The Cult of Cars, Racing and Everything That Moves You.
We may earn a commission from links on this page

How Koenigsegg Nearly Built Subaru's Flat-12 F1 Engine

We may earn a commission from links on this page.

My obsession with Subaru’s failed F1 engine, a flat-12, has gone on for years, and I had often heard that the motor nearly went on to power Koenigsegg’s first supercar. Now I’ve finally heard the full story.

I should have heard this a year ago; that’s when Christian von Koenigsegg first explained it. But I missed it tucked away on Koenigsegg’s own blog. I’ve been busy listening to how the totally-not-a-Bond-villain Swedish supercar genius has been explaining why gravity works backwards, why camshafts don’t need to exist and other technical points.


In any case, here’s as much of the history as I can explain. Back in the late ‘80s and early ‘90s, the Japanese economy was in something of a bubble period and Japanese carmakers were on a tear developing some of the most extreme and fantastically pointless cars in modern history.


In the world of road cars, we’re familiar with how Honda took on Ferrari with the NSX and how Toyota took on Mercedes with Lexus. The stories from racing are a bit more arcane. We do remember Mazda’s victory at the 24 Hours of Le Mans with a quad-rotor Wankel engine, but few recall how Subaru briefly entered Formula 1 with a flat-12.

It was a 3.5-liter naturally aspirated engine designed by the Italian legend Carlo Chiti, who had designed the flat-12s that powered Alfa Romeo to a pair of world constructor’s championships all the way back in the mid 1970s. This Subaru-branded engine was at the tail end of Chiti’s career, and it was never successful. It weighed too much, didn’t make enough power, and its wide flat design meant that it compromised the aerodynamic underbody shape of the Coloni F1 car it was built for. The engine was so bad, it never even qualified for the starting grid at an F1 race. It was ultimately abandoned before its debut year was over, with a string of firings in its wake.

Let’s watch this thing fail, preserved on video forever at Silverstone 1990:

Most histories of this engine explain that the rights to the design got bought up by Koenigsegg at the end of the decade, but few explain exactly how that all worked. I assumed it was a speculative purchase that never went very far for Koenigsegg. I was wrong.


Christian von Koenigsegg himself explained how close he came to running the engine in his first production supercar last fall in this ask me anything-style blog. Koenigsegg said that he met a friend of Carlo Chiti’s in the mid ‘90s, a few years after the Coloni-Subaru F1 effort flopped. This was also right after Koenigsegg’s deal to use tuned Audi V8 engines, like the one in his first prototype, fell through.

While the flat-12 that Chiti and his Moderni Motori company built wasn’t good for racing, it seemed like it was a surprisingly viable choice for his road car, as Koenigsegg remembered:

So we had this prototype, based around a particular setup that we thought we could use and we had no engine supplier. We could have sued but I didn’t want to go down that path.

I met a guy who knew Carlo Chiti via a chain of friends. Chiti ran an engine company called Motori Moderni. They were in a bit of trouble but they used to do Formula 1 engines for Minardi. They had this boxer-12 engine, made in cooperation with Subaru, that had hardly ever raced. They had trouble with the weight and getting the diffusers to work with the layout of the F1 car once the engine was in.

The 12 cylinder engine wasn’t successful in Formula 1 but Motori Moderni had great success building other engines for Alfa Romeo to use in DTM racing. They had proven their worth, so they were worth talking to.

We went and met with them and they were very open. They said they could modify their 3.5 litre flat-12 engine for us. Similar engines had been used in offshore boat racing with twin turbos on them, so we were confident enough in their durability. They put together a 3.8 litre flat-12 and they lowered the RPM from 12000 to 9000. They put different camshafts on it, they stroked it, put in longer intake tracts. It was set up to get 580hp at 9000rpm and I have the dyno tests from where we ran that engine.

When we first designed the Koengisegg monocoque, it was designed for that engine and it was the first engine we put in there. The engine mount positions for that engine are still used in the Agera today.


Koenigsegg even uploaded a picture of one of the engines, huge and low and wide:


What amazes me is what killed the deal. The motor itself seemed fine. Why it never worked its way into production was no fault of the engine at all:

Unfortunately, around this time, Carlo Chiti died and the company filed for bankruptcy. We had received two engines from them (which we still have) but once again, we found ourselves with no engine supplier.

In early 1997, I ended up winning an auction to buy some of the company’s assets. I got all the tools, the drawings, the castings and some spare parts for the engines. Some of that went up in flames at the old factory. We thought that maybe, with all that equipment, we could build our own engines from Chiti’s designs but when we got the equipment back to Sweden and started sorting it all out, it was a nightmare. There we no computerized drawings. It was all drawn by hand. A lot of the tooling was old and made from wood and it was quite beaten up.


I should not be greatly surprised that the high-tech Swedish Koenigsegg was not compatible with the low-tech Italian wood tooling. But again, the engine itself sounded great, as Koenigsegg himself said:

Would the engine have been viable for the future? Certainly at first. It was quite amazing, actually. The whole engine block was under the centre of the rear axle, which gave us a super-low centre of gravity and looked very cool when you opened the rear hood of the car. There was no vibration whatsoever, which is why we decided to bolt it to the monocoque – it was wide, low and solid, so it acted just like a chassis member. We still do that with the V8’s we use today but with a tiny compromise because they’re not as smooth.

The downside of that engine is that it would never have taken us to the level we are at now. It would have been good up to about 750hp with turbo, but that would have been it. As it turns out, the engine package we finally engine layout we went for in 1998, has allowed us to go much further than we could have with Chiti’s flat-12.


Finding any good picture from Koenigsegg of their work on this engine is difficult. I spotted this picture of an engineer working on a full engine with what looks like a CC prototype in the background and I found this picture of an engine mounted in a Koenigsegg car, complete with extremely wonderful multicolored tennis balls protecting the individual throttle bodies. I couldn’t find any full pictures of the outside of this running Koenigsegg prototype at all.

So there are still a few mysteries surrounding this incredible engine that never was. I would have loved to have heard it run.