Engine Costs (and price):
(A continuation/expansion of comments above in my post #388)
According to BMWfans.info, which I believe is just data directly from BMW repackaged, a 2009 (no newer data) S65 motor costs about $22.5k, an N54 costs $13.5k. Now this is the cost to a consumer. We can not deduce actual cost to the manufacturer from this but we can get a rough idea. Obviously these consumer costs are much higher than actual cost as you can not have an engine that costs over $20k in car that has a base sticker cost of $58k. You also have both OEM and dealer profit margins included in these figures. This factor of almost 2 between the engines cost should not be surprising. The N54 is less expensive because (not counting the FI system):
- Has less parts
- Key parts are smaller
- Entire engine is lighter (293 lb vs. 330 lb)
- Last and most importantly - the volume produced is immensely higher
Obviously the turbo system and associated plumbing contribute to making the engine more expensive but this is far outweighed by the economy of scale.
Now we know the version of the N55 in the M3/4 will be fairly heavily massaged. Probably a different head, different turbos, more costly cooling system and maybe even some internals. This will raise its cost above the base engine. What I would be willing to guarantee is that the first digit of the total replacement cost of this new engine will be 1 not 2!
The twin turbo V8 from the X6 and 7er (again 2009) has a cost of $21.9k. It cost more than the N55 due to the same points above but the opposite:
- Has more parts
- Key parts are larger
- Entire engine is heavier (354 lb vs. 293 lb)
- Last and most importantly - the volume produced is much smaller
This should answer definitively why a V8 turbo is not in the M3. BMW M does not believe the V8 is required in the application for prestige nor to make the required power and the N55 based engine will be substantially less expensive.
An educated guess on what factor these consumer costs are compared to actual production costs is not easy. Thus I would offer a reasonable range. Call this a bounding argument. I think the total markup factor is between 2-4. Thus I would place the cost to BMW of the N54 somewhere in the $3.5k-$7k range (probably closer to the lower end). Similarly the cost of the S65 is likely in the range of $5.5k-11k. That means the likely savings of a N54 priced N55 (using N54 cost estimate) is about 100,000 x ($2k - $4k) or between $200-400 million. 100,000 is the rough production volume of the M3 over its lifespan. This does not account for the other vehicles that may get a slightly tweaked version of the M3/4 specific engine.
I have seen multiple references (internet/google, nothing special/secret) that peg complete new engine development at a major OEM in the $1B dollar range. However, I have also seen more references that an entire vehicle development costs that much. I think the latter is much more accurate. $1B is a stinking lot of cash. That is $100k in salary and $100k in facilities for about 3500 workers working for 18 months... (typical new vehicle development time). An engine is however a truly major component of a new vehicle development. At the big OEMs a figure on the order of $100M is probably reasonable.
With these numbers I would revise my prior statements that the cost of the economy of scale likely dwarfs development costs. I think this is a very reasonable argument that for these particular engines, the savings from taking the M3/4 to an I6 from NA V8 are on the same order as the development cost of an entirely new engine. However, the development cost is largely a moot point since we are talking about a massaged N55 NOT an entirely new engine. Certainly though when considering the possibility of another vehicle specific M engine, possibly an evolved direct injected S65 vs. a massaged N55 we are clearly talking about hundreds of millions of dollars of savings in development and production. This is clearly THE primary reason why the car does not have an existing turbo V8 nor a totally new M specific engine.
I am open to anyone's estimates or data to revise these numbers. However, because of the bracketing and large range of the concluded figures it is probably a very reasonable figure. The number here is certainly not $1B and is also not only $10's of millions either.