P08451 SD1 Final Presentation2 - EDGE

P08451: Feasibility of Thermoelectric Waste Heat Recovery From Large Scale Systems

Sam Haas Project ManagerMark Livelli Thermal AnalysisJacob LaManna System LevelChinedu Chikwem Structural SystemsDavid Ortiz Electrical SystemsBrittany Ray Feasibility Analysis

Project Statement

• To understand how thermoelectrics can be used in power generation and how they operate in a large scale system.

• To gain insight into the technical and economic viability of using thermoelectrics for waste heat recovery in industrial settings.

Introduction to Thermo-Electrics

Solid state devices

Two modes of operation1) Current � temperature gradient2) Power Generation � temperature gradient to electrical

energy

Historically materials in TE modules are off the shelf and have been around since the 1960’s. Advancement in nano-scaled materials in the last 5 years has brought attention to power generation applications.

High Level Needs

• Develop a system model that can relate P07441 Auto Exhaust Test Bed and the Dresser Rand VECTRA Gas Turbine.

• Use the model to design, build and test a small scale prototype unit.

• Refine the model using the prototype.

• Use theoretical data to conduct feasibility study for Dresser Rand Vectra Gas Turbine

Summary of Past Design Reviews

• Concept / System Level Review– Rescope of project focus to include a more modular prototype– Redefine specification reference values

• Detailed Design Review– Need for mitigation of thermal expansion risk on modules

– Need for detailed drawing of electrical control interface– Uncertainty in feasibility analysis

Overall Project Flow

Model Development

Model Verification

Prototype Development

Prototype Testing

Feasibility Study

Model Refinement

Key:

-1st stage

-2nd stage

Thermal Resistive Network for Model

Convection Resistance

Fin-to-Duct Contact Resistance

Duct Conduction Resistance

Contact Resistance

Insulation Resistance

Thermoelectric Resistance

Contact Resistance

Cold Side Overall Resistance

Duct Cross-Sectional Thermal Model

Thermo Electric Module Thermo Electric Module

Thermo Electric Module Thermo Electric Module

Heat Flow From Exhaust (qh) Heat Flow From Exhaust (qh)

Heat Flow From Exhaust (qh) Heat Flow From Exhaust (qh)

Thermo Electric Cold Side Temperature (T1)

Thermo Electric Hot Side Temperature (T2)

Heat Flow Into Coolant (qc)

Heat Flow Into Coolant (qc)Heat Flow Into Coolant (qc)

Heat Flow Into Coolant (qc)

TE Cross-Sectional Thermal Model

Thermo Electric Cold Side Temperature (T1)Thermo Electric Hot Side Temperature (T2)

Exhaust Inlet (Thi)Exhaust Outlet (Tho)

Cooling Fluid Outlet (Tco) Cooling Fluid Inlet (Tci)

Thermo Electric Module

Insulation

Heat Flow From Exhaust (qh)

Insulation

Heat Flow Into Module (qh1)

Heat Flow Out of Module (qc1)

Average Exhaust Temperature (Th)

Average Coolant Temperature (Tc)

Heat Flow Into Insulation (qh2)

Heat Flow Out of Insulation (qc2)

Numerical Model

• Required input parameters– Duct and fin geometry– Exhaust and cooling parameters

– Thermo electric properties

• Preliminary calculations– Geometrical

– Thermal

– Resistance

• Numerical simulation

Plotted Results

Current Automotive Test Stand

Prototype Thermoelectric

Heat Exchanger

Key: • Heat Exchanger Integration Points

Prototype Design

Duct Heat Exchanger

•Designed to be a versatile test stand that can relatively quickly and easily be changed from one configuration to another.

•Heat Sinks can be changed or removed completely by removing the end tanks and separating the two halves of the duct

•Different numbers and types of TEGs may be used in different series and parallel configurations

Overall Assembly

Other Design Considerations

•SS Cooling Cold Plates Available in 8 and 12 inch Lengths, but the 12 inch allows testing of more modules in series while remainingwithin budget

•To ensure that the pressure on the modules does not increase to a point of failure due to thermal expansion Belleville disc springs will be used with the bolts

•PTFE gaskets will be used to seal all flange faces

•Insulation will be used on duct around modules

Model Verification

• Multiple Prototype Configurations– Interchangeable fins on hot-side of duct– Interchangeable TEG modules

• Multiple Heat Flow Conditions– Changing Flow Rate– Changing Flow temperature

• Data Collection– Parameter Sensors– Electrical Sensors– Digital Data Acquisition and Processing

Prototype Verification

• Mechanical Robustness– Stress analysis of materials– Thermal analysis of materials

• Analysis of Operation with Respect to Specifications– Thermal sensors– Pressure sensors– Flow sensors– Power calculations

Connection Diagram

Electrical Test Platform Schematic

• Power Calculated for 5 separate zones

• Voltage measurement in LABVIEW

• High Power Rheostats used to match load at Steady State

Feasibility Study

• Economic

– Predictions of future technology

– Linking it to the mathematical model and validating assumptions with testing of prototype

– The prototype will hopefully validate cost assumptions as it relates to TE modules

• Actual Vs. Predicted– Power Generation– System Efficiency– Needed Delta T to produce enough power to benefit existing system

Current Cost Evaluation

• Assumptions include:– 10000 TE modules– That there will be no maintenance cost for the thermoelectric modules.– The current conversion efficiency is 5%.– There will be no operating cost for the thermoelectric unit– Rated Capacity is 75%– TEG intercepts 50% of waste heat

Approaches

• Using Simulink– Power Output vs. Temperature Conditions

• Looks at the required Delta T to determine maximum energy output

– System Power vs. Efficiency Tradeoff• Understand behavior of trade off for large scale model• Shows what compromise in power might change device

efficiency/energy output– Power Output at different cost per Watt

• Vary $/W ($1/w, $2/w,$3/w)• Power output vs. Number of TE’s

Other considerations

• Examine how pressure drop causes reduction in turbine power– Dresser Rand is able to provide power vs. pressure drop

relationship curves for the VECTRA• Ideas for Thermoelectric System

– Device that measures and records exhaust gas temperature versus time

– Dresser Rand indicated they could use this information as input to low cycle fatigue analyses

• Determine how much power is needed to produce this device and use number as target number of KW to produce

Budget

Issues / Risks

Verification from experimental dataIsothermal surface assumptions in the model

Obtaining proper input data from sponsor for VECTRA turbine

Scaling issues for the analytical model

Use Belleville springs to minimize stresses on modules caused by thermal expansion

Thermal expansion effect on thermoelectric modules

Contacting other group ahead of time to minimize conflict

Combined use of facilities with other groups

Aggressive schedule, pushing for early completion of fabrication

Completing high level project needs

Place order before break or as soon as possible

Delivery of long lead items

MitigationIssue

SDII Schedule• 3/10/08 (week 1-2)

– Finalize all designs and drawings (early week 1)– Begin fabrication of prototype (with received parts)– Begin development of detailed test procedure– Begin creation of feasibility model (with unverified data)

• 3/24/08 (week 3-5)– Finish fabrication of prototype (by end of week 4)– Finalize test procedure– Begin test setup / creation of data acquisition program– Continue feasibility model

• 4/14/08 (week 6-8)– Testing of prototype / verification of model– Use newly acquired data in feasibility model to correlate VECTRA scaling

• 5/5/08 (week 9)– Finalize verification of model, feasibility study

• 5/12/08 (week 10)– Design Presentation, Technical Paper, Poster Completed– Documentation completed