Self-optimizing control Theory

Outline Skogestad procedure for control structure design I Top Down Step S1: Define operational objective (cost) and constraints Step S2: Identify degrees of freedom and optimize operation for disturbances Step S3: Implementation of optimal operation What to control ? (primary CV’s) Active constraints Self-optimizing variables for unconstrained, c=Hy Step S4: Where set the production rate? (Inventory control) II Bottom Up Step S5: Regulatory control: What more to control (secondary CV’s) ? Step S6: Supervisory control Step S7: Real-time optimization

Step S3: Implementation of optimal operation Optimal operation for given d*: minu J(u,x,d) subject to: Model equations: f(u,x,d) = 0 Operational constraints: g(u,x,d) < 0 → uopt(d*) Problem: Usally cannot keep uopt constant because disturbances d change How should we adjust the degrees of freedom (u)? What should we control?

“Optimizing Control” y

“Self-Optimizing Control”:Separate optimization and control What should we control? (What is c? What is H?) Self-optimizing control: Constant setpoints give acceptable loss H y C = Hy H: Nonquare matrix Usually selection matrix of 0’s and some 1’s (measurement selection) Can also be full matrix (measurement combinations)

Definition of self-optimizing control Acceptable loss ) self-optimizing control “Self-optimizing control is when we achieve acceptable loss (in comparison with truly optimal operation) with constant setpoint values for the controlled variables (without the need to reoptimize when disturbances occur).” Reference: S. Skogestad, “Plantwide control: The search for the self-optimizing control structure'', Journal of Process Control, 10, 487-507 (2000).

Optimal operation - Runner Marathon runner c=heart rate J=T copt select one measurement c = heart rate CV = heart rate is good “self-optimizing” variable Simple and robust implementation Disturbances are indirectly handled by keeping a constant heart rate May have infrequent adjustment of setpoint (cs)

Unconstrained optimum Optimal operation Cost J Jopt copt Controlled variable c

Unconstrained optimum Optimal operation Cost J d Loss1 Jopt Loss2 n copt Controlled variable c Two problems: 1. Optimum moves because of disturbances d: copt(d) 2. Implementation error, c = copt + n

The ideal “self-optimizing” variable is the gradient, Ju Unconstrained degrees of freedom The ideal “self-optimizing” variable is the gradient, Ju c =  J/ u = Ju Keep gradient at zero for all disturbances (c = Ju=0) Problem: Usually no measurement of gradient u cost J Ju=0 Ju<0 uopt Ju 13

u J J>Jmin Jmin J<Jmin ? Unconstrained optimum: NEVER try to control a variable that reaches max or min at the optimum In particular, never try to control directly the cost J Assume we want to minimize J (e.g., J = V = energy) - and we make the stupid choice os selecting CV = V = J Then setting J < Jmin: Gives infeasible operation (cannot meet constraints) and setting J > Jmin: Forces us to be nonoptimal (two steady states: may require strange operation)

Parallel heat exchangers What should we control? Th1, UA1   T0 m [kg/s] Th2, UA2   T2 T1 T α 1-α 1 Degree of freedom (u=split ®) Objective: maximize J=T (= maximize total heat transfer Q) What should we control? If active constraint : Control it! Here: No active constraint. Control T? NO! Control T1=T2? Not quite Optimal: Control Ju=0

Jäsche temperature Can show that Ju=0 is equivalent to (see exercise for simplified case) TJ1 – TJ2 = 0 where TJ is «Jäschke temperature» for each branch: TJ1 = (T1-T0)2 / (Th1 – T0) TJ2 = (T2-T0)2 / (Th2 – T0) Th1 , Th2 = Inlet tempetaure on other side («hot») of heat exchanger T1 , T2 = outlet temperature T0 = feed temperature

c = JÄSCHKE TEMPERATURE

Unconstrained variables J copt H steady-state control error / selection controlled variable disturbance measurement noise Ideal: c = Ju In practise: c = H y

Optimal measurement combination CV=Measurement combination Optimal measurement combination Candidate measurements (y): Include also inputs u H measurement noise control error disturbance controlled variable

Nullspace method No measurement noise (ny=0) CV=Measurement combination Nullspace method

Nullspace method (HF=0) gives Ju=0 CV=Measurement combination Nullspace method (HF=0) gives Ju=0 Proof: Proof. Appendix B in: Jäschke and Skogestad, ”NCO tracking and self-optimizing control in the context of real-time optimization”, Journal of Process Control, 1407-1416 (2011) .

Example. Nullspace Method for Marathon runner CV=Measurement combination Example. Nullspace Method for Marathon runner u = power, d = slope [degrees] y1 = hr [beat/min], y2 = v [m/s] F = dyopt/dd = [0.25 -0.2]’ H = [h1 h2]] HF = 0 -> h1 f1 + h2 f2 = 0.25 h1 – 0.2 h2 = 0 Choose h1 = 1 -> h2 = 0.25/0.2 = 1.25 Conclusion: c = hr + 1.25 v Control c = constant -> hr increases when v decreases (OK uphill!)

