MrWizard wrote:
Still a no go for A/C, oh a big inverter will start it
But for a roof to unit your looking at 100 amps or more to the inverter to get the 1400 Watts for the a/c, that means approx 2 hours runtime to 50 percent battery charge
Welcome and have fun
True. Which is why the RV rooftop AC has to go in favor of a >20 SEER mini-split. Let's do the math.
SEER mean Seasonal Energy Efficiency Rating or BTU moved divided by watts consumed.
First, let's look at an RV AC. I had this project planned on my last rig when I got sick. I measured the Coleman Mk III 13.5kBTU rooftop AC. It drew 13.5 amps with a power factor of 0.7 (one reason so much generator is required to run it). So the watts are 120 volts * 13.5 * 0.7PF = 1134 watts. BUT. The generator has to supply the total volt-ampers or 1600 VA. The SEER is 13,500 BTU / 1134 watts = 11.9 SEER. This calculation does not take into account the approximately 60 amps of inrush current for starting.
Now let's look at a 13.5kBTU 20 SEER minisplit. The watts draw is 13,500 watts / 20 SEER = 675 watts at full speed. In other words, a modest 1kW inverter, though I'm using a 2kW because I have several on the shelf.
Furthermore, since all the motors are variable speed, driven by 3 phase inverters, there is no inrush. With the Mitsubishi unit I'm used to installing, when the "thermostat" calls for more cooling, the compressor is sped up. The condenser fan speed is controlled to maintain a constant high side pressure. The evaporator pressure is controlled by a variable expansion valve. This last feature keeps the evaporator cold even at low demand for proper dehumidifying.
Unlike an RV and most home AC units, a minisplit does not turn off and on in response to a thermostat. The speed of all components varies as the temperature deviates from the setpoint. That means that the unit runs at full capacity until the desired temperature is met. Then the temperature controller backs down the speeds until the temperature is maintained. The unit runs all the time but at reduced load.
Back to our calculations. Let's look at worst case, full load - that is 675 watts - all the time.
The 5 battery bank is (120amp-hour * 5) * 12.2 volts = 7.32kWh. To 80% DOD*, the minisplit can run (7,320*.8)/675 = 8.7 hours. A night's sleep. But that's worst case.
I have a duty cycle data logger which I have attached to my new rig's AC. In 85 deg weather, after the rig is cooled down, the unit operates 62% of the time.
So if I take that same average load of 13,000BTU * 0.62 that equals a time-averaged load of 8,000 BTU.
A minisplit's electrical demand varies close enough to linearly with load to assume so. So if I take the above 675 watts and multiply it by 0.62, the unit will draw continuously about 418 watts. That gives a run time of (7,320*.8) / 418 watts = 14 hours.
A typical day and night would be to run the generator in the afternoon (or earlier if it's quite hot outside) to cool down the unit and fully charge the battery bank. When bed time arrives, the generator is stopped and battery operation commences. We sleep about 8 hours so we will have used just a little more than half the available battery capacity.
At this point, all I lack is buying the minisplit and installing it. After I reverse-engineer it, of course. I've traditionally used Mitsubishi but a fellow HVAC friend recently pointed me to the chicom-made Pioneer brand. Same specs at a fraction of the price and available on sleazebay. This is probably the one that I'll be purchasing but with a ceiling evaporator.
http://www.ebay.com/itm/3D302289255519
It is larger than I need but that's no problem. It'll simply slow down a bit more to maintain the set temperature.
* 80% DOD does not harm an AGM battery. I've been told that by manufacturer's engineers. Through the bankruptcy, I gained access to the AVS bus company's battery data. The DOD cutoff for the buses is set at 80%. In Chattanooga, TN, their largest deployment, each bus would typically run to near 80% DOD and then be fast-charged twice a day.