You have to conduct or convect or radiate the heat away. A good medium is highly turbulent gas or even liquid depending on the amount of heat generated. This is why cylindrical batteries like the Li-ion 18650 are still used in large assemblies as the gaps between them allow air to circulate and limit the temperature rise.

Amrit,

At a pinch you can add phase-change materials to the structure that the battery pack is made from (encapsulated PCMs such as these have been studied):

http://www.mikrocaps.com/products/microencapsulated-paraffin/

Plenty of papers, a couple of patents:

http://www.puretemp.com/pcmatters/patent-application-lithium-ion-battery-system-having-temperature-control-function

But fundamentally, you still have to dump that stored heat somehow. So, as Tony said, you have to use a working fluid (air or a liquid) to shift heat either to the outside world via an open air path, or to a heat exchanger.

What stops you from using a conventional liquid cooling loop and a fan-assisted heat exchanger? Or, (again, as Tony suggests) just a simple blown-through air system?

The space will most probably be partially closed, however, changes can be made to create an increased air flow based on recommendations. However, is there any known system that can successfully cool down a battery in a mostly confined space, for example, inside a car? I don't want to rely on air flow entirely to cool down the batteries and liquid cooling seems viable but it can only cool the battery to a level. I expect the battery to get quite hot as it will be providing charge to electrical components while being charged itself at the same time.

Amrit,

The greater density of liquids over gases means that a liquid cooling loop can be very small in diameter and yet, for practical flow speeds, still transport heat rapidly to a suitable heat exchanger.

"... but it can only cool the battery to a level. "

?:)

A liquid cooling loop, feeding a heat exchanger, will result in the same final temperature as a blown air-cooled system - but will transport far more heat per unit time for a given flow speed.

> However, is there any known system that can successfully cool down

None that have not already been endlessly optimized and which are on the road already.