project X alternator progress report

[mobile_bob]

building on what i have learned in working with alternators such as the 110-555jho presolite, and then
the 4400 series leece neville in the remote transformer rectifier projects, i am to the point of the final
unit being produced

the goal is 150amps at 57.6volts which equates to 8640watts, and the expected efficiency should peak
at just over 80% which will require nearly 11kwatts mechanical to drive it.

the 2200watts of loss will manifest itself in the stator as heating, and unless i do some additional cooling techniques
the stability of the alternator will be a huge question.

so i have decided to fit the stator with a water jacket to make for a water cooled stator, the goal is to remove approx
50% of the heat via the water jacket transfer, this amount of heat is the equivalent of 5 gallons of domestic hot water per hour
at full output.

now the manufacture of a water jacket to cool a stator is not a task to be taken lightly and has been something i have been toiling over
for several weeks, finally today i think i have a design worked out that will accomplish the goal and harvest the heat available from the stator.

admittedly this is not KISS, and anyone that knows me should understand that in my opinion KISS is overrated 

the unit will require a water pump, which really is not much more than a tiny fountain pump, so the tradeoff for a few watts of power
to drive a pump will allow the alternator to make some serious power in a thermally stable manner and allow me to make use of the
excess heat rather than simply trying to force more air into the unit which would eat up more power than the water pump anyway.

hopefully by this weekend i will be ready for the first in a series of tests.

more to come

bob g

[mike90045]

Why all the copper loss (high amps)?  Why not try a monster PMA?

Maybe some of the windings could be copper tube - water flow in the middle?  Trying to move 2KW of heat out of a core is going to be tough, how to cool the rotor?  Hollow shaft with air vents?

I guess you've got the answers?

[mobile_bob]

Mike:

surprisingly the field current is about 50-60 watts, the balance is the stator losses at high power
so pm field really isn't going to alter the heat load by much, and what little it does is a good tradeoff
for ease of regulation control.

as for the stator heating, i am projecting based on what i am learning from ealier works, the reality may
be quite different.

when i shoot for 150amps at 57.6 vdc we expect approx 8600watts

if i figure on 80% efficiency, which i think is a fair number, therefore

8600 / .80 = 10750watt input

i know from earlier tests my belt drive consumes approx 200watts, so we are then at about 2kwatts of heating
a good portion of which is in the stator, and a significant amount in the rectifier bridge as well, probably about 300 watts
at full load

that leaves me with about 1700 watts give or take to dissipate via the stator

this assumes an efficiency of 80%, if on the other hand the efficiency is higher to say 85%
the numbers obvious get much lower.

i won't know for sure until i try the watercooled stator.

we have to remember the goal is mainly to stabilize the alternator temperature at full load current, and to attempt to recover
some of the heat from the alternator and put it to a good use, rather than allowing it to be wasted and have to provide more
airflow via an additional fan.

another point of interest, this based on earlier tests
the remote transformer/rectifier based alternator starts out being able to produce very near 100amps at full field and the speed
at which i am driving it, as it gets hot the amperage drops back to a thermally stable 85amps,
that drop of 15amps is a loss of useful electrical production that is going directly to heat, if i add additional blower cooling i can recover
most of the lost 15amps.

so we have a couple things to ponder, reduction of heat increases the ampacity of the alternator, and should increase longevity of the alternator
as well. BSFC is also significantly lower.

now moving on, and assuming my projected numbers are close

what do we do about all the heating in the stator core?  not much we can do in my opinion, because
the unit is relatively small at about 6.5" in diameter and maybe 12" long, with a weight of approx 31lbs
there just isn't much room to get creative with hollow conductors as used in the giant alternators that the utility companies
use.

it is also significantly less expensive to work with off the shelf parts where possible and not send out for custom wound stators.

in the end, my decision to look toward a water cooled stator was about the only avenue left that looked to be plausible, if i could figure
a method of manfucturing the water jacket reasonably.

