The dealer put on new tires (Ecopias) and changed all the fluids. The oil still has that clear/new appearance. I was not trying to hypermile but I did notice something peculiar the other day. There is about 3/4 or 1 mile (never measured it carefully) downhill stretch on the way home. A couple of days ago I finally figured out the pedal pressure to make the display arrows go black and I hit all the lights right for once. SOC fell two bars over that distance and maybe 4 minutes. The car was on, the radio was on, but the headlights weren't. So not a huge power draw, I would think. I would have thought that a healthy 12V alone could have supplied that current, but I think this one is very much on the way out. That much SOC charge drop for so little load does not give me a warm and fuzzy feeling about the big battery. Regarding the 12V battery, using the "signal check method" in ACC-ON first thing in the morning (car sat overnight, ~14 hours) shows a voltage of just 11.1V. Seems mighty low for 12V lead acid battery!
Pop the hood at night, then in the morning: check battery at jump point, with a digital multimeter. Better yet, something like Solar BA5, an electronic load tester. Post your findings.
12.39-12.40V after drive home 12.24V 17h later Car completely off both times. Obviously this battery is on its way out. Depending on the table referenced that last measurement is about 50% SOC. A healthy car would be close to 100%. The dealer says that if I bring it in they will load test it and if it doesn't pass, replace it. (Because the million point check list said the charging system was within spec.) Load test instructions usually say that the battery should be at at least 75% before running it, which it might be, barely, for about 2 minutes after the car gets to their location, otherwise, too low to test. One of the two tabs that holds the fuse box lid down is flopping around, it mostly peeled off the side of the box and is now attached only by a narrow sort of hinge at the bottom. The lid is really only held in place by the latch at the bottom and the other tab, the broken one isn't doing squat. Anybody ever fixed one of these tabs? Superglue, epoxy, or something else?
12V battery replaced this morning by dealer (for free). On first startup signal check showed 12.5 V and in ACC-ON it only went down to 12.2 V. Much better! Probably won't make a bit of difference in the mpg department, but will post an update if it does. To balance our "Carma" my wife's late 90's Accord lit its CEL the very same morning, while I was driving back from the dealer. (It was an evap system code - reseated the gas cap, cleared the CEL, and it stayed out on the next restart.)
Filled it again today, first full tank after new 12V battery and after the compound meter board failed and was replaced. Mostly the same commute. It was at 392 miles shortly after reaching the 1 tick mark, fill up (at the same station and pump) was 8.2 gallons, which is 47.8 mpg. This was in perfect agreement with the consumption display. Between the new battery and the new compound meter board (the one for speed, odometer, etc.) it seems most likely that the improved mpg resulted from changing the former. Others have reported this effect with weak 12V batteries before. Still, ~9mpg improvement with a new battery versus the old one is pretty amazing. I think ~48 mpg is about what this car should be doing under these conditions, when it is working right. It would be hard for it to do much better since the car spends such a large fraction of these short trips warming up.
What do you mean by that? The MPG ratings and tests are defined by the EPA, not by the manufacturers. It has been the EPA, not the manufacturers, repeatedly adjusting the scale down to address many 'typical' consumers' complaints about not being able to reach the ratings, when the real problems are ever higher speeds, more aggressive driving cultures, and greater congestion than when the original scale was created in the 1970s. That leaves a lot of room for more skilled drivers in better conditions to beat the EPA figures. The car manufacturers don't sandbag their ratings, but instead do their best to game the test requirements to get higher numbers. The U.S. test requirements leave less room for these games than do (or did until recently) the European test rules, so are more realistic. Some car makers have actually been caught lying or stretching rules beyond reason, and forced to reduce the numbers for certain cars and give previous buyers refunds or credits towards their fuel purchases.
I'm interested to know what the readings have been on subsequent fills. After the major service coupled with the replacement of the 12v and HV batteries my first fill jumped from 38 to 48 mpg on the first tank over a well documented route. Since then it has settled in at about a 10% improvement from what it was. That first tank was probably due to a reduction in the desert heat not allowing as much fuel into the bladder. As has been so well stressed over the years here it's really necessary to take the figures over multiple tanks if you have the U.S. version of the 2nd generation Prius.
Ford, Kia, VW and others have been posting false claims of high mpg's on car window stickers at dealerships and have been charged. The EPA does not do this, the car companies do.
The later statement does not clarify the former claim. The EPA sets the tests that the car makers must use. The EPA does its own testing on only a certain fraction of the models for auditing purposes, and several car makers have been snagged for posting incorrect numbers. Some for fudging high, some for making use of old rules for 'similar configurations' that the EPA has since disallowed, perhaps retroactively. I am not aware of any carmaker that has fudged its numbers low. That is not how some drivers get better MPGs than their cars' ratings. Beating the EPA numbers can be easy. Beating the higher CAFE numbers (which used to be the EPA numbers too, back in the 1970s and early 80s before the EPA started changing its scale) is much harder.