Reconciling the outputs of a generator with the C/10 limits of a LA battery bank

[Ronmar]

If I was contemplating doing this I would probably go with a 3 phase head, as it should require less filtration.  But then the head would not be as usefull for powering direct AC loads if needed...

Marcus, not having done this, was the transformer to help balance the loads on the generator windings, or did you need to isolate the rectifier from the generator?

[mbryner]

Ronmar,

Totally agree about the 3 phase head -- maybe wouldn't even need the filter choke then, which would increase efficiency a few more percent.   Of course, then it takes away the benefits of having AC gen head to begin with.

If you don't have the isolation transformer, you can't bond the ground of the ST to ground of the DC charge controller input (which is also whole system ground).   You would either have the ST having a separate (different) ground or a big spark.    Ask me how I know!    I didn't think about it at first and had voltage differences of 60 V AC, and the circuit breakers kept tripping.   Good thing I designed it w/ multiple breakers at various places for safety and to make troubleshooting easier.   Also, yes, the transformer also helps balance the loads: it's 2:1,  240 --> 120,  5 kva, surplus off ebay.

Hmm, you just got me thinking while typing this.   If you use the 4:1 ratio of the transformer instead of 2:1, therefore 240 --> 60 V, the rectified voltage isn't high enough for the charge controller to charge well IIRC.   What if I take power off the 50 hz leads in the ST head but run at full 1800 rpm instead of designed-for 1500 rpm at 50 hz.   That would increase the voltage to about 288 --> 72.   (50/60 = 0.833 --> 240 / 0.833 = 288, if the conversion factor is linear.)   Then, you wouldn't have to use a resistor in the excitation circuit.   By flipping a breaker, you could jump between charging via rectified AC and regular AC for backup.   Any comments?

[Tom Reed]

That sounds like a very good idea. BUT do you loose any additional efficiency by using a 4:1 instead of a 2:1 xformer???

[mbryner]

Why would you lose efficiency just going from 2:1 to 4:1?   (I don't know, yet.  Researching.)   Of course the transformer has to sized correctly to handle the higher amperage of the lowest voltage.

[mbryner]

Good diagrams of single phase vs 3 phase rectification (if they don't show, go to wikipedia article on bridge rectifiers):

[Lloyd]

Simtech,

What batteries are you using?

I think you are laboring under some misstatements as to battery charging and source size. Most lead acid flooded types can easily accept a charge current of 25% without ill effect. As long as the charge regulation is voltage controlled, and temp compensated during the bulk charge cycle, and then current /voltage controlled during the absorption cycle.

So a 750 am hr bank can accept @ 25% 187 amp charge current. I have in my bank which is 720 amp hr, charged as high as 250 amps during bulk.

And if you use an AGM or gel bat they can take even higher current.

The other thing to consider is that the charge source, may have to provide loads for the inverter, while still charging the battery.

Lloyd

I'm in the design phase and I've bumped into the disparity between the amperage that the generator can generate vs the C/10 limitation for flooded lead-acid batteries.

For example the generator (after feeding through the controller) can output 111 amps at 36V, however the 36V 750AH bank can only take 75 amps without shortening the life of the batteries.

The genny output will be fed through a outback charge controller which I trust will not overcharge the battery bank, however this leaves the genny only running at just below 50%.  From what I understand optimum loading of a generator set is approximately 80% of max load.

There's a lot of potential power not being utilized.

How do you guys resolve this?

Planned specs:  Listeroid 8/1 derated to 6/1 pushing a 5KW ST AC head rectified to DC feeding a Outback power Flexmax80 which charges the battery bank.   I chose to rectify (and heavily filter) the AC output of a ST head because that gives me a failover option of using an 1000 watt inverter/charger that I already own for the bank and passing the AC load directly to the listeroid gen set.
The reason I prefer recified AC power is my understanding that ST heads AC waveform is rather dirty, and my load contains sensitive medical life-support equipment (a home peritoneal dialysis unit).  Thus a separate inverter for my load.

Edit:  Sorry boogered the post, remnants of the marine alternator backup plan that I thought I'd removed.

