Adventures in pH Control

7,286 views

Published on

Presented by Greg McMillan and Dave Joseph at the 2012 Emerson Exchange in Anaheim, California USA.

Published in: Education

Adventures in pH Control

  1. 1. Adventures in pH ControlGreg McMillan CDI Process & IndustrialDave Joseph Rosemount Analytical
  2. 2. Photography & Video Recording Policy Photography and audio/video recording is not permitted in any sessions or in the exhibition areas without press credentials or written permission from the Emerson Exchange Board of Directors. Inquiries should be directed to: EmersonExchange@Emerson.com Thank you.
  3. 3. Presenters Greg McMillanPrincipal ConsultantEmail: Greg.McMillan@Emerson.com33 years Monsanto-Solutia Fellow2 years WU Adjunct Professor10 years DeltaV R&D ContractorBS Engineering PhysicsMS Control Theory Dave JosephSr. Industry ManagerEmail: Dave.joseph@emerson.com24 years with Rosemount AnalyticalBS and MS in Chemical EngineeringMember AIChE
  4. 4. Key Benefits of Course  Recognize the opportunity/challenges of pH control  Learn about modeling and control options  Optimize hardware implementation  Understand the root causes of poor performance  Prioritize improvements based on cost, time, and goal  Gather insights for applications and solutions 4
  5. 5. Section 1: Measuring pH  Brief theory of pH  Inside a pH sensor  The Smart pH sensor  Diagnostics 5
  6. 6. Top Ten Signs of a Rough pH Startup  Food is burning in the operators’ kitchen  Only loop mode configured is manual  Operator puts his fist through the screen  You trip over a pile of used pH electrodes  Technicians ask: “what is a positioner?”  Technicians stick electrodes up your nose  Environmental engineer is wearing a mask  Plant manager leaves the country  Lawyers pull the plugs on the consoles  President is on the phone holding for you 6
  7. 7. The definition of pH pH is the unit of measurement for determining the acidity or alkalinity of a solution. The mathematical definition of pH is the negative logarithm of the molar hydrogen ion concentration, pH = - log([H+]) pH is measured by various different sensors, H2O most common and economical is the glass H+ OH- OH- H+ electrode/silver reference system. pH measurement requires periodic maintenance to maintain accuracy.
  8. 8. pH Scale vs Moles/Liter IonConcentration pH Hydrogen Ion [H+] Hydroxyl Ion [OH-] 0 Acidic 1.0 0.00000000000001 1 0.1 0.0000000000001 2 0.01 0.000000000001 3 0.001 0.00000000001 4 0.0001 0.0000000001 5 0.00001 0.000000001 6 0.000001 0.00000001 7 Neutral 0.0000001 0.0000001 8 0.00000001 0.000001 9 0.000000001 0.00001 10 0.0000000001 0.0001 11 0.00000000001 0.001 12 0.000000000001 0.01 13 0.0000000000001 0.1 14 Basic 0.00000000000001 1.0
  9. 9. pH Values of Acids and Bases 14 4.0 % Sodium Hydroxide 12 0.04% Sodium Hydroxide 10 Milk of Magnesia 8 0.84% Sodium BicarbonatepH Water @ 25ºC 6 0.00001% Sulfuric Acid 4 0.0001% Hydrochloric Acid 0.01% Sulfuric Acid 2 0.1% Hydrochloric Acid 4.9% Sulfuric Acid 0 1E00 1E-01 1E-02 1E-03 1E-04 1E-05 1E-06 1E-07 1E-08 1E-09 1E-10 1E-11 1E-12 1E-13 1E-14 Hydrogen Ion Concentration (Mole/Liter)
  10. 10. What is pH? – technical stuff  pH = - log([H+])  Kw = [H+]*[OH-] = 1.0x10-14 at 25ºC  pH + pOH = pKw  pH is measured using the Nernst equation  E(mV) = Ex + 2.3(RT/F)*log aH+  ~ Ex – (S)*pH in simple form  Where Ex = calibration constant  2.3(RT/F) ~ slope (S) in mV/pH units  aH+ = activity of hydrogen ion ~ [H+]
  11. 11. Theoretical Response of apH Sensor (25ºC) 500 400 300 Slope of 59.16 mV/pH Unit 200 100 0 mV-100-200 Zero mV at 7 pH-300-400 pH-500 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
  12. 12. pH Sensor Basics•The pH electrodeproduces a potential(in mV’s) • The referenceproportional to the electrode potentialpH of the solution. must remain stable regardless of process or time effects Glass Body Ag/AgCl • Internal element- AgAgCl Internal Shield Wire • Electrolyte fill - KCl/AgCl • Liquid Junctions pH Sensitive Fill Glass Solution 12
  13. 13. Inside the pH Glass Membrane... Alkali Metal Ions Anionic Sites Glass Matrix (unaffected) M M M M S M M Core Glass S S S M Leached Layer + M H Inner Zone (not to scale) + + H H + + + H + H H H Outer Zone + + H HLeached LayerDissolving Hydrogen Ions 13
  14. 14. The reference electrode The Reference Cell maintains a stable potential regardless of the process pH or changes in the activities of other ions in AgCl/KCl solution. Ag/AgCl Fill Internal Solution Wire The Liquid Junction completes the electrical circuit between the pH measuring electrode and the reference cell via the process solution. Liquid Junction 14pH17
