MikeBoxwell said:
Actually, it is true of lithium as well as lead acid, ni-cad or nickel metal hydride.
How many people buy a laptop computer and only rarely take it off mains power? After a year or so, the batteries are useless with only a few minutes life rather than a few hours. The answer is simple - unplug the computer at least once a week and run it on batteries until the batteries are running low. The result will be a laptop battery that will last for 5-10 years rather than giving up the ghost after 1-2 years.
Mike, The laptop analogy doesn't apply here for a couple of reasons. The first is that the lithium cobalt cells most commonly used in laptops do not like to be kept fully charged 24/7 - and a laptop with the battery in and the mains plug in the wall is always on charge and is always kept topped off. This has long been known to drastically shorten battery life. Another factor is heat - more time spent next to a hot processor is a bad thing. To maintain the battery in a laptop, discharge (or charge) to about 50% capacity, put in a plastic bag, and toss it in the fridge. Unplugging will help reset the 'fuel gauge' but will not reverse battery damage.
MikeBoxwell said:
It's true of every electric car I've ever driven as well (and believe me, I've driven a few) and bears up with the tests I've carried out with my Mitsubishi iMiEV. If I charge it up every night having done 10-20 miles, the computed range of the car starts to deteriorate over a period of 10 days or so until it is reporting a total range of around 60 miles. If I drive it so that I'm using most of the range once every week or ten days, the range stays up to over 80 miles. When I drove the car from Coventry to Oxford and back (120 miles with a quick top-up charge in Oxford) the car subsequently reported a range of over 90 miles the following day.
You've absolutely nailed the 'computer drift' that I spoke of earlier. An overall system will track battery capacity best if the battery is completely drained before a recharge. One can get a pretty close approximation of state of charge from lead-acid batteries from cell voltage, but the system cannot read individual cell voltages, so it estimates based on battery voltage - but that introduces the first disconnect. When a battery is healthy and all the cells are tracking normally, the 'fuel gauge' is pretty accurate. But when one cell is impaired and the others are a bit overcharged to compensate, the computer is 'happy' to give a 100% full indication but the battery is silently cursing us from the back seat.
One cannot get an accurate indication of state of charge from most lithium cells
by looking at cell voltage - we need to count current into and out of the cell and that introduces another 'level of separation' between the actual and indicated state of charge and thus range.
There are plenty of folks on this forum with EV experience that goes back to the 1994 thru 2003 production EVs used in California. Hundreds of these vehicles are still on the road today scattered around the country. While one could certainly assume the U.S. to be an EV wasteland if they listen to the press, the 'underground' has fairly significant roots and corporate knowledge.
Cheers!
Andy
[edit to correct lithium cell current measuring text]