i think it will be successful because the stator cores of these alternators have very little steel in them, they are not very thick, so
having an outer surface that is cooled with ~60 degree water should draw the heat from the interior of the stator quite quickly
and effectively.  when you are dealing with a stator temperature of ~250 degree's F that is less than .375" thick (from back of slots
to outer dimension) its not hard to imagine a large percentage of heat moving quite freely from the core to the coolant.

as the water in the recovery tank warms the transfer rate would reduce, but the need for max heat transfer would also reduce
because the batteries need for amperage decreases as they become recharged. i am not sure but i bet that sizing of the coolant tank
can be made to get a relatively linear relationship in operation, or in other words keep the alternator at a fairly constant temperature
from full load back down to part load, and finally low loading.

for me it is something that needs to be tried, it is really nothing new as bmw, mercedes and others (maybe some new GM or Ford too)
have made the move to water cooled stators on their alternators, to allow for higher outputs needed for todays automotive loads.

i won't be doing anything that they are not already doing, other than feeding the stator with colder water, and if there is a significant amount
of hot water, use it to wash dishes with?

it looks to me like if it can be done, and proves to be beneficial and useful, and is cost effective, then it should be at the very least explored.

imho

bob g

ps. no i don't have all the answers, but i am actively looking for all the answers i can find.

[mike90045]

Cool!  Thanks for the detailed explanation. Copper is one of the better heat conductors, so if you can increase the layer to layer heat transfer, that might be enough.   I wonder if the square wire is better at that, vs the round wire ?

[BruceM]

Wow, that's a monster charging machine!
156 charge amps is a lot of juice to a 48V bank.  What's the system that this baby will fit into? 

The other thought that comes to mind for this much 48V charging capacity is why not a correspondingly large 3 phase ST head reconfigured to 60V, with custom regulator (best) or a pair of 80amp PWM charge controllers (no custom electronics) ?

[mobile_bob]

Bruce:

the unit goes into a 48volt system designed to take a fast bulk charge in a limited amount of time per day, and have excess capacity to
provide for inverter loads at the same time.

i do have an stc12 3phase head that will called to duty should this project x alternator fall short of my spec's, that was my interest in seeing
a modified st head for 60volt output.

bob g

[BruceM]

Thanks, Bob.  That's a neat project!  I recall your previous post on running the generator briefly in the AM and PM to meet peak loads- now I see how this fits into your power management scheme. (Very smart.)  Is the battery bank large capacity (to take 150A charge)  or are you planning to use AGMs to be able to take that fast charge?  That is the neat part of AGMs, plus the extra 10% of charge efficiency.

I wish the AGM battery prices would come down.  At one time I had found a source in AZ for $100/ 100AH AGMs- but those prices are now doubled.  They were "sorta" deep cycle batteries from Gruber who does big business servicing big UPS setups for businesses in Phoenix, AZ.

[mobile_bob]

Bruce:

i have been leaning toward the use of AGM batteries lately, mainly because of the increase in charge efficiency and their ability to accept
a large amount of amps in a short period of time.

in my analysis, the added cost of the batteries is offset by the added efficiency and the reduction in fuel consumption during recharge to the
point that they appear to be the clear winner.

still crunching numbers though, and working out how best to get the power levels i am looking for.

the 195 reachs peak efficiency as measure in BSFC at 100% loading, my attempt to load the engine continuously at or near this level
dovetails with the production of domestic hot water nicely, and as the charge rate tapers back i can switch in compressor loads to provide for
refrigeration needs to maintain the loading on the engine.

my thinking is the system will reach peak efficiency about 6-7 months out of the year, where electrical power and the waste heat harvested are in good balance with each other for this class of machine, this leaves the summer months where a smaller unit might be a better match, and the very cold winter months where a larger unit might be a better match as well.

trying to get a well matched set of components all working together to load the engine at my preferred design limit is the goal, lofty as it might be.

for instance, my exhaust heat exchanger is sized to produce its best efficiency with the engine running at above 80% load, and it works best
in self cleaning when the run times are limited to shorter duration with enough time between to allow for full cool down.

the goal has always been 1 hour run time before waking, and 1 hour before the suppertime rush, give or take perhaps a quarter hour or so.

