A basis for further discussion.
Warning: this is undoubtedly incomplete.
Ego left at the door; a couple of weeks chewing over by the brains who visit this site will soon have these
(guided) ramblings knocked into shape.
Acetone to improve bd, svo, wvo mpg.Couldn't see this myself, from a theoretical viewpoint,
but there again I'm not a diesel engineer.
At the addition rates mentioned there is insufficient acetone to react with biodiesel,
or to cause the biodiesel to decompose into other combustible species.
A bit of ferreting around tinterweb.
Have not got complete story as much of the detailed info is only available from papers of (costly £ or more frequently $)
of Learned Societies.
So heres a few links and a (over?)-simplified explanation.
(Got myself hopelessly between biodiesel, svo and wvo:
if in doubt invent a new acronym: vodm : vegetable oil derived molecules.
I know all three are different but for the purposes of this they are close enough and share the share features which distinguish them from fossil diesel.
Vodm is worse than fossil diesel on two readily observable counts;
1) volatility, (or the temperature at which vodm volatilises), cold winter starts aided by fossil diesel or unleaded
2) specific heat of volatilisation, (how much energy does a molecule of vodm require to become gaseous compared to fossil diesel)
diesels on vodm run a bit cooler because more energy from combustion is needed.
Apparently modern diesel engines have two stage combustion.
Fossil diesel tends to be smaller(shorter) molecules than vodm.
The smallest molecules in fossil diesel, (the most volatile), promote the first stage of combustion,
and allow the second stage to start sooner and at a lower temperature.
This results in more complete combustion of the fuel, and reduced soot particles.
Compare this to vodm.
The long fatty chains wait around for the compression and temperature to rise,
and when they do start to combust it is later in the engine cycle.
This results in more soot production, as not all the combustible material can burn in the shortened
timespan available.
More soot equals less energy produced equals less mpg, (quite apart from the fact that around 10% of vodm cannot be burnt).
Acetone to the rescue.
Acetone: very volatile, at low temperatures (for a diesel engine), and importantly it contains the =C=O fragment.
This is kind of carbon monoxide (in-waiting), and is a stable fragment (in terms of what goes on inside a diesel cylinder).
So pop your apparently miniscule amount of acetone into your vodm.
The acetone volatilises on exit from the injector (probably irrespective of the ambient or local temperature).
The vodm will not even notice the small amount of energy used by this process.
Acetone will start to combust at much lower temperatures than any of the vodm.
As the acetone starts to combust, (first stage combustion previously either missing or much reduced),
many of the =C=O fragments are produced.
As the temperature increases some of the smaller vodms start to burn, but incompletely, until we reach the second stage combustion.
Now all hell breaks loose, but we got there quicker than in the absence of acetone, so the vodm has longer to combust
and can combust more fully.
Those partially combusted smaller vodms, they were held by the =C=O and prevented from condensing into soot.
Less soot, must mean more vodms fully combusted, more energy liberated, more power generated, more mpg.
Remember I am not a diesel engineer, so some of the above is almost certainly partially ubderstood mashed potatoes.
Heres a few links:
http://books.google.co.uk/books?id=0jCyWkETt6wC&pg=PA399&lpg=PA399&dq=first+stage+ignition+chemistry&source=web&ots=QBkhs9a74S&sig=o-7EJcpg6m5bsczlad9v73ezXLw&hl=en#PPA385,M1http://www-cmls.llnl.gov/?url=science_and_technology-chemistry-combustionhttp://www.aip.com.au/industry/fact_alt_fuels.htmhttp://www.eia.doe.gov/oiaf/analysispaper/biodiesel/ 
