How weak cells combined with deep cycling destroys the Vectrix VX1 Battery

A “Recharge after BALPOR” is the charging event following a Battery Low Point Reset (BALPOR). I made up these terms due to the complete lack of explanation in the official manual. I feel there is a need to give these worrisome and recurrent events some sort of name which makes them more recognizable.
You can search VisforVoltage for numerous previous discussion of these terms and events.

The “BALPOR” consists of the sudden disappearance of the remaining bars on the right hand side battery charge display (the “hourglass” shaped 17 segment LCD), and the remaining range display also drops to zero.
For example, 5 bars (along with about 20 fictitious remaining kilometers of range) will disappear in one hit, about 10 seconds after the battery telltale started to be lit.
If the throttle remains pulled open when the battery telltale is lit, then the current draw continues and the voltage drops below the acceptable level, which is programmable and different in various firmware versions. The BALPOR is caused by the faulty assumption of the stock electronics that there is still charge left in the battery, when in fact some cells (or all cells) are empty. This causes a rapid voltage drop during the reverse charging of the already empty cells, which finally convinces the stock BMS that the battery is empty. The signal to reset the gauge and the range indicator to zero is sent, this is also known as the disappearing bars syndrome, and the stock BMS now believes that it has an empty battery which is ready to receive a full charge.

This sets the stage for the “Recharge after BALPOR”, which is one of the more damaging events in the life of a VX-1 battery, but not necessarily the worst.

The “Recharge after BALPOR” behavior is due to one of the three VX-1 charger safety cutoff mechanisms in action, namely the high voltage safety cutoff. Another descriptive term for it would be "C/3 overcharging with 15min cooling breaks".
(The other safety cutoffs are capacity cutoff and high temperature safety cutoff. These safety cutoffs terminate or delay charging if either the battery temperature is too high, or the voltage is too high, or the battery is believed to be already full.)

The charger behavior during a Recharge after BALPOR will look very similar each time, but there are two distinct scenarios that may be occurring on the inside:

Scenario 1: The BALPOR occurs very close to a recharging point and driving stops immediately after the BALPOR. Only a little bit of reverse charging is inflicted on the low cells, and the high SOC cells keep the remaining charge. But the BMS assumes that they are all equally empty.
The charger then pumps about 11A through the battery in CP mode, trying to put 30Ah into the high SOC cells which still had about 8Ah left in them at the beginning of the charging process. The safety voltage cutoff has to stop a disaster when the full cells have reached such a high voltage that the total string voltage reaches 153V (or 148-153V, depending on temperature).
153/102 = 1.50V/cell on average.
If the cumulative battery voltage reaches the safety cutoff voltage (shown in the middle instrument cluster display) before 16/17th are reached on the hourglass display, then the charger eases off for 15 minutes, and then repeats the 11A cell abuse again, up to another three times, until either 17/17th are reached or the cutoff voltage level has been reached 4 times.
By then the overcharged, good cells are beginning to get quite hot, but they will still have to put up with a C/10 charge lasting 1 hr following this, called the CC charge stage.
Once that is over, the bad cells will probably have caught up and all are full and hot. The stock BMS does the EC charge stage for 1 hr.
Depending on how large the discrepancy between real and perceived SOC of the battery cells is, the hourglass charge display may show less than full at the end of all this. This happens for example when the whole battery is actually full, but the electronics believe the battery is empty, like I have seen after fuse repairs. If the state of charge (and along with it, the voltage) of all cells is very similar, then only little damage occurs, because all cells will get 4 very short C/3 charges until they are all at 1.5V. Even cells with permanently reduced capacity will not get damaged much by this, but only if they have the same State of Charge (SOC) as the good cells when it occurs.

But even in this first scenario they will not usually reach full SOC at the same time. The good cells have remaining charge, maybe about 8Ah or so, when the BALPOR occurs, and the low capacity cells are empty and hot from the reverse charging event prior to the BALPOR. These “Permanently reduced Capacity Cells” (PREDUCCs) are initially unable to efficiently absorb charge due to their deeply discharged state and their high temperature. Meanwhile, the good cells are cool and soak up the charge current well without heating up. This further increases the SOC discrepancy between the cells as the “Recharge after BALPOR” progresses.
If there are, say, 13 PREDUCCS in the 102 cell string, and their voltage might be 1.4V when the average of 1.5V is reached, then you can calculate the following:
102-13=89 ; 13 * 1.4V + 89 * XV = 153V ; (153-18.2)/89 = 1.5146V per good cell instead of 1.5V. That does some damage to the still good cells, causing premature aging.
So in summary: The first scenario causes moderate overcharging of the many good cells.

The second Scenario is much worse, and much more likely:

The BALPOR happens, unexpectedly as it usually does, whilst still a few km from the intended destination with the recharging point. Getting there in crawling mode uses up the remaining charge in the PREDUCCs at low current rates, until they get reverse charged even at the reduced current which the BMS still allows. They heat up more than in scenario 1. The good cells find nothing wrong with it all, they can easily supply the low current requested by the crawling scooter.
The severe problem will occur during the following recharge because the stock BMS still plans to perform a full 30Ah charge as soon as it gets plugged in. It again incorrectly assumes that the whole battery is empty and balanced.
If the amount of "empty but rechargeable capacity" of the good cells exceeds the total rechargeable capacity of the PREDUCCs, then this results in a much worse SOC imbalance compared to the first scenario, because the fewer weak cells must produce a much higher voltage rise each, before the BMS notices it as the overall voltage reaching the safety cutoff value.
The severity depends on how many Ah were drawn from the battery after the BALPOR. The more, the worse!
As I said, this is unfortunately the more common scenario, and it happens whenever a VX1 that already has PREDUCCs is regularly ridden until empty. It results in total destruction of the PREDUCCs over a relatively short time period (maybe a few dozen to a hundred charges).

Here is how the second scenario destroys the already weakened cells:
Assuming the same number of PREDUCCs in the 102 cell battery as in scenario one, i.e. 13, the voltages now add up like this:
Safety cutoff 153V is the voltage at which the stock BMS will stop CP charging.
Voltage of the 89 good cells when the PREDUCCs are already full is again presumably about 1.4V per cell.
==>153V = 13 * XV + 89 * 1.4V
==> 153V - 124.6V = 13 * XV
==> 28.4/13 = 2.18V per PREDUCC!!! And the BMS believes they are all at 1.5V!!!

That really gets them cooking! These already damaged cells will be heating up rapidly and possibly start to vent oxygen and hydrogen through the safety release valve on top of the cell, further reducing the cells capacity.

The effectiveness of the temperature safety cutoff mechanism during the second scenario depends on sheer luck: If the few PREDUCCs among the 102 cells happen to be located away from the temp sensors, then their rapid heating will go unnoticed. If the worst PREDUCC happens to have a temp sensor on it, then the worst over-charging will be prevented due to the detection of the steep temperature gradient occurring in this cell.