EDIT1: New answer to modified question (original see below)
First of all I need to tell you that this set-up is very risky. Very very very risky. This way of connecting the batteries means there's a much higher risk of damage over multiple charge/discharge cycles.
To your specific problem, if one battery in one chain becomes short circuited, the other three are "requested" to charge up to a 4cell voltage by whatever charges them, seriously over stressing them.
Put simply, if any battery becomes shorted in use, you can bet on it the other ones in the same chain are completely wasted as well, not to mention the severely increased risk of one cell blowing up or becoming shorted in the first place.
This is why battery packs are always built as in my original answer below by the factories that make them. Now, before you go and re-build a pack like that, let me warn you to always balance the cells to each other before you put them parallel, with a set up like this:
simulate this circuit – Schematic created using CircuitLab
Because cells are never exactly charged the same, ever, so you let them charge each other through a resistor, making sure no idiotic current flow. How long you leave them like this before connecting them depends on whether they have been fully charged with the same charger. If they have, 10 to 15 minutes should be more than enough. If they have not, leave them for at least an hour, but probably just half a day or more. You can reduce the resistor to increase the current and reduce the time, but be sure the resistor can handle the power, usually 10 Ohm is safe enough with a 1/4W resistor, lower values you may have to go to 1/2W, 1W or even 3W. If you want to make sure they stay alive as long as possible, don't go over 1/3 their capacity for the current through the resistor.
Also, if you want to prevent all damage as good as possible, look into balance-chargers and how they work, as they try to keep all cells in your pack healthy during the charging, severely reducing the risks of damage over time.
ORIGINAL ANSWER:
If your setup is like this:
simulate this circuit
And one battery, say BAT6, becomes a short circuit you will get:
simulate this circuit
So BAT5 is then shorted between it's + and - terminals. Often when one cell of a parallel set-up goes bad the other one is very damaged as well, because the unbalance that causes premature cell-death affects both the cells, just the slightly weaker one will go first. BUT even if that wasn't the case, shorting BAT5 will cause it to get severely damaged soon enough.
So they only thing you should do to fix this is remove both BAT6 and BAT5 and create a lower voltage pack like this:
simulate this circuit
Or get a pack with half the capacity and two cells left over, like this:
simulate this circuit
Not all rechargeable batteries have a nominal voltage of 1.2V; it is specific to a few chemistries that happen to be popular. According to Wikipedia, the following rechargeable battery chemistries have cell voltages of 1.2V:
- Nickel-iron
- Nickel-cadmium
- Nickel-metal hydride
At a glance, it would appear that nickel is the common denominator, but this is not the case, as nickel-hydrogen and nickel-zinc have voltages of 1.5V and 1.7V, respectively.
So, excerpting the relevant sections of Wikipedia:
Nickel-iron:
The open-circuit voltage is 1.4 volts, dropping to 1.2 volts during discharge.
Nickel-cadmium:
Ni–Cd cells have a nominal cell potential of 1.2 volts (V). ...a Ni–Cd cell's terminal voltage only changes a little as it discharges.
Nickel-metal hydride:
A fully charged cell supplies an average 1.25 V/cell during discharge, declining to about 1.0–1.1 V/cell...the starting voltage of a freshly charged AA NiMH cell in good condition is about 1.4 volts.
In conclusion, the 1.2V is only nominal voltage. 1.2V was a round number close enough to the typical voltage of those three chemistries and thus became the voltage on the label. However, every battery chemistry has different characteristics including voltage across a discharge cycle and open-circuit voltage.
Best Answer
There are two main types of rechargeable batteries - Lithium Ion and its children (such as LiPo) and the rest.
LiIon "Vent with flame" ... - spontaneous "uncaused" laptop self dismantlement because (even) Sony could not get it right.
The no-name brand can't always either ;-) ...
LiIon (Lithium Ion) is a "slow release bomb" waiting to happen. A LiIon battery may be able to be persuaded to "vent with flame" (Gargoyle knows) by charging too fast, charging to excessive voltage, heavy discharge, spike penetration or heavy knock, charging at normal rate when voltage is low, charging at all when voltage is very low.
Protection devices for LiIon calls are the norm. These usually mount INSIDE the battery casing so their presence is not obvious. LiIon cells should ALWAYS use such devices. Most manufacturers will not sell LiIon cells without internal protection devices. Some will.
LiIon cells under fault conditions undergo heavy self discharge at a point which produces Hydrogen gas and first molten and then gaseous lithium metal. Temperatures rapidly rise to the ignition point "and away it goes". Once started the reaction usually runs to completion. Water is welcomed as an additional reactant by the Lithium metal.
Lithium ion safety
Lead Acid, NimH. NiCd, NiFe, LiFePO4 ...
can all do significant thermal damage and will probably damage the cells if you short them and leave them shorted. BUT they usually do not explode and are not usually known to self incinerate.
I have never seen any of these batteries fitted with internal protection.
Note that LiFePO4 = Lithium Ferro phosphate, is in that list. A LiFePO4 cell is about as safe as they come. You can drive a silver spike through its heart, if you must. It will not thank you, but also will not self dismantle. Very briefly, the reason is that the Li metal is held in a spinel structure in the cell proper and does not migrate physically when the going gets exciting.
A lead acid "car battery" will melt about anything metallic that you place across its terminals. This would include eg large crescent wrenches. If you got badly hurt by bits of molten wrench you should not be surprised. The battery would probably never be the same again, but will probably not melt down. You CAN literally explode a lead acid battery by igniting hydrogen generated during charging, but that is a separate issue.
Add say 6 x AA Nimh 2000+ mAh cells, various coins and some keys to a trouser pocket and go about your business. On occasion you will get a short via coins and keys and misc cells so hot that skin burns could result and fire would not be a surprise. If you manage to do this you are an idiot! I've done it two or three times :-). Not again, I think!!!
NiCd about as for Nimh. More robust against shorts probably.
A fully charged 2000+ mAh AA Nimh will provide 10+ amps into a hard short.