From: John De Armond
Subject: Re: Air conditioning with minimized generator usage (batteries and
inverters again, yes..)
Date: Thu, 03 Nov 2005 15:02:12 -0500
On 3 Nov 2005 05:27:58 -0800, "daveola" <email@example.com>
>I know this has somewhat been hashed multiple times, and I understand
>that generally A/C units cannot be run on a normal battery bank with
>normal inverters and no genset.
>But what if I had a large battery bank and inverters and I had a
>genset but I was trying to cut down on genset usage as much as
>Possible design: (this is for an MCI bus conversion)
>A) Battery bank: 12V, 1760Ah (yes, that's 16 Trojan T-105s)
>B) Two Coleman 13000BTU A/C, hoping to generally only use one
>C) Large genset 10kW or more
>D) Biggest chargers I can find (is 400A possible?)
>E) Large inverters: 4kW or more (and probably surge caps to help start
First do some math. 1760ah * 12volts is 21,120watt-hours or 21kwh.
You'll want at least one 13.5kbtu AC unit on the bus. From Camping
world, the spec says it draws 1595 watts under normal conditions (70
inside and 90 outside, if I recall correctly), much more under very
hot weather. That means that under standard conditions, you could run
21,120/1595 = 13 hours on a charge, assuming you took the batteries
all the way down. This does not include the effects of Peukert on the
battery capacity. I'd figure 8-10 real world hours.
13.5kbtu will be marginal on the bus so you might want to find the
specs for a 15kbtu unit and re-do the math.
By the letters:
A) I'd not use the T-105s. Since you have the space, I'd go up to
the larger and more rugged T-225 (I think that's the right number),
the ones made for floor scrubbers. They're a little taller than the
T-105s. They'll last a LOT longer. This is experience from using
them in an EV talking.
Also, I'd NOT do 12 volts. The losses are too high that way. I'd go
to at least 48 volts, where lots of telco inverters are available.
You'll get longer battery life that way, too.
B) One obviously won't do the whole rig. If you use 2 then cut my
calculated time at least in half.
C) At least. See below.
D) A 50kw charger is available.
These have a hugely good reputation in the EV world. 50kw is his
largest standard one. Rich Rudman is building a 100kw one for some
industrial customer right now.
The most important part at this power level is the power factor
correction (PFC). A conventional charger only draws current for a
fraction of each line cycle, when the incoming instantaneous voltage
is higher than the pack voltage. This results in the production of
harmonics and poor power factor. Not a terribly big deal on shore
power but a REALLY big deal on a generator. Rich (the owner of
Manzanita) is involved in EV drag racing. They rapid charge the
dragster between rounds at the 30kw level. He found that a non-PFC
charger will fully load the generator at about 60% of the apparent
rating. That is, only about half as large a charger as one would
think can run from the generator. Thus, a PFC charger is vital. Rich
is the only manufacturer of a charger in that power capacity that I
know of that is at all affordable. About $1500, I think.
There is another problem with ordinary rectifier chargers at this
power level. On a single phase circuit, the input voltage decreases
to zero 120 times a second. This results in high ripple current in
the output. We EVers have found that this ripple is very hard on
batteries. Not a major issue when you're charging 10 or 20 amps but
when you're up in the 1C or better charging rate, it matters a lot. A
three phase charger doesn't have this problem because the voltage
never goes to zero.
There is a significantly cheaper and much better method, however. Make
the power at the desired voltage. Essentially an engine driving a DC
generator or an alternator followed by a rectifier. Generally the
3-phase alternator and bridge is more reliable and can be cheaper.
The most compact system is a permanent magnet 3 phase generator
attached to an engine and followed by a 6 diode bridge - almost like
in a car alternator. In fact, a truck alternator can be used, though
the wound field makes it a bit less efficient. Charging rate is
controlled via the engine throttle. An alternative is to generate at a
higher voltage and let Rich's charger (it works on AC or DC input) do
the charge control but that is the more expensive route.
There are several people out there, myself included, that make custom
chargers. Contact me directly if you want to discuss this. Depending
on the capacity, the price would run somewhat less than a generator
and a Rudman PFC charger.
E) at least 48 volts and a telco or large computer UPS is the way to
go. These inverters have much more surge capacity, both because the
batteries can supply it and because they're designed with larger
margins than consumer inverters.
I'd modify the ACs with unloader valves to make them easier to start
too. This is a normally open solenoid valve connected between the
intake and exhaust of the compressor. When the compressor is off, it
opens, equalizing the pressure between the two. A time delay relay
delays the closing of the valve so that during starting, the motor is
having to overcome only friction and inertia. I've used this
technique extensively to get reefers and ACs to run on marginal
circuits, both generator and shore power.
