Adjustable Governor for Synchronous Generator - SEL

Problem Statement
The team is designing, simulating, and implementing closed-loop frequency control on a small power generator located in the University of Idaho Advanced Power Lab. The control-loop governs the existing system such that it dynamically responds as if it were a different, larger generator powered by hydroelectric, steam, or gas.

Specifications
Hard fast specifications for this project have not been communicated, which is normal for senior design. Instead, we have been given boundary performance metrics and are asked to develop minimum specifications based upon industry standards.

The first performance metric to discuss is Settling Time (ST). ST will be defined (at least for this paper) as the time between a deviation of output frequency from the system setpoint and the moment when the system returns to the setpoint with no significant over or undershoot. Thus, the current system performance will be used to establish existing performance for our project. Dr. Johnson has further clarified that many power system governors have ST values around 5 seconds, and this is considered good performance in the industry.

The second performance metric to discuss is precision. Dr. Johnson suggested we consider how tightly we will control frequency. For instance, will the output frequency be controlled to 60 with an error of 1 hz, or an error of .1 hz, or .01 hz. The key here is to what precision with the output frequency be held to its set point.

Objective
A power system is used to develop power at a specific voltage and frequency. Load changes on the output of a power­generating system will cause fluctuations in both voltage and frequency. The fluctuations are undesirable and must be prevented or at least mitigated. The power system can be monitored in terms of its electrical frequency or its mechanical frequency.

The power system that has provided to the University of Idaho will ultimately be used for an Advanced Power Lab. SEL has requested that the team work at cleaning up the lab and making it a professional, clean atmosphere that is safe and conducive to learning.

The specific power system for this project consists of a VFD driving an induction motor. The induction motor is coupled to a synchronous generator. At the output of the synchronous generator is a shaft­position digital­encoder. (There is also a DC source for supplying the generator field, but that is part of another project.) Currently, the VFD controls the induction motor as a governor. The fact that electrical power, Real(3*EA*IA) is equal to mechanical power T*(W ­angle a) is used by the VFD. Since it knows the voltage applied (E) and the current drawn (IA) it can infer what the torque (T) and radial frequency (W) are. The inference that the VFD must do can be improved by experimentally determining exact values for system characteristics, like the inertia of the system or its intrinsic time constant. The IEEE publishes models that include critical characteristics of power systems that allow governors to act with greater response and accuracy. A major deliverable of the team is to experimentally improve the values for these characteristics and update the VFD parameters with these values. The team intends to write experimental procedures for each characteristic requiring testing and providing these test procedures as part of our final deliverable.

A second method for controlling the power system is using an external system that dynamically adjusts feedback from the output encoder of the generator to the input of the VFD. The method of feedback control requires an external system to monitor the mechanical or electrical frequency of the output and comparing that against a set point. Depending upon the magnitude of the error, a compensating feedback signal must be applied to the VFD to drive the system back to the power as quickly as possible. The VFD provides inputs for implementing this control scheme as well.

MODBUS & EFB


The code to construct and send simple MODBUS data packets has been written and is currently undergoing tests on the hardware and VFD. With the code, a user has the ability to specify three different field-attributes that make up a MODBUS frame; node address, function code, and an array of data bytes to be sent. The code will automatically calculate the frame’s 16-bit CRC and append it to the end of a frame transmission. In its current form, the MODBUS implementation results in the VFD displaying an increasing number of UART errors each time a packet is sent. As a demonstration, I have attempted to form and send packets in attempt to cycle the VFD through it’s start-up procedures, accelerate to a reference, and then coast to a stop. The search for why UART errors exist is an ongoing effort and all layers of implementation are being scrutinized (from physical to data-link). An interface logic-analyzer (by IFTools) loaned to me from an SEL-coworker is being used to investigate.

RTDS Simulations
RTDS, REAL TIME DIGITAL SIMULATOR simulates the control-loop governs the existing system such that it responds as if it were a different, larger industrial power system like hydroelectric, steam, or gas. 2015 Byte Back RTDS Simulation



SEL 411L
We are using SEL-411L Relay Protection and Automation System for detecting the change in frequency in the system while applying load. SEL 411L allows us to capture any change happing over period of time for further analysis.



Document Archive

 * Team Meeting Minutes


 * [[File:091115 Byte Back minutes.pdf]]
 * [[File:091815 Byte Back minutes.pdf]]
 * [[File:092515 Byte Back minutes.pdf]]
 * [[File:100215 Byte Back minutes.pdf]]
 * [[File:100915 Byte Back minutes.pdf]]
 * [[File:101615 Byte Back minutes.pdf]]
 * [[File:102315 Byte Back minutes.pdf]]
 * [[File:2015_Byte_Back_minutes_103015.pdf]]
 * [[File:2015_Byte_Back_minutes_110615.pdf]]
 * [[File:2015_Byte_Back_minutes_111915.pdf]]
 * [[File:2015_Byte_Back_minutes_120315.pdf]]