Extending the Lifetime of Fuel Cell Based Hybrid Systems

1
Extending the Lifetime of Fuel Cell Based Hybrid
Systems

• Jianli Zhuo
• Dept. of Electrical Engineering, Arizona State
  University
• jianli_at_asu.edu

Chaitali Chakrabarti (Dept. of EE, Arizona State
University) Naehyuck Chang (School of CSE, Seoul
National University, Korea) Sarma Vrudhula
(Dept. of CSE, Arizona State University)
43rd DAC, San Francisco, Jul 26, 2006
2
More Functions ? More Energy
1st cell phone on market Motorola DynaTAC Price
3k, 1983
1st cell phone, by Martin Cooper,
Motorola (www.fcc.gov)
Motorola V60, 2000
Motorola razr, 2005
Motorola Moto-Q, 2006
Source www.motorola.com
3
Fuel Cell - High Energy Density Source

• Fuel cell
• an electrochemical energy conversion device
• uses external supply of fuel and oxygen
• electrodes are catalytic and relatively stable
• Advantages of Fuel cell
• High energy density ? longer lifetime for same
  weight/size
• Instant recharge
• No performance degradation during discharge
• Clean, zero emission

4
Fuel Cell Stack
Fuel cell stack anode, cathode, electrolyte
At the anode 2H2 ? 4H4e-
H ions pass through the electrolyte (membrane)
e- electrons generate electricity
At the cathode, 4H4e-O2 ? 2H2O
5
Fuel Cell System

• Fuel cell system
• fuel processor generates hydrogen
• oxygen is from the air
• power conditioning component.
• temperature/humidity/air control

6
Fuel Cell - Hybrid Automobiles
Mercedes-Benz Citaro fuel cell bus (Source
DaimlerChrysler AG)
Honda (www.edmunds.com)
Toyota (www.edmunds.com)
Chrysler (auto.consumerguide.com)
7
Fuel Cell - Portable Applications
They have proven the functionality, but have not
been optimized
Ballard power system (www.ichet.org)
Fujitsu (pr.fujitsu.com)
Toshiba (www.engadget.com)
Toshiba, KDDI, and Hitachi (www.ubergizmo.com)
8
Work in Fuel Cell

• Membrane and fuel cell stack
• Hydrogen generation
• Control of hybrid power source
• most previous work is in the context of hybrid
  automobiles.
• hybrid cars are different from embedded systems
  because their power demand are usually high
  depending on the user, and not easily
  controllable.
• No prior work on system-level optimization of
  fuel cell powered embedded systems

9
Hybrid System Optimization

• Maximize the lifetime of the fuel cell
• Fuel cell lifetime is determined by the fuel
  consumption
• Fuel consumption rate is proportional to the
  current value
• Total fuel consumption is proportional to the
  total charge consumption
• Maximizing the lifetime ? minimizing the charge
  consumption

10
Our Prototype
SIGDA / UBooth Station 4, Demo 22 MTH, 1000AM
1100AM
11
Outline

• Fuel cell characteristics and hybrid system
• Fuel cell efficient scaling algorithm
• Power model
• Motivational example
• Optimization framework
• Algorithm
• Simulation results
• Conclusion

12
Fuel Cell Characteristics

Figure 1. Polarization I-V-P curves for room
temperature fuel cell

• Current , voltage , the power first
  and then .
• The operating point is set at the power 2/3 max
  power
• Bad load following capability, small range, slow
  speed
• We consider a fixed output of the fuel cell (no
  load following allowed)

13
Why Constant Fuel Cell Output

• Allowing load current variation requires an
  expensive control scheme
• Dead-end anode and purge operation requires a
  sophisticated control scheme
• Constant fuel cell current setup is suitable for
  portable, low-cost fuel cell systems
• Instead of fuel flow control or open-end anode,
  bleeder current makes the load current constant
• A proper charge management scheme can minimize
  the bleeder current while keeping the fuel cell
  current constant

