Micro CoGen.

Thanks for the info and link Marcus. I think you may be right, but I don't have any information on the engine protection relay so can't be sure. I'll try starting it with the oil pressure switch disconnected (ie. the relay thinks there is oil pressure) and see if it starts.

Cheers,
Nick

Hey,

I've been working on a project for my brother for a while now. He's got a remote house with no utility connections. I've set him up with a Trace 4548 inverter / charger, a 180Ah 48V ex substation battery bank (hoping to get more soon), and have just recently acquired a  Kubota diesel genset (GL6500S rated at 6kW continuous) to replace the unreliable and fuel hungry Hyundai petrol gen they're currently using.

Photo of the same model (but better condition) genset


The Trace can charge at a maximum of 60A (DC), so around 3.5kW AC input.

Speed Control
The alternator is 2 pole (3000RPM over here in Aust), and is belt driven with a ZB600-GL-S, also running at 3000RPM. All speed control seems to be electronic with feedback from the alternator.

Is adjusting the run speed as simple as changing the pulleys and belts given the speed control seems to be all electronic?

Any opinions on how this engine would run if slowed down to say 2400RPM or so. I know I'd have a loss of power of around 20% (possibly more/less depending on the torque curve of the engine), but it'd run quieter and I would expect be a bit more efficient on fuel. I can limit the input to the trace to 4kW which would still not be running at 100% load. The batteries won't accept full charge for very long anyway and then the load will drop off unless stuff in the house is drawing. Given the gen will be running while the house is unoccupied this isn't likely.

Cogen / heat exchange
The genset will also be set up as cogen. My plan is to circulate the coolant through a plate heat exchanger (like the below). A secondary loop will be pumped through the heat exchanger and back to the hot water tank, drawing from the bottom and returning to the top of the tank.

http://www.ebay.com.au/itm/270833027611?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1423.l2649#ht_7313wt_1002

Will the 3/4" connectors restrict the flow too much for the coolant circulation? Would this be worse than the pressure drop through a radiator? I'm not expecting any problems but would be very happy to defer to someone with some experience!

I'm also planning on setting up a heat exchanger on the exhaust. Details haven't been fully firmed on how, but probably a pipe in a pipe kind of system with some sort of ridge to break up the flow, and easily removable for cleaning. Has anyone had a go at an exhaust heat exchanger before?

Safety
Finally, has anyone got any recommendations on engine safety? The gen has cutout on coolant temp and oil pressure (both tested and working fine).
Is there a way of adding a sensor so that the gen will shutoff on a broken belt? At the moment the engine would keep running, and then the Trace would keep trying to restart it. Not a good option! I'm thinking a couple of time delay relays that won't allow the gen to be restarted for 10 mins after being started initially (crank inhibit: 10s on delay, 10 min off delay).
Any other bright ideas - perhaps a SPL sensor?

Mods, if this should be in a different section please move it.

Hey Mr Fixit,
Out of interest do you get much usable heat out of the SC Yanmar?

Looking at the RC circuit you are correct that this is a LPF. However looking at the capacitor alone (with no load) this is a HPF. Regardless, we both mean the same thing.

Getterdone,
The capacitor could be put anywhere on the circuit, and should be put where convenient. This will probably be at the switchboard where it can be fused / protected with a circuit breaker. It will be (very) slightly more effective at the dryer, but not enough to justify a difficult installation.

Without seeing waveforms / knowing THD and knowing the requirements of the dryer the cap size can't really be calculated. Cap's are cheap, so you can probably use the trial and error approach.

The reactive power supplied by the capacitor will be Q = V^2*(2*pi*f*C) where V = volts, f = frequency and C = capacitance. For example for a 10 uF cap on a 110V 60Hz system, Q = 110^2*2*pi*10*10^-6 = 45VARs. Using this as a baseline a 30uF cap will supply 135VARs.

Did you try using a bench grinder or something and see if that would get it started? An unloaded motor would be better than a loaded one (ie air con)

Clarksonguam,
I would strongly recommend against adding capacitors to the system. Generator voltage regulators are not designed to drive a capacitive load and voltage fluctuations can result with resultant damage to equipment (fluctuations will be greatest where there is no resistive load).

getterdone,
Firstly, i would recommend trying the flywheel method as discussed above.
If the problem you are having are to do with harmonics (due to PWM) a capacitor would have a positive effect. The capacitor acts as a high-pass filter, smoothing out the waveform. It is similar to how a capacitor works in a DC system, providing stored energy at the point of the wave where it is required. As for the level of harmonics, and the amount of filtering required, I could only guess.