[Simtech]

Lloyd:  All the various battery maint documents that I've read state not to exceed the batteries C10 rating for maximum life of the battery bank.
Of course it will take a faster charge, but at the expense of premature wear on the batteries.

mbryner:  If I understand you correctly, you bonded your inverters ground to the ST's ground via the isolation transformer?  Why didnt you bond directly to it instead? 

Also why did you go the transformer route at all - wouldnt it have been easier to wire the 4 poles of the ST head in parallel (giving you 120AC) and avoid the losses inherent in a transformer?

If all else fails in an emergency with 120AC from the ST head I can get power into the battery bank using a dang Black-and-Decker automotive smart charger (of which I have 2 of) if I have to (after a reconfig to 12V of course).

Thanks for all the constructive comments so far.  I'm a bit dissapointed that all I'm going to get is 2000 watts into the batteries w/ ~4.5hp of input (6hp derated 25%) but 2KW delivered for 8 hours is still 16KWH which is actually a lot of power, more than the household uses in a standard day.

I'm particularly drawn to the lister(oids) because of its multi-fuel capability.  If my mass quantities of free hydraulic oil doesnt work out, I can still use home heating oil - which costs less than even off-road diesel.  I have a few ideas on ultrafiltration of waste motor oil using filters and electrophoresis.  It'll be fun to tinker with anyway.

[Tom Reed]

Rectified 120vac is around 180-190vdc, it had to do with RMS or something like that. It's an accepted standard that it takes 2 engine HP to make 1kw.

I've ran my Listeroid on WHO (waste hydraulic oil) with marginal success. It was filtered, unheated and of an unknown source. The engine would start and run fine on it, but the injector would carbon up with up to a quarter inch long chunk in the stream. Your oil may be different, WMO does similar in my engine. It only takes a few min to clean the injector, but it needed it about every 10 hours of run time.

Also if your going to use heating oil, please add some lubrication for the IP. Around 5% B100 bio-diesel is probably best.

[mbryner]

Hi Simtech,

I keep the ST head in 240 VAC configuration because I have 240 V well pump and I want the listeroid to be able to power anything in an emergency.   Yes, you could go to 120 V config on the ST, but you will still need to put a resistor in the excitation coil circuit to decrease the voltage if you are feeding a Outback charge controller.    Rectified 120 V RMS is ~180 V peak as mentioned above.

No, the ground of the ST is directly bonded to the battery ground.   To understand the isolation transformer issue, take a bridge rectifier and hook it up to AC (without isolation transformer).   Now measure the voltage between the (-) output on the DC side and which lead you pick for ground of the 2 input leads on 240 VAC.   Also compare the (-) at the rectifier output to battery ground with that setup.   

What Tom said:  2 HP per kilowatt is generally accepted.   If you get everything tuned good, then maybe 1.5 HP / kilowatt.   I said 2000 watts because that's what I run, and it seems to be a good load for a 6/1.

Listeroids are easy to work on, fun, and will run a lot of different fuels!   Just remember to expect 1920's technology, crude machining, etc. etc. etc.

Marcus

[Ronmar]

Hi Simtech,

No, the ground of the ST is directly bonded to the battery ground.   To understand the isolation transformer issue, take a bridge rectifier and hook it up to AC (without isolation transformer).   Now measure the voltage between the (-) output on the DC side and which lead you pick for ground of the 2 input leads on 240 VAC.   Also compare the (-) at the rectifier output to battery ground with that setup. 
Marcus

Marcus
  Why would you pick one of the 240V input legs to reference to ground?  I got to thinking about your comments about the unexpected voltage issue you ran into when trying to rectify directly from the generator output and tying the inverter ground back to the generator.  The only way I could see you having an issue is if the generator center tap winding/neutral terminal is tied to ground at the generator(needed for 120-240 ops).  If the center tap is not bonded to ground, the 240V winding is floated/isolated from ground just like the isolation transformer is...

