The aim of control in a process plant is to try to ensure that plant operation meets production requirements while observing certain constraints of safety, environmental protection and limits of efficient operation.
3.1. Control requirements
In order to be able to evaluate how effective the control is, the control objectives need to be stated. The operational objectives of control as taken from (Luyben, Tyreus & Luyben, 1997) are as follows:
Establish control procedures.
Determine control degrees of freedom.
Establish energy management system.
Set production rate.
Control product quality and handle safety, operational and environmental constraints.
Control inventories (pressures and levels) and fix a flow in every recycle loop.
Check component balances.
Control individual unit operations.
Optimize economics or improve dynamic controllability.
3.2. Identification of control elements
Process output variables are normally streams that leave the process or conditions inside a process vessel. These are controlled variables and they are normally specified.
Process input variables are independent variables and they affect the output variables. They can be divided into two subgroups, namely manipulated variables and disturbance variables. The manipulated variables can be adjusted in order to control the output variable. The disturbance variables are determined by the external environment and upstream processes and can therefore not be manipulated or controlled. From the above discussion it is of utmost importance to be able to identify which variables can be controlled, manipulated and those which are disturbances.
(Newell and Lee, quoted by Seborg, Edgar & Mellichamp, 2004: 241-243), give guidelines for selection of controlled manipulated and measured variables (both input and output variables may be measured variables). These guidelines were used in the development of the control strategy.
For controlled variables:
All variables that are not self-regulating must be controlled.
Choose output variables that must be kept within equipment and operating constraints.
Select output variables that represent a direct measure of product quality or that strongly affect it.
Choose output variables that interact with other controlled variables.
Choose output variables that have favourable dynamic and static characteristics
For manipulated variable:
Select inputs that have large effects on controlled variables.
Choose inputs that rapidly affect the controlled variables.
The manipulated variables should affect the controlled variables directly rather than indirectly.
Avoid recycling of disturbances.
For measured variables:
Reliable, accurate measurements are essential for good control.
Select measurement points that have an adequate degree of sensitivity.
Select measurement points that minimize time delays and time constraints.
3.3 Control strategy
3.3.1 Main control variables
For this design, the most important control variables are in the furnace where the reforming reactions occur. This is due to the harsh operating conditions and heat that needs to be transferred. The major variables affecting the overall safety of the furnace are identified as follows:
Pressure control within the furnace is essential because it can compromise the
performance. Pressure drop within the reformer tubes has to be monitored and kept within the required range as a too high drop affects conversion. Backflow within pipelines has to be avoided as well.
The reforming reaction is overall endothermic and heat needs to be transferred to the reformer tubes for the reactions to occur. Temperature control within the reformer is essential from both a reaction and safety point of view. The reaction temperature should be kept in the range which allows reaction to proceed smoothly and the required conversion is achieved. The temperature should also not exceed the range that can be handled by the construction materials and the catalyst. The flue gas exit temperature also needs monitoring in order to comply with environmental guidelines.
Composition and Flow
The two important flows that need to be controlled are the steam/carbon ratios for the reactors and the fuel/air ratio in the reformer furnace. The steam/carbon ratio has to be monitored and kept above a certain value in order to drive the reactions and to prevent coking. The fuel /air ratio has to be kept at a certain value that ensures as high a combustion efficiency which in turn leads to a higher overall furnace efficiency.
3.3.2 Detailed strategy for unit operations
11.5 DETAILED CONTROL DESIGN FOR REFORMER FEED
11.5.1 Steam-to-Methane Ratio Control
One important control parameter in steam-methane reforming is the steam-to-methane
ratio. A low ratio is undesirable as it promotes the side reaction of coke formation on the catalyst,
which deactivates it and requires expensive replacement. Nonetheless, a high steam-to-methane
ratio will result in better conversion, but at the expense of elevated operating costs due to the
high cost associated with superheated steam. Hence, a compromise between methane conversion
and operating expense has to be made and in industries, this ratio is typically kept at 3:1.
