Wind Turbine Design and Implementation Phase III
Download
Report
Transcript Wind Turbine Design and Implementation Phase III
Wind Turbine Design
and Implementation
Phase III
Senior Design May 2011-01 Team
Andrew Nigro (EE)
Chad Hand (EE)
Luke Rupiper (EE)
Ryan Semler (EE)
Shonda Butler (EE)
Advisor: Venkataramana Ajjarapu
Problem Statement
The objective of the senior design project is to simulate an environment in which
a wind resource provides renewable energy. The electricity generated from the
wind turbine can be used to power a direct load and potentially be integrated
into the Iowa State University power grid.
System Block Diagram
Overview of the Wind Power System
Operating Environment
•
•
•
•
Ideally would be mounted to the roof of a building
Controlled environment inside power lab in Coover
Wind simulation with AC motor controlled by LabVIEW
Input data provided by wind sensors
Requirements
Functional
•
•
•
•
•
•
•
•
The turbine will generate a 24V DC output
The turbine will generate a 400W peak output
The test-bed connection will serve to simulate the load
The motor will simulate outdoor wind speed
The sensors used to gather wind data will be an anemometer and wind vane
The RPM sensor will accurately reflect the speed of the motor within ±5%
The wind turbine will supply a load after charging the batteries to 23V
The user interface will display accurate measurements of DC voltage and
current, RPM, and real power produced
Non-Functional
• The project will comply with all state and federal electrical regulations
• The turbine will be remounted to a new stable operating platform
• The project will be documented through technical manual and in-depth
schematics
• Wiring and connections will be redone in a professional manner
Mounting & Platform
• Old platform issues:
• Stability
• Vibration
• Bowing
• Materials:
• 80/20 Inc. Rectangular Extruded Aluminum Rods
• ¾” slab of Medium Density Fiberwood
• L-Brackets and Hex Screws
• Aluminum Plate
Bowing
• Cost:
• $40-50 – Extruded aluminum
• $10 – Aluminum plate
• $25 – MDF platform
• $5 – Misc – Screws and Washers
• 20$/hour Installation – 5 Hours
Strapped
For Stability
CAD Model
Turbine Mount
• Construction
• Cut aluminum plating to cover faceplate of the turbine
• Cut extruded aluminum rods to match CAD measurements
• Bolt motor to aluminum rods
• Connect rods to faceplate to stabilize
• L bracket brace 90 degree connections
• Screw to MDF board
CAD Model
Wind Turbine & Inverter
• Southwest Windpower AirX 400
Turbine
• Designed to charge batteries
• Minimum of 7V DC input to power on
• Regulation Mode maximum 27V DC
input
• Wind input will be simulated through a
coupled 3 phase motor
• Outback GTFX2524 Inverter
•
•
•
•
24V DC input
120V AC output
Internal controller for stable output
Specifications are able to handle the
24V DC output from the turbine and
the 120V AC, 10A load
• Hybrid, stand-alone or grid connected
modes
Sensors: Anemometer & Wind Vane
• Data from wind sensors are collected from a transceiver that were provided by Senior
Design Team SD Dec10-05
•
•
•
•
•
Wind profile data collected from a self-healing mesh network on the roofs of Coover and Durham
Each node in the network uses an anemometer, a wind vane, a microcontroller, and a wireless transceiver
Real-time data is averaged every 10 seconds in order to avoid excessively ramping the motor for short
fluctuations
Sensors’ output transmitted through a transceiver with a range of 1-2 km
Transceiver output is passed through the microcontroller and then connected to our interface through USB
• Microcontroller support software exports sensor data to a text file
• Text file is imported into LabVIEW by analyzing each line as a string then passing this data to our
LabVIEW interface
Sensors : CT & Voltage
•
•
•
•
•
Hall effect current transducer
Powered by 12V DC power supply
Calculate current from wind turbine
Calculate current to inverter
Tested with multi-meter across resistor
• Voltage divider gives safe voltage to
DAQ
• Tested with multi-meter
Rpm Sensor
•
•
•
•
Hall Effect Sensor
Digital Output
Powered by DAQ
Tested by passing magnet causes a change
in voltage
• LabVIEW uses edge counter to calculate
RPM.
• Pull up resistor of 2.2KΩ gives ≈5V output
• Sensor gives ≈0V output when magnet is
near
User Interface
Motor Control
• LabVIEW interface takes in text file from wind sensors and uses it for
motor speed
• This was tested by running several different profiles through the
interface.
Full System Testing
LabVIEW Control of Motor
• Test that LabVIEW controls the motor at given wind speed
• Verify the RPM sensor matches with expected RPM
Full System Testing
Turbine Testing
• On/Off testing - Spin turbine by hand while shorting turbine
leads to test breaking ability.
• Battery Connection Testing - Connect appropriate leads to the
terminals and observe if the indicator light turns on.
• Rotation Response Testing - Spin the turbine with motor while
keeping the turbine under 500RPM indicator light should be
off.
• Rotation Response Testing - Spin the turbine with motor while
keeping the turbine over 500RPM indicator light should come
on which indicates power is being generated.
Full System Testing
Battery Testing
• Batteries measured to determine if they held the rated charge
of 14.5-14.9V each using a multi-meter.
• Series connection of batteries yields 29.0-29.8V maximum.
Full System Testing
• Inverter Testing
• 18V DC needed to operate – confirmed through multi-meter, LEDs, and
LabVIEW interface
• Jumper connects to pins to make operational
• LEDs show the status of the inverter
Inverter Modes Courtesy of Outback
Full System Testing
Load Testing
• Inverter being switched on will automatically switch on 1x75W
light bulb and power the additional outlet
• Light switch controls additional 75W light bulb
• Multiple devices will be plugged in to the outlet
• Circuit breaker stops power from being delivered to load if
30A current is run through load
Project Continuation
•
•
•
•
Direct real-time connection from wind sensors
Dedicated computer
Connection to grid
Mounting on roof of Coover
Project Members
• Luke Rupiper – Project Leader
• Motor Control lead
• Andrew Nigro –Equipment Procurement Task Leader
• Mounting and Construction lead
• Shonda Butler – Communications Task Leader
• Wind Sensor lead
• Ryan Semler – Website Task Leader
• RPM Sensor lead
• Chad Hand – Documentation Task Leader
• LabVIEW lead
Acknowledgements
• Advisor:
• Dr. Ajjarapu
• Client:
• Iowa State University Department of Electrical & Computer
Engineering
• Coover Hall Parts Department:
• Leland Harker
• Wind Sensor Design Team:
• SDDEC10-05
• Senior Design Group:
• SDMAY10-17
• National Instrument Forums