14
Advantage of Hybrid Power Source

• Fuel cell/battery hybrid power source
• Fuel cell provides high energy density
• Battery provides high power density (peak power
  value) and fast load matching

15
Fuel Cell - Battery Hybrid System
16
System Power Model and Definitions
17
Motivational Example
IF 500mA, Bmax 2000 mA-min,
Bkini1000mA-min, Task Tk has execution time
10min. ik(1)500mA DVS scales from 1 to 2.5 with
steps of 0.1,
18
Charge Optimization

• Objective
• Minimize the total charge consumption of the
  hybrid system
• Subject to
• Configuration of the hybrid power source
• fuel cell output and the battery capacity
• Task specification of the embedded system
• deadline requirements

19
Optimization Single Task
20
Optimization Multiple Tasks
21
Algorithm

• According to
  when battery is ideal
• We develop a 3 step algorithm
• Step1
• distribute the slack evenly among all tasks
• Repeat step 2 and step 3 for each task
• Step2
• based on result of step 1, calculate skopt by
  considering charge minimization, deadline
  constraint and battery constraint.
• Step3
• run task Tk, then calculate the final state of
  the battery, which is the initial battery state
  of the next task.

22
Simulation Setup
Compare the two algorithms fc_scale the
proposed fuel cell efficient scaling
algorithm en_scale it minimizes the energy of
the embedded system
Embedded system scaling factor from 1 to 2.5
with steps 0.1 ik(1) 700 mA, Task
specification all tasks share a same deadline
D number of tasks n 50100 for each
simulation execution time of each task is
randomly chosen 12 min task density
is varied from 0.3 to 0.9 Battery B1ini
0.5Bmax
23
Experiment 1
Power source setting IF 800mA, Bmax
250mA-hr (load current is 700mA on highest
frequency)
24
Experiment 2 Effect of IF
In experiment 1, IF 800mA gt load current
700mA We reduce fuel cell current to 600mA and
compare the results
A-hr
Qtask
Qwaste
2
1.5
1
0.5
0
600mA fuel cell is more efficient if we combine
it with this battery (Bmax 250mA-hr)
25
Experiment 3 IF Bmax
Now we vary both fuel cell current and battery
capacity IF 250mA 700mA with steps of
50mA Bmax 100mA-min to 204,800 mA-min (
3500mA-hr)
26
Conclusion

• Proposed a fuel cell-battery hybrid power system
  for embedded applications
• Developed a charge based optimization framework
• Developed a task scaling algorithm that minimizes
  the total charge consumption and thus maximizes
  the lifetime of the fuel cell based system.
• Thanks to
• Dr. Don Gervasio (Flexible Display Center, ASU)
• Sonja Tasic (Flexible Display Center, ASU)
• Kyungsoo Lee (School of CSE, SNU)

27
Thank You!
SIGDA/UBooth Station 4, Demo 22 MTH 1000AM
1100AM
28
Fuel Cell-battery Hybrid System
29
Effect of Each Parameters

• Effect of fuel cell current
• IF too small, constraint violation
• IF too large, waste charge-gtlow cost-efficiency
• Effect of battery capacity
• Bmax too small, constraint violation or waste
  charge
• Bmax too large, waste of size and weight
• Effect of power ratio
• which minimizes Qtask

30
Experiment 1
Simulation results
Power source setting IF 800mA, Bmax 250mA-hr
31
Experiment 4 Choose IF Bmax
Based on experiment 3, we can choose combination
of the fuel cell and the battery according to the
system specification
Assumed VF VB 1.5V in the calculation
32
Basics of Fuel Cell
How fuel cell stack works
Fuel cell system
32/24
33
Fuel Cell Power Generations
1 MW fuel cell power plants, WA, 2004
(dnr.metrokc.gov)
Toshiba 200 KW power plant, 1997 (www.toshiba.co.j
p)