Chemical Processing - April 2008 - (Page 46)

>> PROCESS PUZZLER This line could operate during start-up. I would recycle the flow until the reactor temperature is achieved. Amit Neogi, project engineer Fluor, New Delhi, India Maintain the temperature The simplest solution would be to heat and maintain the liquid at the desired temperature. The liquid can circulate to the raw material tank. When it is time to charge the reactor, use the existing system, which can be simplified by taking out some of the controls that now would become redundant. Maintaining temperature should allow automatic controls. Girish Malhotra, president EPCOT International, Pepper Pike, Ohio Ramp it up First, I would look at the proportional, integral and differential (PID) tuning of your temperature controller. It sounds like the controller is not fast enough on the throttle when the temperature exceeds the set point. Try increasing the gain of the controller (P) and backing off on the reset action (I). It might also be that your heat input may be too much during start-up. You might want to program an output clamp on the temperature controller during the start-up period to limit the heat input (valve opening) and avoid rapid temperature increase. Also try using a set-point ramp on the temperature controller, which will limit the heat input during start-up. If these don’t work, and if you have a batch sequence controller, you can program your controller to mimic what works for the operators during a startup. The program can set the controller in manual with a set output for a period of time until the start-up phase is over and then automatically switch to auto mode. Mike Dobrowolsky, utilities leader SABIC Innovative Plastics, Selkirk, N.Y. Measure fluid condition In the old days, steam was such a bitch to control. Adding to that a shear-sensitive, possibly heat-sensitive fluid and you have a complicated control problem. I have a few ideas worth considering. Ammeters are nice to measure work being done on changeable viscosity materials, if one can find suitable space to insert some device that reflects viscosity — gear pumps come to mind as long as shearing isn’t a problem. If viscosity can be inferred, then condition of the fluid could be measured before it becomes an issue in the reactor. The heat exchanger could be seeing fouling. A fluorocarbon coating may be needed to limit adhesion to the 46 • April 2008 tubes or plates. Here’s another thought: a static mixer between the heat exchanger and temperature probe. Tom Murphy, CEO Puritrol, Inc., Centerville, Mass. Review the procedures I am assuming that the process is a single pass through the heat exchanger and on to the reactor. Also you did not mention what the alarms indicate. With that said, I would suggest the following. Take a look at your start-up procedures: 1. Can you bring the heat exchanger up to temperature prior to introducing the product? 2. Can you run the feed pump slow during start-up, allowing temperature equilibrium? 3. Tie the pump speed to the outlet temperature — with more heat exchanger dwell time. 4. Can you run a recycle loop when the discharge temperature is too low? Scott E. Crosby, president SKL Process Solutions, Grass Lake, Mich. Recirculate I would install thermostatic valves in a bypass line after the heat exchanger. Connect the bypass line to the pump suction line. Circulate the fluid through the pump until the liquid, the piping and the heat exchanger are heated. C. M. Pakhale, superintendent Oil India, Duliajan, India Improve shutdown I don’t have any comments on the start-up. For the shutdown, incorporate a time delay in the control loop. The delay can be set depending on the process sensitivity. If the heat exchanger has spare heating capacity, consider recycling part of the stream from the outlet of the heat exchanger to the storage tank — I assume that a storage tank supplies the liquid to the heat exchanger. The increased turbulence from the higher velocity will reduce the response time of the resistance temperature detector (RTD) or thermocouple. Typically, the response of an RTD is somewhat slower than a thermocouple. You might consider replacing an RTD with a thermocouple as part of your solution. From the problem statement it is not very clear as to when these shutdowns occur. If they occur randomly, the problem may be with the installation of the sensor and the wiring. The signal to the control system could be distorted by noise. Some of the items to check may be rounding, electromagnetic noise and tightness of connections. If the pump has adequate head, consider swaging down the pipe where the temperature sensor is installed. This www.chemicalprocessing.com http://www.chemicalprocessing.com

Table of Contents Feed for the Digital Edition of Chemical Processing - April 2008

Chemical Processing - April 2008 Contents From the Editor ChemicalProcessing.com Field Notes In Process Energy Saver Compliance Advisor Is It a Tragedy or Comedy for Engineers? Better Understanding Boosts Mixer Scale-up Don't Err With Air Compressors Control Performance Supervision Enhances Revamp Process Puzzler Plant InSites Equipment & Services Product Spotlight/Classifieds Ad Index End Point

Chemical Processing - April 2008

For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.