Hybrid Car – More Fun with Less Gas

Hypermiling on a continually hilly terrain

register ::  Login Password  :: Lost Password?
Posted by Neo on July 5, 2011, 3:42 pm
On a hill terrain scenario where a vehicle is going
downhill from the top of one hill and then continue
without stopping uphill to the top of another hill
of equal height, the following hypermiling technique
can be used to save fuel/energy used regardless
of what type of energy the vehicle is using.

From the top of the hill apply energy gently and
make no attempt to increase velocity until the
vehicle is moving downward the top of the hill.
With the vehicle is pointed downhill, apply additional
energy to increase the velocity of the
vehicle so that when it is at the bottom of the
two hills its velocity is 1.5x the speed that
desired at the middle of the uphill climb
approaching. At the bottom of  hill pull back
on the energy being applies so that  vehicle
gradually decelerates to the desired velocity
as it passes the middle of the uphill climb.
Allow the the vehicle to further decelerate
to about 3/4 to the  speed when it was
in the middle of the uphill climb so that when
the vehicle is on the top of the next hill its
velocity is 2/3  of when it was in the middle
of the uphill climb.

Why does is this method more energy efficient?

One word - gravity.

At any given velocity V, it takes less power
to propell a vehicle downhill than it does
uphill.  If a vehicle is on cruise control
and going a fixed velocity on a hilly terrain
where all hills are of equal height then the
uphill energy savings will approximately
equals the downhill energy extra cost.  Hence,
if the downhill energy cost requirement, E1,
is about 20% less ( 0.80*E )  THEN the downhill
energy cost requirement, E2,  is about 20%
more (1.20*E)  where E is the energy requirement
to move the vehicle at a velocity V if the
vehicle was moving on a flat terrain. Hence
the total estimated energy requirement, ET, for
going downhill  and then uphill using cruise control
look like this:

ET = E1 + E2
     = .80 * E(v)  + 1.20 E(v)
     = 2 * E(v)

We also know that it takes more energy to move
a vehicle the higher the velocity. Suppose that at
the velocity range we are using the vehicle power
plant has a linear power/velocity performance (air
resistance is considered neligible if V < 50 mph)
then if V is increased in by 20% then E is increase by
20% and that if V is decreased by 33% then E is
decreased by 33%. Then  E(v') = 1.20 *E(v)  and E(v'")=0.66*E(v)

ET(hypermiling hill)   = .80* E(v') +1.20(v")
                              = .80* (1.20*E(v) + 1.20(.66*E(v))
                              =  .96 *E(v) + .79* E(v)
                              = 1.75 E(v)

The would result in a hypothetical results of
a 12% savings in energy over using a constant
velocity regardless of what type of power the
vehicle is using. Of course if there is a stop
sign or a red light at the bottom of the hill then
this technique does not work (9_9)


Posted by bwilson4web on July 6, 2011, 12:53 pm

NIce but there is another approach, assuming regular highway traffic
mix including large trucks. It is based upon keeping the engine in a
fuel efficient power range and how large trucks have power-to-weight
ratios that match Prius efficient power levels.

Find a large truck or semi-trailer, moving vans are excellent
candidates, and use them as the pacing vehicle. As they climb a hill,
follow them at the same speed at least 200 ft. behind. On steeper
grades, this will be in the 50-55 mph and tends to keep our Prius
engine in fuel-efficient power ranges. This is especially true for the
1.5L models that use fuel enrichment to avoid overheating and damaging
the catalytic converter. For example, this chart shows the fuel burn
up an 8% grade hill at different speeds:

It may be tempting to go up the hill very fast but this is a false
economy. The traction battery is providing the extra hill climb power
while the engine is struggling at maximum power, a fuel wasting power

When the truck crests the hill, follow and match speed on the descent
shifting between "N", "D" and "B" with preference to "B" and "N". The
reason is descending a hill in "D" and braking will put a significant
charge and heat load on the traction battery ... heat is the enemy of
traction batteries. This is especially true with the older NHW11,
2001-03 Prius. Maintain the 200 ft. and take pains to make sure the
truck driver can easily see that you are following at a safe distance.

If you see evidence that the truck driver is acting anxious, back off
and find another truck pacing vehicle. Always be friendly and smile
and 'thumbs up' to the truckers. Whatever you do, keep the distance at
least 200 ft., way back so you can react to road debris and won't be
seen as 'drafting' the trucker. After all the trucker is providing
'cover' for your fuel efficient driving.

Following traffic will see the truck as soon as they see you and
prepare to pass. If you are out there by yourself, following traffic
will often come right up behind your bumper before realizing you're
driving your speed, not theirs. So let the following traffic curse the
truck and snicker that you're stuck and move on down the road. You'll
pass them soon enough when they are refueling. <GRINS>

What if there isn't a 'big buddy' pacing vehicle?

Climb the hills at 55 mph and descend shifting between "B" and "N" at
a safe and reasonable speed for the conditions. Don't crowd the bumper
of slower traffic and generally be courteous.

