[S Keith]

U.S. Gen1 modules are different than Gen2 and 3. There was a flaw in the terminal design that allowed KOH to leak out the terminal to case interface at a pressure lower than the blow-off pressure of the vent. Toyota developed a "sealing" procedures that did nothing but make it all gooey as heck, IMHO. The Gen1 failures are most associated with this loss of electrolyte causing reduced capacity.

They improved the terminal design and internal resistance in Gen2 significantly by providing an additional connection between cells on the bottom of the module in addition to the top. Gen3 internal resistance is a touch better than Gen2.

Prius uses 40-80% SoC with an aggressive effort to maintain 60% thus minimizing cycles. IMA originally used 20-80% SoC with a tendency to keep the SoC as high as possible. The new scheme is unknown, but it is expected to use less than 20-80% SoC range for improved life.

The same manufacturer made both Prius and IMA battery packs. The cells are very different. The IMA are literally "D" cells. The Prius are prismatic rectangular modules that have a common "vent" space. All NiMH produce Hydrogen when charged. When the cells settle the Hydrogen is reabsorbed. Within an IMA pack this means that each cell behaves independently. In the Prius module, the six cells vent to the same space, so they all reabsorb according to their potential. This improves uniformity across cells, particularly when an actual high-pressure blow-off occurs. All six cells are affected the same. In an IMA module, only the individual cell(s) is(are) affected. In the prius, this creates a diminished capacity, but it's fairly uniform across all 6 cells. In the IMA, the capacity hits only the affected cells thus creating an imbalance between cells. The reduction of capacity of the affected cell(s) has the same effect as ALL cells having reduced capacity. Of the hundreds of IMA sticks I've had in my hands, almost all module failures are due to 1-2 cells in the group performing very poorly while the other 4-5 perform very well.

With the '06-11, the modules are now 12 cell, so the failure rate essentially doubles. the design was changed in '09-11 with a greater spacing between the sticks for better cooling. This should have improved life over the 06-08 packs, but they were about 3X as bad. Personally, I attribute this to a quality issue at Primearth (Toyota) as they didn't give a squat about "D" cells anymore.

When cell imbalance worsens, individual cells are run outside the 20-80% range... literally 0% for the weakest with polarity reversals and 100% for the strongest cells and a greater risk of venting.

IMHO, a grid charger as a preventative maintenance tool conducted every 3-6 months is KEY to maximizing life of an IMA pack. Regular grid charging and discharging will help keep the state of imbalance minimized, thus preventing individual cells from being operated outside the 20-80% SoC range and avoiding the accelerated deterioration.

Steve