White Paper - Capacitors across a battery

How to do it. Does it make sense?

Introduction

Batteries are good for energy storage, while some capacitors are good for power storage.

Naturally, some have proposed ways to combine the two to obtain benefits from each.

However, some approaches are technically ineffective, or not economically viable.

Directly in parallel

For there to be a benefit, the system must be able to use the energy in the capacitor: the capacitor must be able to charge and discharge.

By connecting the capacitors directly across the battery, the voltage of the capacitors is constrained by the battery; since the capacitor voltage cannot vary* no energy* can be extracted from or returned to the capacitors. Therefore, the capacitors have no effect*.
* (not much, at least)

Therefore, direct connection of a capacitor to cells or to a battery is ineffective.

The BMS doesn't care if you use capacitors.

There are two approaches:

• A low voltage capacitor _directly_ across each cell
• A high voltage capacitor across the entire battery

A low voltage capacitor _directly_ across each cell: as far as the BMS is concerned, the parallel combination of a Li-ion cell and a super-capacitor is no different from a simple Li-ion cell without a capacitor

A high voltage capacitor across the entire battery: the BMS will handle the battery and won't have anything to do with the capacitor.
(For a string of ultra-capacitors, you'll have to come-up with some way to prevent overcharging each individual ultra-capacitor.)

Through DC-DC converter

For there to be a benefit, the system must be able to use the energy in the capacitor: the capacitor must be able to charge and discharge.

The only way that the capacitors can charge and discharge is if the capacitor voltage is allowed to increase and decrease as energy is extracted from and returned to it.

The only way to allow the capacitor voltage to increase and decrease is if the battery bus (fixed voltage) is isolated from the capacitor bus (variable voltage).

The only way to isolate the two voltages is through the use of a bidirectional DC-DC converter between the two buses.

The DC-DC converter isolates the battery bus (fixed voltage) from the capacitor bus (variable voltage), therefore allowing the capacitor to charge and discharge.

In general, the cost of complexity of a bidirectional DC-DC converter capable of such power levels, plus the cost of the high power capacitors, are such that, in practice, it's cheaper, smaller and simpler to replace the Li-ion cells with other cells that are capable of delivering the required power. There are exceptions, yes, but, for most applications, using good, high power cells is better.

That's why you hardly ever see any production level product using capacitors directly in parallel with a battery. When you do see such designs, is almost always in prototypes that are never brought to production, because the designers soon realize it's a bad idea.

Conclusions

In most applications, selection of good power cells is a better approach than coupling batteries and capacitors.

In the rare exceptions, you must use a high power, bidirectional, DC-DC converter between the battery and the capacitor, which is bulky and expensive.