conventional power is more efficient.
Figure 4 represents a revised version
of the previously developed theoretical
crossover map used in Part I. It is still
optimistic in practice, representing
what we think are achievable targets—
not what has already been achieved.
In Figure 3 I have plotted fuel
consumption versus boat speed for
our test boat, a 48-foot, 18-ton sailboat
with an optimized conventional
propulsion system. This system was
made as efficient as possible to raise the
bar as high as possible for the hybrid
systems. Figure 3 also shows two other
curves: one for a serial hybrid system in
diesel-electric mode and one for either
a serial or a parallel system in battery-driven mode using TPPL batteries. The
crossover point for diesel-electric mode
occurs at 5. 8 knots. In battery-driven
mode, the crossover speed drops to
4. 55 knots. Figure 4 is an expanded
version of the lower end of Figure 3.
When looked at in percentage
terms, below 3. 5 knots, both diesel-electric and battery-driven hybrid
systems achieve massive gains in fuel
efficiency compared with the optimized conventional system.
If we extrapolate down to 3 knots (i.e., harbor maneuvering
speeds), hybrid fuel consumption is less than half that of
the conventional system, and there will be additional gains
through the elimination of all dockside idling.
Above the crossover speed, in percentage terms, the increase
in fuel consumption of the hybrid system is nowhere near
as great as the percentage gains below the crossover speed.
However, in absolute terms (liters per hour), the hybrid losses
rapidly dwarf any gains below the crossover speed. Parallel
systems do not care, because when the vessel is operating
above the crossover speed, it will be under engine power,
whereas serial systems pay the full penalty.
This is the Achilles heel of a serial system. If vessel
operation requires sustained running above the crossover
speed—and often, cruising speed will be above the crossover
speed—the losses will rapidly overwhelm any gains below the
crossover speed, for a net loss of efficiency.
To avoid this serial-system efficiency penalty, you have to
raise the efficiencies in the serial system to a level that shifts
the crossover speed in diesel-electric mode to cruising speed
or higher. Then, the serial system only pays an efficiency
penalty on those occasions when it is operated above cruising
speed. I have seen a number of efficiency claims for larger
serial systems that fit this scenario, but I have not seen data to
determine whether this has been achieved in practice.
The parallel system must pay the battery loss penalty,
which lowers the crossover speed, but as long as electric
operation is used only below the crossover speed, the parallel
system will always show a net gain in efficiency. This may be
quite substantial in percentage terms, although, as we have
seen, it will typically be modest in absolute terms.
There is something important missing from this analysis.
Up to this point, I have presumed that all propulsion energy
comes from an engine on the boat—either the main engine
in the conventional installation, and at higher loads in the
parallel system, or a generator supplying the energy for
electric propulsion. This is necessarily true in the conventional
system, but not necessarily true for the hybrid. At the core of
any hybrid system (serial or parallel) is an electric bus. Energy
for propulsion (and house) loads can come from any electrical
source. In practical terms, on boats this means shorepower,
solar, wind and, on a sailboat when under sail, regeneration
Our experiments required sophisticated data monitoring,
including a custom-built datalogger, which is here being
programmed and tested by people from TML, its developer.
Figure 2 Cross-Over speeds revised
A revised version of the theoretical crossover map from Part I, with lower
estimated crossover speeds. The crossover speed is the point at which a
conventional diesel propulsion becomes more efficient than a hybrid system.
SFC or specific fuel consumption is measured in grams per kilowatt-hour. The
higher the SFC, the lower the efficency.