1 INTERACTION OF PROCESS DESIGN AND CONTROL Ref: Seider, Seader and Lewin (2004), Chapter 20

2 PART ONE: PART ONE: CLASSIFICATION OF VARIABLES, DOF ANALYSIS & UNIT-BY-UNIT CONTROL Ref: Seider, Seader and Lewin (2004), Chapter 20

3 The design of a control system for a chemical plant is guided by the objective to maximize profits by transforming raw materials into useful products while satisfying: –Product specifications: quality, rate. –Safety –Operational constraints –Environmental regulations - on air and water quality as well as waste disposal. PROCESS OBJECTIVES

4  Variables that effect and are affected by the process should be categorized as either control (manipulated) variables, disturbances and outputs. Process Outputs Manipulated variables Disturbances  It is usually not possible to control all outputs (why?)  Thus, once the number of manipulated variables are defined, one selects which of the outputs should be controlled variables. CLASSIFICATION OF VARIABLES

5  Rule 1: Select variables that are not self-regulating.  Rule 2: Select output variables that would exceed the equipment and operating constraints without control.  Rule 3: Select output variables that are a direct measure of the product quality or that strongly affect it.  Rule 4: Choose output variables that seriously interact with other controlled variables.  Rule 5: Choose output variables that have favorable static and dynamic responses to the available control variables. SELECTION OF CONTROLLED VARIABLES

6  Rule 6: Select inputs that significantly affect the controlled variables.  Rule 7: Select inputs that rapidly affect the controlled variables.  Rule 8: The manipulated variables should affect the controlled variables directly rather than indirectly.  Rule 9: Avoid recycling disturbances. SELECTION OF MANIPULATED VARIABLES

7  Rule 10: Reliable, accurate measurements are essential for good control.  Rule 11: Select measurement points that are sufficiently sensitive.  Rule 12: Select measurement points that minimize time delays and time constants. SELECTION OF MEASURED VARIABLES

8  Before selecting the controlled and manipulated variables for a control system, one must determine the number of variables permissible. The number of manipulated variables cannot exceed the degrees of freedom, which are determined using a process model according to: N D = N Variables - N Equations N D = N manipulated + N Externally Defined Degrees of freedom Number of variables Number of equations N Manipulated = N Variables - N externally defined - N Equations DEGREES OF FREEDOM ANALYSIS

9  Number of variables. N variables = Externally defined (disturbances) : C Ai, T i, and T CO 10 EXAMPLE 1: CONTROL OF CSTR

10  Material and energy balances: N Equations = 4 EXAMPLE 1: CONTROL OF CSTR (Cont’d)

11 N Manipulated = N Variables - N ext. defined - N equations = = 3 EXAMPLE 1: CONTROL OF CSTR (Cont’d)

12  Selection of controlled variables.  C A should be selected since it directly affects the product quality (Rule 3).  T should be selected because it must be regulated properly to avoid safety problems (Rule 2) and because it interacts with C A (Rule 4).  h must be selected as a controlled output because it is non-self-regulating (Rule 1). EXAMPLE 1: CONTROL OF CSTR (Cont’d)

13  Selection of manipulated variables.  F i should be selected since it directly and rapidly affects C A (Guidelines 6, 7 and 8).  F c should be selected since it directly and rapidly affects T (Guidelines 6, 7 and 8). F o should be selected since it directly and rapidly affects h (Guidelines 6, 7 and 8). EXAMPLE 1: CONTROL OF CSTR (Cont’d)

14  This suggests the following control configuration:  Can you think of alternatives or improvements ? EXAMPLE 1: CONTROL OF CSTR (Cont’d)

15 PART TWO: Plantwide Control System design PART TWO: Plantwide Control System design Ref: Seider, Seader and Lewin, Chapter 20

16 PLANTWIDE CONTROL DESIGN Luyben et al. (1999) suggest a method for the conceptual design of plant-wide control systems, which consists of the following steps: Step 1: Establish the control objectives. Step 2: Determine the control degrees of freedom. Simply stated – the number of control valves – with additions if necessary. Step 3: Establish the energy management system. Regulation of exothermic or endothermic reactors, and placement of controllers to attenuate temperature disturbances. Step 4: Set the production rate. Step 5: Control the product quality and handle safety, environmental, and operational constraints.