To achieve this, a ratio control depicted in Fig 10-2 is employed to maintain the ratio
between steam and methane at 3:1 as stated in our design problem. The flow rate of the natural
gas stream is measured and transmitted by FT-101 to the ratio station FY-101. At the ratio
station, this signal is multiplied by an adjustable gain whose value is the desired ratio. The output
signal from the ratio station is then used as the set-point for flow controller FIC-102. This feed-
forward controller then adjusts the flow rate of the imported superheated steam by manipulating
the opening of the diaphragm valve FCV-102 using pneumatic signals.
In the preliminary design of this ratio control, it is assumed that molar flow rate is equal to the
volumetric flow rate which implies that possible pressure and temperature fluctuations in process
streams are not compensated.
3.3 Interaction analysis
Each controlled variable has been assigned a manipulate variable which will be used to control the controlled variable. Ideally, changes in a manipulated variable should affect only the controlled variable which it was paired with. Unfortunately, sometimes it happens that a manipulated variable affects more than one controlled variable. In order to make a choice with regards to pairing of the manipulated variable and controlled variable, the effect of the manipulated variable on the controlled variables must be determined.
The Relative-Gain Array (RGA) method can be used to determine the effect of inputs on outputs and can therefore assist in determining which controlled and manipulated variables to pair for control purposes (de Vaal, 2009).
The distillation column for which the control strategy was discussed above is a good example of a process unit with interaction and it will be used to explain how interaction analysis for the process was done. As an example, changes in the reflux rate for the column may affect the liquid level in the reflux drum, the column pressure and the top product composition. To make a choice as to which of the controlled variables to pair with the reflux rate, an RGA must be calculated.
After a choice has been made with regards to pairing of manipulated and controlled variables, the effect on other controlled variables can be eliminated or at least minimized by adding decouplers. A decoupler is a controller that is added as an addition to a conventional loop and is designed in such a way as to reduce interaction between control loops (Seborg et al, 2004:500).
3.4 Type of control
Unless otherwise stated, feedback control will be utilised on the plant. Feedback works on the principle that the controlled variable is measured and compared to a set point value. If there is a difference between the measured and the set point values, an error signal will be generated. This error is fed into a controller which takes action to remove the error by moving a final control element. The final control element is usually a control valve which manipulates the flow of a stream.
There are three common types of controllers and they are the proportional (P), proportional-integral (PI) and proportional-integral-derivative (PID) controllers. Each type of controller is superior to the others for certain types of applications. For instance, proportional controllers are mostly used to control liquid level in tanks (Luyben, 1990:228). PI controllers are used in most flow loops in industry as tight control is normally required (Luyben, 1990:231). Temperature control especially in reactors is controlled using PID controllers as they compensate for the lag in response (Luyben, 1990:231). The choice of controller types for pressure control loops depends on whether tight control is required or not. For this particular process the operating pressure is high and therefore tight control is not required. For this reason a PID controller can be used.
In some instances control strategies other than feedback are used. These include ratio control, cascade control and feedforward control. Ratio control involves keeping the ratio of flows constant and is normally used for feed flows into a reactor where stoichiometry is important. This is the case for the for the polymerization reactor where the feed ratio of VDF and CO2 must be maintained at the required point.
Cascade control is a variation of feedback control in that there are two feedback controllers. The primary controller changes the setpoint of the secondary loop. The output of the secondary loop goes to the final control element. The purpose of cascade control is to improve the dynamic performance of the control loop. (Luyben, 1990:255). An example of cascade control is the pyrolysis reactor temperature control loop. The reactor temperature controller is the primary controller and its output is the setpoint for the fuel gas flowrate controller.
4. Control Instrumentation
The instrumentation in process control includes sensors, analyzers and transmitters. Sensors interact directly with process variables such as temperature and pressure. Transmitters convert the measurement of the sensor into a signal which can be easily understood. Instrumentation for this process is described using the distillation column discussed in the control strategy section.
Sensors are devices that can measure individual process variables such as temperature, pressure and flow rate. Each type of variable has a different sensor and these are discussed below.
