PPS Lab 5 - PID Controller tuning

Certification Assistance for Marine Engineers

Brought to you by www.dieselduck.net, comments to webmaster@dieselduck.net

Propulsion Plant Simulator, Level One
British Columbia Institute of Technology
Pacific Marine Training Campus
By: Martin Leduc Jan 22, 1999

Persons wanting to obtain the Canadian marine engineering license, or chief endorsement, need to complete the PPS L1 and L2 courses, in which they will have to carry out five labs. These are my results for one lab for your comparison. Findings are based on observation carried out and graphs not presented here.


(PID) Proportional, Integral, Derivative
Controller Tuning
Lab 5



This lab was performed to observe the reaction of a process controller. Controllers need to accurately adjust to input and react to various changes in the process, and must do so in a timely fashion.


Method of investigation

Procedure . The procedures set forth in the PPS Level 2 Course booklet were followed. The main ideas are as follows: In this case we are using a controller on the Heavy Fuel Oil purifier. It controls the purifier inlet temperature.

- In exercise 1, Altering the proportional ( gain ) from 13 to 5, and 19, to observe it reaction to the changes.
- In exercise 2, Altering the Integral control from 40 to 20, and 80. Exercise 3, altering the derivative from 0 to 20 sec, and 80 sec.



Equipment Used. The simulator used for this particular experiment was the NOR Control TEC 2000, simulating a Sulzer RLB. The RLB is a slow speed, six cylinder, two stroke diesel / heavy fuel engine. The engine drives a fixed pitch propeller directly couple to the engine. A shaft generator is also couple to the main drive shaft, but was not engaged at time of experiments. Various controllers are used throughout the simulated ship. The ship has three purifiers with these controllers. The HFO purifier was use for these exercises.

Data. Plotting of the controller respond to change was recorded using the "pen recorder" function. Results for exercice one can be found in Appendix A - B, D - F graph results from experiment two and three.



- In exercise 1A the gain, set @ 5, reacted sluggishly. In about five minutes, the desire value was reached, with a slight offset, as can be expected. When set at 19, in exercise 1B, the gain had trouble reacting to input value especially when an increase in temperature was input. The controller still had not settled after ten minutes of fluctuations.
- Exercise 2A. The controller, with Integral action set at 20 seconds, had trouble stabilising, even after 11 minutes. Once again the input for an increased in temperature created wild fluctuations in the controller. In 2B, the controller responded efficiently and within 4 minutes. An offset, more define in the temperature increase input, was prominent.
- In excercise 3A, with Derivative set at 20 from 0, the controller responded within three minutes, once again the controller, on the temperature increase, fluctuate dramatically. In exercise 3B, the response was very similar to 3A, but the offset was more dramatical. The desired value was never reached in either exercise, a definite offset was present.

The defaults settings of the controller are P @ 13, I @ 40, and D @ 0. This provides the quickest respond from the Integral action , the least overshoot from the Gain, and quickest correction of the offset by the Derivative action.

Brought to you by www.dieselduck.net comments to webmaster@dieselduck.net