Check to make sure the xantrex can handle leading (capacitive) loads. I expect it should be able to.

High harmonic currents may cause capacitor heating and premature failure. Make sure you fuse them, and perhaps buy a higher voltage capacitor than you need.

Change of plan. I didn't buy the hawkers because I got in touch with a contact of a contact. I now have access to pretty much all the batteries I want for free 
I'm Picking up 8 x Exide marathon M12V90's on Monday, about 5 years old out of a substation DC supply. VRLA / AGM batteries.

Datasheet is found at http://www.networkpower.exide.com/exidepdfs/Section%2022.60%202010-12.pdf

Is there any requirement for a multi-stage charge for these batteries? The literature seems to point to a single charge rate (and bloody fast one at that).

The Recharge Characterization says that they should be charged at approx 2.27V/cell = 13.62V for the battery. It gives an acceptable range of 2.25-2.3V/cell. It's hard to make out the charge current draw in the lower regions of the graph, but it's pretty high. My charger is good for 60 amps (Trace 4548E) which will be C/3 for 2 parallel strings. This won't quite give the quickest recharge, but will give sufficient to get them going. The trace is also temp compensated.

I'm expecting the daily discharge won't be more than 20 or 30% max.

I am proposing a bulk charge limited to the max setting of the trace (60A), an absorption voltage limit of 54.6V, and absorption time at a couple of hours. If they're not getting enough juice I'll extend the absorption charge time, and may drop the charge voltage down to 54.5V or 54.4V. It'll take a bit of trial and error to see how the batteries respond.

I haven't bought them yet. Was going to buy 8, so 2 48V strings. Price was AU$1200 for the lot, which seemed good value.

However I've been able to get in touch with some people who work for our local rail network. It looks like I can have my pick of the used substation batteries at no cost. These are changed out every 5 years on a regular replacement program, so should have at least another 5 years of life left in them.

This is looking like a more attractive option 

That sounds promising 
What bulk current and absorption voltage do you use on them? The spec sheet only gives a float voltage (as that's what they're designed for).
Cheers

Quote from: BioHazard on July 18, 2011, 01:15:18 AM
What type of batteries are they? Some standby type batteries will be "lead calcium" batteries which don't like to be cycled. But not all telco batteries are this way, some are just standard AGM batteries or other types.

Original post edited with battery type. Cheers

I can get some ex telco batteries at a good price which have been on float for 3-5 years. Batteries are "Hawker supersafe 12T155F" Valve regulated lead acid type.

I know the batteries are designed to maintain a float charge rather than regularly cycle. Has anyone had experience using stationary / float / telco (call them what you will) batteries for a cyclic application?

Hi all,

First time post. Thanks for the great information provided in these forums.

I'm setting up an off-grid power system for my brother in Australia (Victoria). As he is completely off grid and LPG is expensive I am considering using a water-cooled diesel generator as both heat and electricity source. They are surrounded by trees so solar isn't an option. At this stage I've purchased a Trace SW4548E (4.5kW, 48V) inverter/charger, and will shortly be purchasing a battery bank (around 300Ah). This part of the installation is pretty much set in stone. Power requirements are pretty small at approx 5-6kWh per day (accounting for efficiencies in the charge / discharge cycle).

Hot water heat input requirement is around 10kWh per day (120L raised 60degK + allowance for losses in the heat exchanger & tank).

The trace will be capable of pulling 3.5kW during bulk charging. Given alternator & drive efficiencies and auxiliary losses in the generator (assumed loss of 30%) this requires a 5kWm input.

I can obtain a Yanmar 2TNV70-PGA for a good price brand new. This is rated continuous at 9.76kWm @ 3600RPM / 8.16kWm @ 3000RPM. I could belt this engine down where it would run efficiently (ie. 2000RPM, continuous output of 5.4kWm). I could also belt it to the peak of it's torque curve at around 2800RPM (continuous output of 7.7kWm).

Questions
1. Has anyone got experience with similar size engines (~13HP) running at lower load? Will it glaze / gum up at 50% maximum load?
2. Will the heat output from the cooling water drop off too much to be usable?

Any pointers or information would be greatly appreciated.

Another option is to add a 1.2kW water heating element, pushing the engine loading up to about 6.5kWm. This would also push 2.4kWh into the hot water system over the gen's 2 hour run time.

Thanks for looking