>My hope is to generally run everything off the battery/inverters, and
>then have the genset autostart to charge them back up - my goal is to
>have the genset run as little as possible (a few hours a day would be
Completely doable. Tom Fisher (I don't think he's on this group
anymore) designed a VERY nice auto-start system around a $150
programmable logic controller. His is much more versatile than the
commercially available ones. If we get that far, I can put you in
contact with him or I could design you one. Or you could do it
yourself - PLCs are very easy to work with.
>The big questions I have to answer are:
>- How can I estimate (even back-of-the-envelope) the duty cycle my A/Cs
> will need?
> . What kind of duty cycle do other people see on their A/C units?
> Do they need to run both for the whole day? 4 hours a day? ??
> . I was up in the desert in 90+ degree weather with no insulation on
> my bus and it was reasonably comfortable with only one A/C unit,
> on low, running most of the day. I was surprised that I didn't
> need both units. Do I just have a higher tolerance? How can I
> approximate heat loss on my bus (I want to keep most of my windows)
You have a higher tolerance than me. My 15kbtu runs all the time in
hot weather keeping my 20 ft rig at 70 with it 90+ outside. The only
way to know what yours does is to measure it. The easiest method to
measure the duty cycle is to get a 120 VAC wall clock and hook it in
parallel with the compressor. It runs only when the compressor is on.
Run it for a known period. The clock time divided by real time is the
duty cycle. I suggest a wall clock instead of an hour meter because
it has better resolution and is dirt-cheap.
Modeling the heat loss on the bus will be far more complicated than
it's worth. I'd just do a duty cycle test. Once you have the duty
cycle for a given temperature (hopefully near your actual operating
temp) then you can use charts in the ASHRAE handbook to extrapolate.
> so I can approx BTU needs per degrees of temp change I desire?
>- Can I just buy a number of 100A chargers and hook them in parallel?
NO, see the power factor discussion above.
>- Does anyone make a large genset that just has a 12V alternator on it?
> It seems silly that I convert to 120V from the genset, then convert
> back to 12+V for charging, then run the batteries to the inverter
> back to 120V. A 10kW genset that had an 800A 12V charger (about
> 10kW) and no 120V would be absolutely perfect.
yep, except that you'll want the higher voltage.
>- Failing that - how many chargers can I hook up to a 10kW gen?
Depends on whether the chargers are PFC'd. I don't think the
Progressive Dynamics one is. The IOTA one is, I think. You'd have to
ask. With PFC chargers, you can pretty much stack 'em up to the
>- Do square wave inverters really shorten A/C life span?
Nope. First off, nobody I know of makes a true square wave inverter
anymore. I ran all kinds of stuff from a true square wave inverter
(Tripplite) many years ago. Most everything except a microwave oven
worked just fine. The modified square wave inverters now so common
output a wave that resembles a sine wave well enough that the motor's
inductance will filter out the higher order harmonics with no
problems. At worst, it will run SLIGHTLY hotter and perhaps you'll
hear the sharp edge of wave.
>- Would this actually work?
Sure. Computer centers everywhere do essentially that. large battery
bank, suitable inverters and efficient ACs.
>>From a math perspective it seems okay - for days that I use only one
>A/C unit for 4 hours of the day, plus other appliances in my house, I
>figure I need about 24kWh of juice for the day. That averages 1kW
>each hour, and the batteries hold about 21kWh. If I recharge at 60%,
>then I need to recharge about three times a day (to regain ~8.5kWh).
>A 10kW genset should be able to do that in about an hour or so, yes?
>I'd expect a max load of 2kW (except for the A/C starting surge which
>is brief), so at worst my batteries would be discharged at a 10hr
>rate, so the battery bank seems big enough to not be overwhelmed.
Good batteries will show a charging efficiency of 90% or better. An
E-meter/link 10 can actually measure and compute that for you. The
parameter in the E-meter is the CEP. You're going to be discharging
at more than a C/10 rate but that's fine. The batteries can handle
>Sorry about the lengthy, questiony post - if anyone has answers to
>even some of the questions, please let me know. I'd love to get some
>experienced feedback before laying this kind of $$ down.
No problem. As an EV'er, this is one of my areas of specialty and
interest. The power levels you're proposing are normal for lead-acid
battery powered EVs. The only problems I foresee are weight and
space. Your bus should handle both easily.