17 PLANTWIDE CONTROL DESIGN (Cont’d) Step 6: Fix a flow rate in every recycle loop and control vapor and liquid inventories (vessel pressures and levels). Step 7: Check component balances. Establish control to prevent the accumulation of individual chemical species in the process. Step 8:Control the individual process units. Use remaining DOFs to improve local control, but only after resolving more important plant-wide issues. Step 9: Optimize economics and improve dynamic controllability. Add nice-to-have options with any remaining DOFs.

18 EXAMPLE 2: ACYCLIC PROCESS  Maintain a constant production rate  Achieve constant composition in the liquid effluent from the flash drum.  Keep the conversion of the plant at its highest permissible value. Steps 1 & 2: Establish the control objectives and DOFs: Select V-7 for On-demand product flow Select V-1 for fixed feed

19 EXAMPLE 2: ACYCLIC PROCESS (Cont’d)  Need to control reactor temperature: Use V-2. Step 3: Establish energy management system.  Need to control reactor feed temperature: Use V-3.

20 EXAMPLE 2: ACYCLIC PROCESS (Cont’d)  For on-demand product: Use V-7. Step 4: Set the production rate.

21 EXAMPLE 2: ACYCLIC PROCESS (Cont’d)  To regulate V-100 pressure: Use V-5 Step 5: Control product quality, and meet safety, environmental, and operational constraints.  To regulate V-100 temperature: Use V-6

22 EXAMPLE 2: ACYCLIC PROCESS (Cont’d) Step 6: Fix recycle flow rates and vapor and liquid inventories  Need to control vapor inventory in V-100: Use V-5 (already installed)  Need to control liquid inventory in V-100: Use V-4  Need to control liquid inventory in R-100: Use V-1

23 EXAMPLE 2: ACYCLIC PROCESS (Cont’d) Step 7: Check component balances. (N/A)  Install composition controller, cascaded with TC of reactor. Step 8: Control the individual process units (N/A) Step 9: Optimization

24 EXAMPLE 2 (Class): ACYCLIC PROCESS Try your hand at designing a plant-wide control system for fixed feed rate. Select V-1 for fixed feed

25 EXAMPLE 2 (Class): ACYCLIC PROCESS Possible solution.

26 EXAMPLE 3: CYCLIC PROCESS The above control system for (fixed feed) has an inherent problem? Can you see what it is?

27 EXAMPLE 3: CYCLIC PROCESS (Cont’d) The above control system for (fixed feed) has an inherent problem? Can you see what it is?

28 EXAMPLE 3: CYCLIC PROCESS (Cont’d)  Maintain the production rate at a specified level.  Keep the conversion of the plant at its highest permissible value. Steps 1 & 2: Establish the control objectives and DOFs:

29 EXAMPLE 3: CYCLIC PROCESS (Cont’d)  Need to control reactor temperature: Use V-2. Step 3: Establish energy management system.

30 EXAMPLE 3: CYCLIC PROCESS (Cont’d)  For fixed feed: Use V-1. Step 4: Set the production rate.

31 EXAMPLE 3: CYCLIC PROCESS (Cont’d)  To regulate V-100 pressure: Use V-4 Step 5: Control product quality, and meet safety, environmental, and operational constraints.  To regulate V-100 temperature: Use V-5

32 EXAMPLE 3: CYCLIC PROCESS (Cont’d) Step 6: Fix recycle flow rates and vapor and liquid inventories  Need to control vapor inventory in V-100: Use V-4 (already installed)  Need to control liquid inventory in V-100: Use V-3  Need to control liquid inventory in R-100: Cascade to FC on V-1.  Need to control recycle flow rate: Use V-6

33 EXAMPLE 3: CYCLIC PROCESS (Cont’d)  Install composition controller, cascaded with TC of reactor. Steps 7, 8 and 9: Improvements

34  Outlined qualitative approach for unit-by- unit control structure selection  Outlined qualitative approach for plantwide control structure selection SUMMARY