Perhaps, there's also the fact that a good amount of energy is lost in the mechanical drivetrain (transmission, differential gears, etc.) compared to a fuel cell system that might have electric motors directly in the hubs, but still, I would expect the ICE system to be much more cost effective for a very, very long time-- especially if there are no major breakthroughs in terms of battery technology.

The major flaw I've always thought of in the whole H2 scheme is not the fuel cells themselves, so much as the fact that the process of making H2 is somewhat wasteful if all you're going to do is run it in a car -- you have to take a bunch of methane, then extract the hydrogen from it, a highly energy-consuming process. You'd be better off just burning the methane in an engine, especially at the rate engines seem to still be improving.

Perhaps a better solution would be to use heat recovery on a hybrid-electric vehicle to spin a generator instead of directly spinning the crankshaft. Of course, with the heat-recovery parts, it may be hard to even fit a hybrid drivetrain into a vehicle at all...

Diesel is an incredibly poor candidate for this for two reasons. First is that there's almost no naturally aspirated diesel vehicles anymore, they're nearly all turbocharged. Turbocharging actually increases volumetric efficiency under load since you're recapturing waste energy and turning it into artificial induction. This lets you get away with running a much smaller engine for any given peak power output. As such, the turbo soaks up a ton of the heat and exhaust energy potential before it would even have a chance to reach the exhaust heat exchanger.

The other major "problem" is that diesel is inherently very thermally efficient. While you may see peak exhaust gas temperatures of 1500-1600 degrees at the exhaust port on a gasoline engine, most diesels run peak EGT's closer to 1100 degrees. In both cases, low-load EGT's are much lower; for diesel they are FAR, FAR LOWER due to the super lean mixtures. Factor in the heat loss from being downstream of the turbocharger and the temps may be so low as to be completely unusable. Gasoline is a more suitable alternative because it generally is NOT used with turbocharging and it has inherently higher EGT's to start with.

Hi AZCivic, the low heat availability of diesel engines is something I'd been thinking about too. One of the most often complained about features of diesel models over here is that in the winter the cabin heater is often pathetic compared to the heater on a gasoline car. (In fact this is such an issue that Honda actually engineered a heat pump into their new diesel engine to overcome it!)

But even with a diesel, 55-60% of the incoming chemical energy gets dissipated into heat. While I accept that a lesser proportion of that will be present in the exhaust gases (for all the reasons you stated), there is still a fair amount of heat to be had in the water coolant system (which the BMW system also uses). Downsize the main radiator and switch to a low boiling point working fluid (like propane, maybe?) and I think a fair amount of energy could still be recovered, despite the lower temperatures.

But I do agree, the gains won't be as noticeable as with a gasoline engine.

http://mysite.wanadoo-members.co.uk/ecotech/ecot.htm

Clett-AZCivic-Cummins is doing something like this with a TD in an attempt to reach 50% efficiency.They are using the exhaust gas-heat-to vaporize a liquid(the name sounded like it was some sort of flurocarbon) and this liquid then drives a small turbine.My impression is that this turbine then "hooks" into the transmission and drives whatever is being driven.Uhh,it isn't being used on a moving vehicle-some sort of in place Diesel driving whatever-pump etc.I guessing it is too heavy-complicated-bulky-to use in a car-truck etc.
I THINK it is a turbo diesel,so it is really working the exhaust gas hard.Charlie
I think they called it a RANKIN OR RANKINE??

Rankine cycle is the regular steam cycle- although any condensable fluid can be used. Not all that different from a heat pump/refrigeration cycle, except the energy is used to do work instead of simply move heat around.

Refrigerants make very good working fluids (replacing water) for low temp-pressure applications. Good chance that is what would be used. Water has many practicality issues that don't exist with refrigerants.

Clett- look at the BMW animation- the secondary/low temp loop is heated by jacket water as you suggest, superheats in the exhaust, and rejects to the radiator post-turbine. The radiator looks like it is completely removed from the primary engine cooling loop. (The primary loop flashes in the exhaust, superheats further up in the exhaust, and rejects to the 2ndary loop)