Extension: ”Exact local method” (with measurement noise) CV=Measurement combination Extension: ”Exact local method” (with measurement noise) General analytical solution (“full” H): No disturbances (Wd=0) + same noise for all measurements (Wny=I): Optimal is H=GyT (“control sensitive measurements”) Proof: Use analytic expression No noise (Wny=0): Cannot use analytic expression, but optimal is clearly HF=0 (Nullspace method) Assumes enough measurements: #y ¸ #u + #d If “extra” measurements (>) then solution is not unique

Good candidate controlled variables c=Hy (for self-optimizing control) CV=Single measurements Good candidate controlled variables c=Hy (for self-optimizing control) The optimal value of c should be insensitive to disturbances Want small dcopt/dd = HF The value of c should be sensitive to changes in the degrees of freedom Want large dc/du = Hgy (same as wanting flat optimum) c should be easy to measure and control Proof: This keeps the gradient close to zero since BAD Good

Example: CO2 refrigeration cycle pH J = Ws (work supplied) DOF = u (valve opening, z) Main disturbances: d1 = TH d2 = TCs (setpoint) d3 = UAloss What should we control?

CO2 refrigeration cycle Step 1. One (remaining) degree of freedom (u=z) Step 2. Objective function. J = Ws (compressor work) Step 3. Optimize operation for disturbances (d1=TC, d2=TH, d3=UA) Optimum always unconstrained Step 4. Implementation of optimal operation No good single measurements (all give large losses): ph, Th, z, … Nullspace method: Need to combine nu+nd=1+3=4 measurements to have zero disturbance loss Simpler: Try combining two measurements. Exact local method: c = h1 ph + h2 Th = ph + k Th; k = -8.53 bar/K Nonlinear evaluation of loss: OK!

CO2 cycle: Maximum gain rule

Refrigeration cycle: Proposed control structure CV=Measurement combination Refrigeration cycle: Proposed control structure Control c= “temperature-corrected high pressure”

Importance of optimal operation for CO2 capturing process Dependency of equivalent energy in CO2 capture plant verses recycle amine flowrate An amine absorption/stripping CO2 capturing process* *Figure from: Toshiba (2008). Toshiba to Build Pilot Plant to Test CO2 Capture Technology. http://www.japanfs.org/en/pages/028843.html.

Case study Control structure design using self-optimizing control for economically optimal CO2 recovery* Step S1. Objective function= J = energy cost + cost (tax) of released CO2 to air 4 equality and 2 inequality constraints: stripper top pressure condenser temperature pump pressure of recycle amine cooler temperature CO2 recovery ≥ 80% Reboiler duty < 1393 kW (nominal +20%) Step S2. (a) 10 degrees of freedom: 8 valves + 2 pumps 4 levels without steady state effect: absorber 1,stripper 2,make up tank 1 Disturbances: flue gas flowrate, CO2 composition in flue gas + active constraints (b) Optimization using Unisim steady-state simulator. Mode I = Region I (nominal feedrate): No inequality constraints active 2 unconstrained degrees of freedom =10-4-4 Step S3 (Identify CVs). 1. Control the 4 equality constraints 2. Identify 2 self-optimizing CVs. Use Exact Local method and select CV set with minimum loss. *M. Panahi and S. Skogestad, ``Economically efficient operation of CO2 capturing process, part I: Self-optimizing procedure for selecting the best controlled variables'', Chemical Engineering and Processing, 50, 247-253 (2011).

Proposed control structure with given nominal flue gas flowrate (mode I)

Mode II: large feedrates of flue gas (+30%) (kmol/hr) Self-optimizing CVs in region I Reboiler duty (kW) Cost (USD/ton) CO2 recovery % Temperature tray no. 16 °C Optimal nominal point 219.3 95.26 106.9 1161 2.49 +5% feedrate 230.3 1222 +10% feedrate 241.2 1279 +15% feedrate 252.2 1339 +19.38%, when reboiler duty saturates 261.8 1393 (+20%) 2.50 +30% feedrate (reoptimized) 285.1 91.60 103.3 2.65 region I region II max Saturation of reboiler duty; one unconstrained degree of freedom left Use Maximum gain rule to find the best CV among 37 candidates : Temp. on tray no. 13 in the stripper: largest scaled gain, but tray 16 also OK

Proposed control structure with large flue gas flowrate (mode II = region II) max

Conditions for switching between regions of active constraints (“supervisory control”) Within each region of active constraints it is optimal to Control active constraints at ca = c,a, constraint Control self-optimizing variables at cso = c,so, optimal Define in each region i: Keep track of ci (active constraints and “self-optimizing” variables) in all regions i Switch to region i when element in ci changes sign

Example – switching policies CO2 plant (”supervisory control”) Assume operating in region I (unconstrained) with CV=CO2-recovery=95.26% When reach maximum Q: Switch to Q=Qmax (Region II) (obvious) CO2-recovery will then drop below 95.26% When CO2-recovery exceeds 95.26%: Switch back to region I !!!

Conclusion optimal operation ALWAYS: 1. Control active constraints and control them tightly!! Good times: Maximize throughput -> tight control of bottleneck 2. Identify “self-optimizing” CVs for remaining unconstrained degrees of freedom Use offline analysis to find expected operating regions and prepare control system for this! One control policy when prices are low (nominal, unconstrained optimum) Another when prices are high (constrained optimum = bottleneck) ONLY if necessary: consider RTO on top of this

Sigurd’s rules for CV selection Always control active constraints! (almost always) Purity constraint on expensive product always active (no overpurification): (a) "Avoid product give away" (e.g., sell water as expensive product) (b) Save energy (costs energy to overpurify) Unconstrained optimum: NEVER try to control a variable that reaches max or min at the optimum In particular, never try to control directly the cost J - Assume we want to minimize J (e.g., J = V = energy) - and we make the stupid choice os selecting CV = V = J - Then setting J < Jmin: Gives infeasible operation (cannot meet constraints) - and setting J > Jmin: Forces us to be nonoptimal (which may require strange operation; see Exercise on recycle process)