mobile-bob -
  My largest bat set is 1000 AH, I try to go conservative since many manufacturers suggest 18 to 30 amp recharge rates for each 100AH at 8 hours. I know this suggests that 300 would be ok, but monitoring the temp rise suggests otherwise. I use a Betatherm 10K3A thermistor for the front end, through a Burr Brown 1NA118 precision instrumentation amplifier to a PIC 16C74B processor. I limited myself to 8 sensor inputs cause I'm lazy. The PIC will input 8 directly, so that became my choice.
   1. High on the back wall of the battery room.
   2. Low on the front wall of the battery room.
   3. High over the battery rack.
   4. Just above the equipment rack.
   5. Battery case 4 in from the start of the series.
   6. Battery case 4 from the end of the series.
   7. Center battery case.
   8. And one just outside the door for outside measurement.
  By comparing these inputs I can keep a close tab on what is heating up and how fast. The center battery case is also used to make voltage adjustments for charge rate, and the average case rise every ten minutes controls the current input. I use 4 360 watt Peltier Effect Devices for both dump load and to either heat or cool the battery room as needed. They respond very well to PWM control so I can fine tune where the power in from the wind generator goes, either to the batteries or dump load or split operation.Â
  A second set of sensors monitors the voltage on every 6th cell. I once thought of monitoring every cell, but the processing time for the PIC would get a little high, and this method seems to work fine.
  An old Compaq Aero laptop runs the system, does the logging and sends alarms. If any upper or lower limits are reached in temperature or voltage or current I use contactors to open the inputs to the batteries and the inverters and clamp the generator to full dump load.
  Three good sources for more information would be:
  Â
   R. Jones, Recharging VRLA Batteries for Maximum Life, Proceedings of Intelec (99) Copenhagen
   S. Despande, Intelligent Monitoring Systems for Continuous Monitoring on VRLA Batteries, (99) Copenhagen
   P. C. Butler, Utility Battery Storage Systems, Sandia Report, SAND92-2272 (1993)
 Another good read is the Professional Papers put out by Absolyte, although I realize that they might have a bias on what they print. However, after about 8 years of monitoring, my numbers match theirs very closely, and I would not hesitate to base a charging system on what they recommend.
  I understand that all this may be just a wee tad overkill for the job at hand, but remember there is no one more anal than a retired engineer with time on his hands, and it does keep me out of the bars.