[mbryner]

Ahhhh,  yes, Ron you are correct.   I have a breaker panel where I can switch between 120 and 240 config on-the-fly w/ 2 breakers.   The center tap is neutral and, yes, it is bonded to ground.   I'm inbetween a few things at work right now, and can confirm later.   My apologies, to the original poster: you could get rid of the isolation transformer and gain some efficiency.    Now I'm interested in testing this later.....

Also, to simtech, please don't misunderstand me when I say to expect 2000 watts.   You can get over 3000 W easy enough w/ a 6/1, but above 2000 my simple thermosiphon cooling system starts to boil.   With better cooling, I can sustain longer runs at maximum load.

Marcus

[Simtech]

I'm just grateful for the comments so far!  The more you guys talk the more I learn, feel free to ramble on about anything 

Glad to hear that 2kw isnt a hard limit.  It looks like I'm going to need a variable pot for the field coils when feeding the battery charge controller.  Or maybe find a charge controller than can handle a higher DC input.

Any opinions on the ST head with the cleanest waveform?  I've been considering just spending the money on a 5KW winco generator head.  In a failover mode during a LONG term emergency it's possible my wifes dialysis unit will be running directly from the gen head and the batteries will be getting a charge via a marine alternator. (layered backup plans)

Short term power outage will most likely be handled by a generac 7.5kw natural gas genset with auto switch over - assuming that natural gas is still available and something like an earthquake hasnt severed the supply lines.

The listeroid/genny is more of a fun hobby than aligns with my long-term disaster prepping plans.

[Lloyd]

Lloyd:  All the various battery maint documents that I've read state not to exceed the batteries C10 rating for maximum life of the battery bank.
Of course it will take a faster charge, but at the expense of premature wear on the batteries.

Hi Simtech,

The C-10 is a a result of so many off grid solar sites. Because most couldn't afford a solar system that could generate 25% of amp hr capacity.

I have installed dozens upon dozens of flooded banks with charge system design of 25%. I can tell you that it will not shorten the life  of a bank. Conditioned on a proper charge regulator, that is temp compensated and and voltage controlled during bulk.

A battery bank isn't fully charged until the the return amps drops to 4% during absorption. Most damage will be done if overcharged, but with proper regulation that isn't the case.

Have a read at this link http://www.cruisersforum.com/forums/f14/evaluating-12vdc-flooded-cell-batteries-12707.html , the guy that wrote it is an EE, and formerly worked at Trace/Xantrex.

Lloyd

[mobile_bob]

in most cases, the reduction in battery lifespan (which is questionable) will be offset by 2-4 times in the savings in fuel burned.

in other words if you lose a thousand dollars worth of battery lifespan, you likely will have saved 2 to 4 grand in fuel.

bob g

[Ronmar]

I'm just grateful for the comments so far!  The more you guys talk the more I learn, feel free to ramble on about anything 

Glad to hear that 2kw isnt a hard limit.  It looks like I'm going to need a variable pot for the field coils when feeding the battery charge controller.  Or maybe find a charge controller than can handle a higher DC input.

Any opinions on the ST head with the cleanest waveform?  I've been considering just spending the money on a 5KW winco generator head.

Yep, 2KW of electric load is about 66% load on a 6/1 IMO.  As I approach about 3.4KW of load on my 6/1-ST5 I get dark smoke and start to see excessive RPM/Frequency droop indicating to me that I have started to exceed my engines HP limit.  I can get 3KW indefinitely(and nearly 18,000 BTU/HR in hot water from the engine:))... 

As for generator quality, sadly all ST-5's are not created equal, so unless it is sold by someone who tests and has standards, you get what you get...  I have only got to o-scope my own, but it's output compared pretty favorably in terms of wave shape, to the domestic generators I work with at our remote sites(onans and generacs).  it's design(harmonically regulated) may give you starting voltages that are outside some precision equipments ability to accept.  It's voltage output is also RPM dependent.  Because of this IMO, probably the biggest issue with an ST-5 is the choice of driving it with a 6/1, and the resultant single cylinder flicker(voltage flux due to rotor speed changes over the engines power cycle)... IF you must have a little more precise power, it is probably best to go with a brushless AVR type head.  The AVR will also help to dampen the voltage swings seen with the single cylinder RPM flux.