But if not the battery, then what?

More likely, poor packaging in the engine compartment and exhaust routing generated excess heat. When combined with a fluid leak, that would be enough to create a fire, said Jon Bereisa, CEO of consultancy Auto Lectrification.

Bereisa was chief engineer of General Motors' EV1 and was the systems architect for the Chevrolet Volt, so he knows his way around these sorts of complex problems.

Bereisa has driven the Karma and has nosed around the car's inner workings. When he saw the cramped engine compartment of his test car, he was immediately alarmed.

"That engine is shoehorned into that bay, because they had to use a larger engine, because it was too heavy a car. As a result, there's no room for exhaust routing and heat shielding to route the heat away," Bereisa said in an interview.

Thermal condition

The Karma is "using the hell out of that motor-generator," Bereisa said. As a result, a "thermal condition" would be created under the hood or along the tightly packed exhaust routing path With that sort of heat, an oil, fuel or coolant leak can cause a risk of fire. A major ingredient in coolant is glycol, itself flammable.

Jeremy Gutierrez, the owner of the Fisker Karma, said he smelled rubber when the fire started, according to an account published by Autoweek.

Bereisa said: "You don't smell rubber with batteries, but you will if it's something on the engine."

Why else wouldn't it be the battery?

Bereisa said the battery pack's state of charge likely was mostly used up during its errand run, so it would take a lot of energy and heat to make a heavy battery pack hot enough to ignite. The compartment would have had to been breached -- unlikely for a car that showed no faults during its drive.

By contrast, the recent Chevrolet Volt crash-test battery pack fires were started when engine coolant splashed across a printed circuit board with live voltage, Bereisa said. When a coolant leak runs across a printed circuit board, it makes a conduction path and creates its own short circuit, making the board hot enough to ignite, Bereisa said. But no such event should have happened with the Karma.

"If the pack were to burn down the car, you would see where it started and reached the [battery] case," Bereisa said.

Like the rest of us, Bereisa is awaiting the fire department's official report, but he says, "There's more odds that it's a conventional, heat-related problem in packaging and heat-related leaks."

Fisker's position

So far, Fisker is denying that its battery pack is at fault for the fire, but has dispatched its own squadron of engineers to reach a conclusion.

Fisker spokesman Roger Ormisher said Thursday the cause of the fire still "has yet to be ascertained."

"There are myriad combustible materials that could be in the garage, in the wheel arch, or picked up on the roadside. They think the source is around the Karma, but they have not determined any cause yet. We have investigative teams, three insurance companies and the local fire chief all with their opinions. There are some question marks," Ormisher said.

Ormisher said he did not want to debate Bereisa's theories, but he said that, "The Karma has been through all regulatory and certification checks."

Robert Baker, the chief fire inspector for Fort Bend County, Texas, told Autoweek that the fire started in the Karma, not elsewhere in the garage.

"Yes, the Karma was the origin of the fire," Baker said. "But what exactly caused that we don't know at this time."

This article was originally published in Automotive News.

Content provided by Autoweek.