course is how many Watts of power a battery can produce for one hour.  To caclulate power, we multiply Amperage times Voltage: 175A * 12V = 2100Wh (or 2.1KWh).   So each battery can run a 2000 Watt appliance for 1 hour, or a 1000 Watt appliance for 2 hours. 

 We have 4 of these batteries, divided into 2 sets of 2.  Within each set, the batteries are connected in parallel using #2 AWG wire, which increases the amperage capability from 175Ah to 350Ah, but not the voltage.  The power rating for 350Ah set is 4.2Kw.   The two sets are then connected in series, which leaves the amperage the same, but increases the voltage to 24V.  So we now have a 350Ah system at 24V, or more importantly an 8.4KWh battery system.

At this point, you may be wondering why I am going from 12V to 24V.  If all 4 were hooked up in parallel, I would still have gotten an 8.4KWh system.  The reason has to do with wiring and efficiency.  DC voltage can drop quite easily over short distances.  So if you're going to work with DC voltages, you want to get your voltage up as high as possible before transmitting it.  In practice, most people have to choose between 12V, 24V, and 48V systems.  You usually only see 48V in grid-tie systems and industrial applications.

Connected to this battery pack, I have 3 devices:
  1. A 24V 1500W A/C Inverter  - this will convert my 24 volt DC electricity into 120V house current and can power 1500W devices (with a peak of 3000W).  This is connected via #4 AWG wire.
  2. A 12 volt converter.  This little box regulates the 24V voltage down to 12 volts, which is quite useful since a large number of devices are designed to operate at 12 volts.  The box claims to be rated at 30A, but I have it fused at 25A.
  3. A Xantrex C35 - 24-volt DC charge controller.  This charge controller is capable of 35A @ 24V or 840W.  This is then connected to the solar panel.  Due to the higher voltage/lower wattage, this is connected with #10 AWG wire -- technically called MC4 Interconnect wire, which is standard on most solar panels these days.
The Inverter powers my air conditioner.  I also have another power strip connected to it that I use on occasion for small devices, such as the fan or my netbook.

The solar panel is an HQRP  24V, 180W solar panel  (7.5A).  Most guides on designing solar systems recommend that you get a solar panel array capable of producing 30% of your system's capacity.  This guideline is for a daily use system. When the guides discuss this, they typically refer to your battery bank's capacity (and the battery capacity should always be 2-3 times your load). Apparently 30% has been determined as the optimal amount for recharging batteries.  If you recharge with too much power, you could damage the batteries.  If you recharge too slowly, you could end up never recharging the batteries.

My solor panel is way below this recommendation.  At the 8.4KWh rating, the 180W panel is around 2% capacity.  If I built my array out to the full 840W that the charge controller can handle, I would still only be at 10% capacity.  However, I do not use this system on a daily basis.  During last August's super hot weeks, I ran the air conditioner 4-5 hours each day.  So on a peak long week, I will use the air conditioner for 3 x 5-hour stretches, for a total of 15hours each week.

My current air conditioner peaks around 1000W while the compressor is running.  The compressor will run for 1-2 minutes, and then run the fan for 4-6 minutes. During the one measurement I recorded (1hr, 20 minutes), the air conditioner averaged 0.31KW/hour.  A 15-hour week should pull 4.62KW, or just over half of my battery bank's capacity.   I do want to take more measurements of this load to get better figures.  This air conditioner is a 5200 BTU with an EER of 11 - see it on my equipment page.

To be fair, the inverter and wiring do loose some power, just as any system does.  Also, I will from time to time run some 12V appliances off of the system.  All of these impact my calculation.  Also, during the morning and evenings when I am not utilizing my air conditioner, the solar panel is hard at work recharging the system.  I have only run the air conditioner on a few occasions, and only 2 days in a row so far (cooler weather outside).  But on day 2, the battery pack is recharged before I start the air conditioner.  On day three, the battery pack is recharged again even though I haven't started the air conditioner.


I don't know if that will remain true when the weather heats up and the air conditioner has to work harder.   At 8.4KWh, the battery pack can sustain around 0.5KW/hour for 15 hours, then it would need completely recharged. If hotter weather begins taking me beyond 0.5KW/hour, my best solution is to increase the size of the battery bank to match.  If I increase the size of the battery bank and find that the batteries are not getting recharged in suitable time, I will have to increase the size of the solar array to shorten the recharge time.  Neither is exactly cheap.  The batteries run around $85, plus the expensive wiring (call it $100/battery).  I also need to find room for the batteries (somewhere where I can tie them into the existing set).  The solar panel itself can run $500-$600, but can be easily mounted on the roof with the other one.


The 12V converter was purchased on ebay for around $30 and comes from Hong Kong. It is rated at 30A, but I have it fused at 25A (the fuse goes between the converter and the 24V battery pack) for added safety.  The converter has two accessory wires which I have no use for (but come in handy for a radio setup).  The 12V output goes into a fuse block.  I paid way too much for this on ebay for $48 shipped.  If you are smart, you'll go with my original idea and pull one off a junkyard car for much less, perhaps free.  This fuse block has two independent positive buses, a negative bus, and 14 fuse slots.  I could technically run separate 12V and 24V devices off of this, but I have no 24V devices, so the two positive buses are linked together.

Currently, I have 3 x 5A circuits coming off of the fuse bus.  One goes to the fan vent in the ceiling.  Another goes to a set of lights over the desk (which I pulled out of the van originally).  The final one goes to a 3-port power socket (cigarette lighter socket).  This lets me plug in common 12V car accessories such as cell phone chargers.