My take on how it should work...
Okay, here's what actually happens on the road...
http://www.autospeed.com/cms/A_110216/article.html
and refer to pretty picture...
Now, I'm picking favorable conditions here, to show how it can work, any idiot can pick conditions where it won't (Mythbusters probably qualifies here)
Now load is scaled in BMEP, max is 150. Imagine a vehicle cruising at 20% load on the highway, that would be 30 on the side of the graph, using about 20HP to keep moving at 3000 rpm 60mph, that puts us in the magenta region at a BSFC of about .65 lb/bhp-hr, a gallon of gas weighs about 6.5lb, so our car is getting .65x20HP=13lb/hr, dead on 30mpg. Now we've got a decent sized alternator in the thing 160 Amp, sounds a little large, but that size is on many light trucks now and is optional on vehicles like the Dodge Magnum, dB draggers are putting 200+ amp units in. Using this size because it makes the numbers convenient, it gives us about 2250W, which is around an even 3HP, which at 60% efficiency makes for a 5HP load on the engine, when fully loaded, so we fully load it.....
So we've increased load by 5HP, so now we're using .65x25HP=16.25lb/hr = 24mpg right? WRONG
No, we're going back to our graph... we had to crack the throttle a little more to hold 60mph, so we've reduced pumping loss... so reading across from our new 25% load, 37.5 on the left, ye gods! now we're using about .55lb/bhp-hr so .55x25HP=13.75lbhr, 28.4mpg...
So we're effectively getting 2250W of useable electrical power at a "cost" of only .75lb of gasoline an hour...
So, the electrolysis process for H2 is about 80% efficient in terms of energy input, however, ~20% of the energy comes from environmental heat, usually presumed to be at a temperature of 20-25*C.... but our electrolysis tank is sitting in a toasty warm engine bay, soaking up a proportion of that 75% of the wasted energy in the gasoline that doesn't get to the road. If the electrolyte temperature gets up to about 40-50C from heat soak, we can expect about 40% of the energy of the process to come from heat, electricity needs only to supply the remaining 60%. So we need to supply 1.25x the energy we get out of the hydrogen, to overcome the 20% efficiency loss, but since 40% is supplied by heat, we only need 75% of the energy to come from the alternator... so 2250W of electrical energy, makes 3000W worth of hydrogen...
So, the energy before losses in that extra .75lb of gas we're burning is about 3800Watt-hours.... we see from the graphs here
http://www.transportation.anl.gov/tr...en_engine.html that 50% efficiency is achieved by pure hydrogen in conversion to mechanical energy in an IC motor, where it's limiting factor is the mechanical limits of IC engine design, in diluted quantities, it is reasonable to expect it to exceed 50% efficiency, perhaps approaching 80% or more. So at best case efficiency for the gasoline of 25%, and at this point on the graph (3000 rpm 25% load) we're only seeing 20% efficiency, that 3800Wh of gasoline is only worth 950 Wh at the wheels, whereas our 3000Wh of hydrogen would be worth 1500Wh at the wheels, this is best case for gasoline, medium case for H2, therefore our H2 is worth the equivalent of 1.2lb of gasoline... 13.75-1.2=12.55lb/hr consumption, our mileage went up to 31mpg...
However... doing it as a "better case" for the H2 at 80% burn efficiency and using actual gasoline efficiency at this load point of 20% the gas is only worth 760Wh at the wheels, where the H2 is worth 2400 Wh at the wheels... making the H2 worth 3.16lb of gasoline, taking us up to around 37mpg of actual gasoline consumption.
This also neglects synergistic efficiency increases of the H2 and remember the "extra free" O2 on the burn efficiency of the gasoline mixture, and the improved efficiency of heat to cylinder pressure conversion which is likely to occur with extra steam in the chamber as a working fluid, all which may be worth another couple of mpg at least...
There is no "free" energy here, it's just improving the efficiency of the system as a whole, the car is getting 30mpg at 20% efficiency, that means it would get 150mpg if all that energy were harnessed as propulsive effort. All you're doing is clawing some of that back.
MIT knows that H2 increases combustion efficiency, they are working with Arvin-Merritor to crack H2 from gasoline in a plasma chamber called the plasmatron, they are conservatively estimating efficiency gains of up to 33% when 25% of the gasoline is cracked to H2. This process appears more efficient than electrolysis cells.
Remember the gasoline has to remain in the picture to provide the inefficiencies to scavenge the rewards from, you're not going to get your car running on 100% water electrolysis like this. On hot days with a favourable engine design (one with relatively high piston speeds) you might take a few hours to run your battery down with a really efficient cell, but it ain't gonna run constant. It could be said that using electricity this way is overall more efficient than many electric hybrid motors.... in warm climates at least.
There's gonna be a point where making more H2 doesn't get you any more MPG and MPG starts dropping, don't know where that is, most of these commercial kits are making relatively small amounts of H2 off stock alternators. Some of them seem to be giving apparently freaky results on small cars, like 70mpg in an Escort... well... the Escort could get 40mpg highway for a start, driven carefully, and with a relatively small motor has further to jump across the BSFC graph when loaded and is requiring relatively less H2 production to hit peak potential.
Other ways of getting H2 into the motor might be investigated to get better overall efficiency than these electrolysis cells. Waste heat might be scavenged directly for H2 generation, and then you don't even drop that .75 lb/hr that you have to make back again.
A possibility I will be investigating shortly is that ethanol can be cracked to H2 at relatively low heat, ~100C... with a catalyst... copper... doped with lanthanum (found in lighter flints) for improved specificity for H2 and reduced coking... so I'm going to copper plate my intake ports.. reasoning that it will crack the ethanol in the fuel to H2 as it is evaporating off the port walls, waiting for the intake to open (injectors are spraying at a closed valve during 90% of driving)
So anyway, if you wanna screw with it, do so, use a big alternator, put lots of cells in series for best results (Best efficiency is at low voltage per cell, electrolysis starts somewhere around 1.7V so 6 cells across 12V is about 2V per cell. you can probably go to 7 cells, since alt output is probably at 13.5-13.8V)
Road Warrior
Actual car that ran on water
Linear Mode
