Makes me wonder how transmitting 1 kWh via power line from a plant to a car, would compare with transmitting 1 kWh of gasoline from a refinery to a car. I know it wouldn't really matter, but it'd be an interesting number.
This is one of those places where the math is a little surprising. I was all ready to estimate that the fuel cost of transporting fuel was on the order of a few percent, then...
Some time ago on another project, I worked out an efficiency number for semi trucks--around 100 ton-miles per gallon of fuel. Based on a very limited search just now, that might be a little low, but it makes the math easier. Semi trucks are also the least efficient means of transporting fuel (as opposed to rail or pipelines), so these are absolute worst case numbers.
A typical double tanker truck is about 11,500 gallons, gasoline is 6.3 lbs/gallon, so a load of gasoline is right around 36 tons.
For the sake or argument, let's say that the gasoline is moved 150 miles by semi truck. That's probably really high except for very rural areas--most places are closer to an oil pipeline than that.
That rate means that the tanker truck would burn 54 gallons (36 tons * 150 miles / 100 ton-miles/gallon) to move 11,500 gallons of gasoline to its destination, or about 0.5%.
I will grant that this doesn't account for movement of bulk oil to the refinery or moving refined products to terminals for road shipment, but I also note above that I'm making some very unfavorable assumptions about road transport as noted above. Based on these numbers, I'd be surprised if the total transport cost of gasoline/diesel was more than 1% of the total.
That said, I can't exactly dispute
@mach7's numbers of 8-15% electricity losses in power transmission, but I do note that the federal government thinks it's more like 5%. I can't say from the level of information available whether the federal government numbers include all of the losses in the post the Mach7 cites.
https://www.eia.gov/tools/faqs/faq.php?id=105&t=3