Now if you have instrumentation that shows engine RPM, try to keep the
RPM under 2,400-2,600 (NHW11.) During a climb, try to keep it under
3,800 rpm (NHW11). Anything over 4,000 really burns the fuel for not
that much extra power so if traffic permits, bleed off speed. Here is
a chart of mixture and brake specific fuel consumption:

This is properly called "load driving" because you are keeping the
engine always in a fuel efficient region. Speed will vary but not
nearly as much as ballistics driving. Some of us have ideas for
adaptive cruise control that would handle the set points automatically
but that is something for another day.

Bob Wilson

Posted by Neo on August 2, 2011, 7:14 pm
The technique that I've describe is a variation of what is
known in Hypermiling terminology as NICE and DWL.
For the 2010 Prius (gen 3),  I've found that the Scangauge
II's Gallons per Hour (GPH) Xgauge to be a better indicator
of fuel efficiency than the Revolutions per Minute (RPM)
Xgauge. In non-highway environments, sustained burning
fuel at a rate higher than 1.50 GPH makes it much more
difficult to maintain  a FE of greater than 50 mpg. So
the underlying premise is to increase the fuel efficiency
by accelerating and  maintaining  the velocity while not
pushing the GPH meter over 1.20 GPH for any sustained
period of time. In a hilly driving terrain, the most fuel
efficient method of gaining  velocity to accelerate while
the car is travelling downhill until the desired speed is
achieve and to start driving up an uphill (from the bottom
of the hill) at a speed about 10 to 20 mph greater than
the targeted  top of the hill velocity while lessening the *load*
gradually (pressure on the accelerator) allowing the velocity
to incrementally decrease as the vehicle climbs
up the hill.

In non-winter and non-superhighway driving environments,
the most fuel efficienct load range (+50mpg) for the 2010
Prius (gen3) internal combustion engine (ICE)  appear
to be  when  the ICE is burning  from 0.80 to 1.20 gallons
per hour in short burst sessions while the car is in
drive mode and where the velocity is from 15 to 45 mph
where the optimum speed is about 35 mph (+60 mpg)
and the  optimum burst load is about 1.10 GPH.

In superhighway driving (+50 mph) when there was no
slow truck to pace behind,  I found that DWL-NICE
-while providing better FE- was much more difficult
to sustain than just setting the cruise control.
I just could not keep it up for more than an hour
on I270-I70 in Maryland.  By the time I was in
Pennsylvania I basically switched to setting
the cruise control between 50 mph to 60 mph
(where 50mph gets about +64 mpg and 60 mph
get about +58 mpg if the car is driven for more than
2 hours per driving session in a spring/summer/fall
driving environment ).   For superhighway driving,
GPH is not as important as acceleration, maximum
speed, and the length of the route (aka length of time
per trip).  For the best FE on the superhighway,
the maximum speed should be from 50 mph to
60 mph (with 50-55 mph giving slightly better
FE than 60mph), the car must be driven non stop on
the route/trip for over 30 minutes/30 miles (with
the longer non-stop driving session more desirable),
and car should gradually accelerate to highway
speed in the beginning and gradually decelerate
to nonhighway speed at the end of the trip.

Posted by bwilson4web on August 10, 2011, 12:32 pm
 . . .

Fortunately the ScanGauge can display four gauges and GPH is one of
three that work well in both the 1.5L NHW11 and the 1.8L ZVW30:

 o ICE coolant temperature
 o ICE rpm
 o Gallons per Hour

My GPH thresholds are similar:

 0.30 GPH - the car ICE under 70C while coasting in "N"
 0.60 GPH - during the first ~45 seconds of catalytic warm-up in 1.5L
NHW11 and ~180 seconds in the 1.8L ZVW30. I understand the 1.5L NHW20
has the longer, initial warm-up
 2.00 GPH - what I try to stay under when accelerating, traffic
permitting, after reaching 70C

During the initial, catalytic warm-up, both the NHW11 and ZVW30 Prius
will dip into the traction battery, heavily, if you keep the GPH at
~0.60-0.70. This window means I can often reach 35 mph on traction
battery in the NHW11 and 45 mph in the ZVW30 with very high fuel

What I do is park my commuting car near the exit at work where I can
see the cross-street traffic. Hours later when I'm ready to leave, I
monitor the traffic and when I can pull out without stopping, start
the car and smoothly exit the parking lot and accelerate to the
fastest speed while in the 0.60 GPH, EV mode. This ends at over 50 MPG
and when the ICE kicks in, shift into "N", traffic permitting, and
getting very nice mileage.

BTW, this early warm-up is described in the Toyota paper on the NHW11
but it takes instrumentation like the ScanGauge. If you have a cheaper
scanner, the mass airflow rate can substitute for the GPH.

Bob Wilson

This Thread
Bookmark this thread:
  • Subject
  • Author
  • Date
please rate this thread