4.1.1 Temperature measurement
Thermocouples are the most commonly used temperature sensing devices. Iron-constantan thermocouples are commonly used over the 0 to 1300 ÂÂ° F temperature range (Luyben, 1990:209). Resistance thermometers are receiving increasing usage because they are about ten times more accurate than thermocouples (Perry & Green, 1997:8-45). However the cost of these resistance thermometers will exceed that of a typical thermometer. The location of a temperature sensor depends on the controlled variable and required accuracy of the instrument. Temperature is normally used to infer composition and the sensor is placed at the top or bottom of the column. For this process the bottoms purity is controlled and therefore the temperature sensor is located at the bottom of the column. The bottoms streams must be pure R152a and this implies that the accuracy of the sensor must be high with an allowable error of about 1-2 ÂÂ°C.
4.1.2 Pressure measurement
Bourden tubes, bellows and diaphragms are used to sense pressure and differential pressure (Luyben, 1997:209). The C-spring Bourden-tube pressure gauge will be used as it is available in a wide variety of pressure ranges and materials of construction (Perry & Green, 1997:8-47). Pressure sensors are located in the rectifying section near or at the top of the column. The accuracy of the sensor is not much of an issue as the aim is not to keep the pressure constant. Considering that the column is meant to operate at about 5 atm, errors of less than 0.1 atm are acceptable.
4.1.3 Flow measurement
According to Luyben, 1990:208, orifice plates are by far the most common type of flow-rate sensor. It operates by placing a fixed-area flow restriction in pipe that carries the fluid (Perry & Green, 1997:8-48). It is popular due to its simplicity and low cost. Turbine meters can also be utilised but according to Luyben, 1990:209, it is more expensive and it gives measurements that are more accurate.
Transmitters act as the interface between the process and the control system (Luyben, 1990:211). The function of a transmitter is to convert a sensor signal, in millivolts, pressure differential, into a control signal.
4.3 Control Valves
The purpose of a control valve is to regulate flow in any pipeline. In majority of chemical engineering processes the final control element is an automatic control valve which throttles the flow of a manipulated variable (Luyben, 1990:213).
4.3.1 Control Valve Choice
There are four basic types of control valves namely; globe valves, ball valves, disc valves and butterfly valves. Globe valves are the most commonly used. The globe valve presents some advantages over other valves in that valve trim is replaceable and another advantage is that it can operate at high pressure and can handle larger pressure drops than the other types of valves.
4.3.2 Valve action
Valves are designed either to fail in a wide-open position or completely shut. Appropriate action depends on the effect of the manipulated variable on the process (Luyben, 1990:213). In controlling a distillation column, there are utilities for condensing vapour in the condenser and for vaporising the liquid in a reboiler. The control valve for the steam into a reboiler would be designed in such a way as to fail in completely shut position while the opposite holds for cooling water to a condenser. In the case of streams going into and out of the column it is not as obvious to decide on the fail mechanism. For example, if the valve on the bottoms stream is to fail wide open, it puts the column at a risk of running dry while failing completely shut may lead to the column flooding. The best way to design the fail mechanism of these valves is in such a way that one compensates for the damage that may be caused by the other. For example if the bottoms valve fails closed, the vapour valve should fail open.
4.4 Materials of construction
Instrumentation just as any equipmentmust be constructed from materials that can withstand the environment in which it is operating. If a piece of instrumentation is located in a corrosive environment or a very hot environment, the construction material must be able to handle the conditions.
Sensors and control valves are the instrumentation that may be exposed to harsh environment. Wherever a sensor or control valve is located, it will be constructed of the same material as the equipment on which it is installed.
Carbon steel will be utilised for the parts of the process where there is little or no corrosion of the control valves and pipes. This includes the valves that deal with the control of the cooling water inlets to the reactors and column condensers.
Cleaning and maintenance is a part of the operation of every process plant. Control instrumentation must always be maintained in a working condition. During the design stage of process plant equipment, provision will be made for installation of control instrumentation. Maintenance will be carried out on a regular basis and the principle of prevention will be applied wherein potential problems with instrumentation are dealt with before they have an effect on the process.