TuhinPresentation_0729.ppt

A CONTROL STRATEGY FOR LOAD DISTRIBUTION IN

HYBRID FUEL CELL SYSTEMS

Dr. Tuhin Das Assistant Professor Rochester Institute of Technology Department of Mechanical Engineering Rochester, NY 14623

Fuel Cell System Overview

Modeling Methodology Hierarchical libraries

System Model

No. of cells = 100

Fuel flow = 0.0035 moles/s

Air flow = 0.035 moles/s

Recirculation = 70% A nominal operating point obtained through simulations Simulation and Model Validation

Reforming Reactions

Internal Reforming & Electro-chemistry

Anode reactions Cathode reaction Reaction (b) at equilibrium

CO to CO2 negligible Observations:

Mass Balance Equations

Coupled dynamics Inputs

Characterization Prediction of steady state conditions

Critical variables:

Utilization

Utilization Definition Coordinate transformation:

With coupled dynamics, utilization is independent of reaction rates

2. Steady-state utilization can be predicted exactly:

Steam-to-Carbon Balance Definition With coupled dynamics, STCB is independent of reaction rates

2. Steady-state STCB can be predicted exactly:

Coordinate transformation:

Fuel Optimization Problem statement:

Given the … Current demand:

2. What will be the minimum fuel operating condition?

Can we determine …

Optimum Fuel Operating Conditions Existence:

Minimum Fuel:

Transient Performance Considerations Fuel control based on steady-state analysis:

Considerations:

How well does the control respond to load transients in presence of fuel dynamics?

How much deviation is observed from target Utilization during transients and for what duration?

How much deviation is observed in STCR/STCB during transients and for what duration?

Fuel supply may be rate-limited or display a specific dynamic behavior (say 1st order dynamics) Fuel supply cannot respond instantaneously to sudden load changes.

Effect of Load Transients Fuel flow follows first order dynamics with time constant of 2s

0 100 200 300 0 100 200 300 140 160 180 200 time (s) time (s) time (s)

0 100 200 300 0 100 200 300 0 100 200 300

C el l c ur re nt

(A m ps

0.04

0.05

0.06

0.07

Fu fl ow

(m ol es /s

U til iz at io n

Effect of Load Transients Fuel flow follows first order dynamics with time constant of 2s

0 100 200 300

1.6

1.8

2.2

140 160 180 200 1.9

2.1

ST

R

Proposed Control Strategy Target fuel based on steady-state analysis:

FC load is based on measured fuel:

The FC load current is shaped dynamically due to feedback

Target fuel is based on the net current requirement

As the measured fuel flow approaches the target, the FC load current approaches the net current requirement

Open-Loop Results

Open-Loop Results

1.94

1.96

140 160 180 200

2.1

Hybrid Fuel Cell System Fuel Cell current draw:

Fuel flow based on required total power:

Current draw from battery/ultra-cap:

FC system

Li-Ion Battery

DC/DC

Conv

DC/DC

Conv

Fuel Cell System

Supplementary Source Batt/Ultra-cap

Control

Load

Preliminary Closed-Loop Results

150 200 250 300 350

0.98

0.984

0.988

0.992

79.5

80.5

Net power demand

FC power

Po w er (k

W ) FC current demand FC current draw

FC

c

Po w er (k

W

Battery power

SO

B te ry c

Preliminary Closed-Loop Results

80.1

80.2 time (s) time (s)

With FC current shaping

Without FC current shaping

Acknowledgement I would like to thank ONR for supporting this research.

I am grateful to Mr. Donald Hoffman, Mr. John Kuseian, Mr. John Heinzel, and Mr. Steven Miller of ONR for their valuable comments and suggestions.