At the risk of sounding self-serving, I highly suggest retaining the
services of someone experienced in this area. It'll save you a bunch
of money in the long run. Rudman can help you if you want to go the
generator/charger route. I can help if you want to go the direct
generator route. I love doing this stuff so my consulting rate is
quite cheap. Contact me offline if you're interested.
One other comment. If you have the money to spend, you might want to
consider a high capacity LiIon pack. Probably 5X at least, the cost
of a lead-acid pack but probably only 10% of the weight and a much
longer cycle life. The chicoms are making some decent LiIon batteries
at quite reasonable prices. I can put you in contact with some folks
who have extensive experience with high capacity LiIon batteries and
From: John De Armond
Subject: Re: Air conditioning with minimized generator usage (batteries and
inverters again, yes..)
Date: Fri, 04 Nov 2005 00:53:53 -0500
On 3 Nov 2005 15:22:00 -0800, "daveola" <firstname.lastname@example.org>
>Will Sill wrote:
>> IMO (not to be insulting) you're nutz.
>I thought being nutz was a prerequisite to doing
>a bus conversion in the first place. :-)
>> An AC makes as much noise as the really quiet diesel gensets
>I don't believe this is true. I'm having trouble finding exact dBA
>numbers for AC units, but I think they are around 50dBA? That
>would make them at least a few times quieter than a quiet diesel
>genset (around 67dBA).
Dave, you can't directly compare the two because the measurement
conditions are not the same, nor are the noise characteristics
The inverse square law applies to point sources of sound. That means
that the first few feet make a lot of difference. Consider how close
your ear is to an overhead AC air outlet compared to the generator
located in a (hopefully) sound dampened box somewhere outside the
I can tell you that with my roof AC going, my Generac Quietpack (the
quietest generator I've ever witnessed) sitting on a cart right
outside my window is inaudible. I can hear my built-in Generac Impact
inverter generator but it's in a box right under my couch. The noise
all adds up and I'd love to be rid of the genny noise at night.
Where your plan will really pay off is when you have a load to run,
say, the AC and/or an electric oven or something similar while you're
outside in the evening. Being able to do that without any generator
noise would be wonderful. I can't think how often I've avoided
cooking because I didn't want to listen to the generator run.
You ought to shell out a few dollars on a Rat Shack decibel meter.
About $40 for the analog version, I think. The meter is amazingly
accurate for the money. I have a General Radio sound calibrator that
I periodically test my meter against. In 15+ years of ownership I've
never had to touch an adjustment.
A db meter will let you measure and characterize sound sources and
tell you when something you did made a difference. Being able to see
the effect of, say, applying some damping rubber to a sheet metal
panel will greatly speed the quieting project.
>> 16 Trojan T-105s with a combined weight of over 1 1/2 TONS
>Incorrect. Total weight is 976 pounds, about 1/2 ton.
>Not a problem.
>> you're wasting fuel with your scheme because of the inherent
>> inefficiencies of batteries and inverters - not to mention chargers.
>Yes, but consider the efficiencies of running a genset at optimal load
>and only at optimal load, and then shutting it off. That's pretty good
>compared to losing 10% on inverters and 10% on chargers. In fact,
>this is the same thinking behind hybrid vehicles.
You're right, of course, as anyone who has owned an EV or used an
off-grid power system knows. Or anyone who has read around the net,
for that matter.
Something that may not have occurred to you yet is that with a water
cooled engine, you can charge your batteries and heat your water at
the same time. This "combined cycle" setup is incredibly efficient.
For commercially made units, I've seen figures in the 80% energy
utilization range quoted. Even more heat can be stored if the hot
water tank is partially filled with phase change heat storage spheres.
This is a fancy way of describing small plastic balls filled with a
material that melts at the desired temperature. The heat of fusion is
MUCH greater than the specific heat of water. Paraffin is a
convenient phase change material, though there are much better ones,
at more cost, of course.
Tom Fisher and I worked on this quite a bit last year for his coach.
He uses engine heat, water cooled generator heat and a small diesel
fired water heater to provide coach comfort heat, hot water and engine
preheating. We had gotten through testing and were ready to order the
spheres and fabricate a tank when he had to go off and do some other
things. I haven't followed up but I bet he has the system running by
His goal was to have a single fuel (diesel) coach and get rid of the
propane. He was essentially there the last time we talked.
You could store up a day's worth of heat for hot water while charging
the batteries. That would save a ton on propane.
Before I go, one other comment. Trying to argue with Sill is like
trying to teach a pig to sing. It invariably fails and it entertains
the pig. He never lets ignorance impede his arguments. Do like many
of us have and kill file his posts or simply ignore them.