I've got a Prius MPG problem that I'd like to relate to the community
for some feedback and ideas, because the evidence is sort of bizarre
and inexplicable.
.
This is also being posted to the prius_technical_stuff yahoogroup.
.
I've been doing fairly well over the winter in my '04, turning in tank
averages across the high fifties to low sixties MPG, but a friend with
an '05 has only been managing 45-ish MPG averages despite trying to be
a little more mellow on the road. [It's Jesse aka "300kmile prius", who
some of you know.] Now, he does have a somewhat more aggressive driving
style than me, but he's been trying to think about MPG a little more
lately and both of us agreed that the difference in results shouldn't
be *that* far apart in what's essentially the same car. I've got 102K
miles on mine, he's up around 150K.
.
Well, yesterday we got to explore that further. As a test, I drove
us down to an event and back, a couple of hours each way, in Jesse's
car, and my results were in the crapper too, making us think that
something's amiss with his Prius. It was raining for much of the way
down on the first leg of the trip but the roads were close to dry on
the way back, but even with *my* right foot in play I could only pull
49 or so MPG. Heck, on my run down to meet up with everyone else
earlier that same morning on a fresh tank, I was showing 64.0 on
arrival [which with my known +1.2% tire revs error should still
equate to about 62 real-life].
.
There are no obvious differences between the vehicles. Their setups
are the same as far as drivetrain -- stock system, a little extra
instrumentation added. '04 and '05 have the same minor ECU bugs
with regard to how they report certain parameters via OBD-II, but we
know how to ignore some of that. He's running 0W-20 oil where I'm
still running 5W-30. Similar tire pressures, all mid to high 50s,
but he's running the stock Integrities where I'm on my 50,000+ mile
old Hydroedges [which anecdotally deliver lower MPG but I think that's
basically bunkum once they've worn in and are run at decent pressure].
No hints from the helm of any misalignment, and tire wear looks
normal and even. With three of us and a bit of luggage in the car
we might have been running a little heavy, but not that bad.
.
I set up an Xgauge on his Scangauge to monitor fuel injector open
time, as detailed in my
writeup on rudimentary CANbus analysis,
and the Xgauge programming I'm using now has actually evolved a little
past that to be more selective:
_ 07E0 21F3 / 0306 8461 05F3 / 3808 / 000A 0008 0001 : inj
and to add the +1 offset that seems to make it more accurately match
the professional scantool. I've been using this as one of my main
guidelines for several months now, and it really helps to determine
the boundaries of efficient running.
.
The parameters that I've come to consider "normal" are thus: injection
time reaches about 5.0 or 5.1 milliseconds as the engine becomes fully
torque-loaded and I see the manifold vacuum drop, and stays on
something of a plateau between 5.1 and 6.2 - 6.4 ms over a wide RPM
range from there on. On pushing harder with RPM climbing toward
3000, I see peaks of 7.0 ms. We have a fair amount of evidence at
this point that this is the most efficient running range for the
Prius engine, and on the highway where the battery and electrical
system is largely out of the picture, it's all about how the engine
is run that dictates efficiency. My own driving has been lately
influenced by monitoring RPM, injection time, and kilowatts, and
the fact that I've been turning high fifties and low sixties with
significant highway travel through a fairly cold winter [and only
sporadic block-heater use] bears that out. I'm eagerly waiting
for warm days and summer gas blend to get here, and start my
fifth-year graduate-level curriculum on Prius driving.
.
When I lock my foot at 2000 engine RPM on the highway, letting the
terrain determine my actual travel speed within reason, in my car I
see 6.2 - 6.3 ms injection time, maybe as high as 6.4, and speed
tends to stay generally between 55 and 65 MPH over gentle terrain.
This is what I refer to as the "60@60 point" where MPH meets MPG.
.
In Jesse's car, at the same 2000 RPM and observing about the same
range of travel speeds and performance, injection time was more like
6.8 - 6.9, with occasional excursions to 7.0 or more. And the highest
I could push the average on the MFD was 49.5 -- I never broke 50 MPG
the entire day, in fact, even on the backroads near our destination.
I was pretty much driving the way I always do, and the car felt
like it was doing about what I'd expect at any given RPM in terms
of power output and speed of travel. But there was this notable
disparity in fuel usage -- 10% more squirt-time in his, basically,
leading to something approaching a 20% fuel-economy difference.
I think that's more than the passengers or the weather would have
donked it; something has to be wrong.
.
We stopped carefully [regen-only as well as I could determine] a
couple of times so I could get out and feel all the brakes, seeing
if I could detect any heat from one dragging. All stone-cold, and
if that was a factor the speeds I observed from holding at 2000 RPM
would likely have been lower anyway. RPM vs. ballpark speed and
the miscellaneous seat-of-the pants determinations in his car
matches what mine does; it's just injector time that doesn't.
He's got a relatively new air filter, and we changed his plugs
and transaxle fluid at about 120K a little while back. I didn't
check the oil *level*, in hindsight.
.
I didn't have all my Scangauge notes with me or I would have set
up to monitor his short-term and long-term fuel trims, to see if
any of that was off but not off enough to throw a MIL. Maybe next
time we get together. What I'm looking for here, from those who
have bothered to read this far, is some additional ideas on what
to check for. Compare TPS deltas between idle and steady-state
running? Inspect for vacuum leaks? Dump all the crap out of the
car and try a solo loop on a nicer day? That sort of thing, geared
toward whatever might affect MPG in a relatively serious way.
.
_H*
- Dale
Linear Mode
