Thursday, March 4, 2010

Inverter and Air Conditioner Test


The confusion below regarding AH and run time has beencleared up.

A 175Ah battery provides 175A @ 12V (2100W) for 1 hour. 

The portable 9000 BTU air conditioner actually pulls 9-10A @ 115V (900-1000W) (giving it an EER or 9).   However, my inverter is rated at 1500W with a peak of 3000.  I believe at the moment the compressor kicks on, it pulls around 2000W and exceeds the batteries capabilities, coupled with losses from the inverter.  It would be an interesting idea to take a fully charged starting battery with a high CCA rating to see if it could activate the compressor for a few minutes.

I have ordered a 5100BTU air conditioner with an 11.7 EER value.  This should pull 435W when the compressor is running.  It is a window unit, so I'll have to make a special box/duct for it.

Fun Fact: Take an air conditioner's BTU and divide it by its EER to get the electrical power usage (watts).

Today (at home) I did a somewhat expensive test.  I connected to 175Ah deep-cycle batteries in series to a 24v / 1500W power inverter.  I then used this to power my air conditioner.

A few things first.  It's 30F outside, so this test would not work in my van.  I discovered it wouldn't even work in the basement of my house (too cool).  I left it running for 4 hours in my basement connected to the battery/inverter.  It was running great.  So great that I couldn't believe it.  After 4 hours, the voltage from those two batteries was holding at 22.75.  I hooked up my kill-o-watt and discovered I was only pulling around 100W, sometimes less -- it was using about as much as a bright light bulb.  I had been taken.  It was just pretending to run.  In the process, it caused me to run the fan against my batteries for 4 hours, reducing their fully charged status.

So ultimately, I had to bring it up and set it next to my radiator.   First, I ran it on the utility power.  This time, the  kill-o-watt showed 600-650W, on occasion kicking up to 750W.  Not bad.  I then brought up my battery setup and ran it off of that.  Similar wattage.  However, this time I only was able to run it for 20 minutes.  I suspect that it would have ran longer had I not drained the batteries all day.

So I am going to charge these batteries and try this again.  However, in the meantime I decided to do some theory.  My batteries are 175Ah.   Connecting them in series only increases the voltage -- the  two should be considered one 24V battery.   Amp-Hour is a rating of how long a battery can supply 10Amps in an hour.  To me, this means I should take my wattage, calculate the Amperage, and divide the Ah rating into it.  According to my inverter manual, I should take my wattage, divide it by 10, and divide the Ah rating by that number. At 750 or 650 watts, the manual's method comes to 2.3 or 2.9 hours, respectively.

The amperage draw varies from 5.7 to 6.9. I have read various documents on Ah rating, and I am more confused than when I started. From various sources, I am supposed to divide my Ah rating by the Amperage draw of the device.  The thing about that method is though, I would be looking at 175/7 for 25 hours.  In this case, I expect the manual's calculations (approx 2.5 hours) to be more accurate.  I also checked other batteries out on the web.  US Battery sells a 130Ah battery and calculated that it will run a 25A load for 205 minutes (3.4 hours).   Since my load is approximately 1/4th of that, I should get 820 minutes (13 hours).   I don't believe that I will get anywhere near that amount of runtime.

I also believe that Ah requirement ratings are explained deep in a textbook that takes you  several weeks of studying.  This knowledge has not made it properly to the web, but has been "summarized" by a lot of people that don't understand the subject matter.  To make it worse, battery manufacturers spin these details to make their batteries look better.

Here is what I know.  My device draws 600-750W (or 6-7A) during normal operation.  It needs to run for approximately 5 hours/day over the course of 3 days (15 hours).  In theory, that is 90-105 Amp-Hours, so I don't see why a theoretical 175Ah battery won't provide that charge.

In reality, this power should have been provided using my 12V inverter.  We are dealing with wattages much, much lower than 1000, and my 12V inverter was a 1500W inverter (with 3000W peak).  Now, I believe a single 12V battery is going to have trouble exceeding 2000W, but I am not hitting above 800W.

In this case, there are a couple possibilities.  1) My Cobra 1500W inverter is junk.  2) The #6 AWG wires going to the inverter are a larger impediment than suspected, or 3) My Kill-O-Watt meter is junk.

In any event,  I can run this air conditioner off of 2 12V batteries and 1 inverter.    This summer, I am going to implement the following:

  • Install the two batteries ($90ea) in the van in their current setup with the 24V inverter ($190) to run the air conditioner.  Depending on further tests this week, I may add more.
  • Solar Panel 185W, 24V solar panel (approx $600)
  • Charge controller, 24V (unknown price: $80 - $150)
  • Estimated cost: $1,060.
I know that the 2 batteries will power the air conditioner.  I don't know for how long.  I suspect it could be from 2-4 hours.  I will be conducting additional tests (hopefully this week) to figure out how true this is.  All the calculations in the world won't show me what a real life test can do.

Most posts agree that you need about 30% of your operating load to recharge a battery bank in any decent time.  My solar panel is on the lower side of that 30% (30% of 600W = 180W,  30% of 750W = 220W).  However, I will be operating the air conditioner considerably less than 50% of the daylight time.  Also, I hope that once the air conditioner cools down the area of the van, it will drop to fan and be in the 100W range (at which point, the solar panel will power it completely).    I also plan to make use of a fan and install a vent in the roof to let heat escape.

If more power is needed, additional batteries, or even a second solar panel could be added to the system.  I would much  rather start small and improve on it than to go overboard on this system.