when its all assembled the last component is the control system, in my opinion it will be as important as any single component that makes up the
hardware side of the system, without adequate control i don't see me being able to optimize the system with any consistency.
working on the hardware has taken the lead focus because i feel like it is going to be much easier to build and implement the control system after
i know exactly what needs to be controlled.

bob g

[mike90045]

Just adding a wrench to the pile of batteries -  there is a thread here
http://www.wind-sun.com/ForumVB/showthread.php?t=7344
about   "Lithium Iron Phosphate Batteries; LFP; LiFePO4 discussions"  batteries and how to "drop them in" as replacements for a PV based system.  EV battery prices seem to have fallen to where their deeper discharge and 10x cycle count may make them competitive with AGM batteries.

[BruceM]

Great link, Mike, thanks.  I didn't realize the EV folks have something new and available that might work for off grid folks.  The LiFePO4 batteries do look very interesting.  The number of cells (about 40) for my 120V bank is a bit daunting.  I'll be watching these and their prices closely!

I note that the LiFeP04 batteries require individual shunt regulators to avoid cell overcharge.  

Bob- Since you haven't committed your controller and charge regulation hardware yet, you might want to look at the LiFePO4 batteries as an alternative to AGM.  They do look to me like they have AGM's beat.  Extremely high charge efficiency and decent charge rate, greater cycle depth and life.  

[Cornelius]

Amanda over at Fieldlines/Otherpower also have some experience with LiFePo4:

http://www.fieldlines.com/story/2009/8/26/71752/6793

If one have the money, i'd say that LiFePo4 are the way to go...

[BruceM]

Great follow up link about two slightly different individual battery shunt regulation/charge control schemes, thanks Cornelius.

I use an analog shunt regulator scheme per 12V battery on my setup, to avoid introducing EMI.  I would stick with that again for LiFePO4, but simpler, since no temperature compensation is needed. Building and wiring 40 regulators for my 120V bank wouldn't be fun, but 16 for a 48V system is reasonable.   Looks like a charger with just 2 output levels would suffice, too. (Full power, low amp).

I hope the prices on LiFeP04 will come down soon as more EV's start to hit the streets.

[mobile_bob]

those batteries look quite interesting, but
the cost looks to be quite high, unless i am  wrong?

and i would like to see a mainstream manufacture and/or supplier in this hemisphere
because i am not at all sure i want to be a guinea pig for several thousand dollars worth of battery.

perhaps over the next year or so the battery will become more mainstream, and common around here.

bob g

[Cornelius]

The initial cost is high, yes.
A supplier near me sells 3,3V, 150Ah cells for around (US)$340 (Mainly as replacement batteries for Electric Cars.), but consider the discharge rate down to at least 80% and 1000x charge cycles you'll have a battery that doesn't cost much more than good AGM's over time.

I haven't done any precise calculation, but lead acid batteries shouldn't be discharged much more than 40-50%, and then a good one lasts maybe 300x charge cycles, so one have to buy a larger LA battery and replace them at least 3x over time, compared to LiFePo4 batteries... (Now, there are exceptions, like the S5000 series LA from Rolls/Surette, but... There's another point; weight and maintenance...)

I've just ordered a new 36V, 16Ah LiFePo4 battery for my electric bike (36V, 750W motor); my brother got one, and he gets a 45km ride with that one. Currently i have 3x 12V, 17Ah SLA, and i get 15-20km out of them. My SLAs weighs 21kg, and his battery weighs 5,5kg and are under half the size...
(On a side note; my lawnmower uses a 36V, 2,7Ah LiIon battery... )

So it's starts to 'stream' here in Norway at least, and i think we'll see more of them, at lower prices soon. (I hope...)

Edit:
Another point are charge time: most LiFePo4 batteries has a standard charge current of 0.5C, and have no trouble accepting a 1C charge rate without shortening the life.

[BruceM]

Bob, I feel the same way  about expensive new batteries from China, essentially single source.  It looks like LiFePO4 will be the EV battery tech of the near future, and we should see more sources and  price improvements.  My thanks to Mike and Cornelius!