  15. 15. The sum of all potentials… Assuming a preamp with low leakage current, the pH sensor Ex = Eoutside of glass (in process solution) - Einside of glass (in glass fill solution) - Emeasurement wire (in glass fill solution) + Ereference wire (in reference Ag/KCL solution) + Ejunction potential (sum of all interface potentials)  Glass fill solution typically formulated to cancel out effects so that 7 pH is 0 mV at any temperature.
  16. 16. Double Junction Combination pH Electrode - Circuit Diagram Em W R3 Er W R4 solution ground silver-silver chlorideDehydration, loss of active sites, internal electrodechemical attack, and premature E4 aging reduces efficiency and secondmakes sensor dramatically slow W R5 junction potassium chloride (KCl) electrolyte Gel layer is used as a term in salt bridge between junctions for the glass surface that has water molecules primary R6 W junction inner E5 gel silver-silver chloride layer internal electrode E3 outer W R2 gel pH fill solution layer E2 Process ions may Measurement R1 W migrate into porous becomes slow Ii reference junction if glass gets coated E1 while electrolyte ions Process Fluid migrate out W W R10 W W R9 R7 R8 High acid or base concentrations can affect glass gel layer and reference junction potential Increase in noise or decrease in span or efficiency is indicative of glass electrode problem Shift or drift in pH measurement is normally associated with reference electrode problem 16
  17. 17. Life Depends On Process Conditions Months >100% increase in life from new glass designs for high temperatures 25ºC 50ºC 75ºC 100ºC Process Temperature  High pH conditions decrease glass life at any temperature  Degraded accuracy and response time is also common  Leads to unreliable feedforward control 17
  18. 18. New Glass preserves response time After 120 hours exposure at 140ºC 200 mV 150 New Glass 100 Other 50 minutes 0 0 50 100 150 200 Glass electrodes get slow as they age High temperatures cause accelerated aging New glass formulations can resist this effect 18
  19. 19. Review: pH Measurement loop Analyzer (not part of the sensor) 4. Solution ground 2. Reference electrode Liquid junction 1. Glass electrode 3.Temperature element
  20. 20. What is a SMART sensor?  SMART sensors store calibration data on an embedded chip.  SMART sensors record the initial calibration data of the sensor and all data from the last 5 calibrations  They allow trending the performance variables of the sensor to determine how healthy the sensor is and what work is needed on it before venturing out into the field.  Trended diagnostics enable Plantweb users to take action before the reading is compromised without any intimate knowledge of how a sensor works or what conditions the sensor may have been exposed to.  Results are reduced maintenance and increased measurement uptime.
  21. 21. SMART loop: instrument-cable-sensor 4-wire Models 56 and 1056 are smart-enabled 2-wire, FF Model 1066 is smart-enabled Smart pH sensorused sensor used OR in “Smart” mode VP8 (or cable) Model 6081pH Smart-enabled Wireless Transmitter
  22. 22. SMART pH Sensors  Plug and Play - Factory pre-calibrated - Calibrate in lab instead of in field - Can restore to factory values  SMART technology - Automatically trend diagnostics - Capture intermittent sensor problems - SMART signal superimposed on mV signal (like HART)  Simple Migration path - Compatible with previous analyzers - Compatible with previous sensors
  23. 23. Calibration history Advanced diagnostics  Last 5 calibration data sets for troubleshooting
  24. 24. Calibration data set – diagnostics Current readings! Calibration Data Time stamp between calibrations Calibration method Slope Offset Temperature at the time of calibration Glass impedance Reference impedance
  25. 25. Calibration History
  26. 26. Plug & Play Convenience Conventional approach: Field calibration with buffers SMART approach: Cal in the lab, Plug & Play in the fieldConventional sensor Field Equipment Smart Sensor Field Equipment
  27. 27. Siemens Water Technologies, WI • Application: spent caustic, pH ~10-12 • pH sensor: 3500HTVP and 396PVP • User comment: “The SMART is somewhat fool proof. I do like the backward compatibility with it, because initially we had the wrong probes hooked to the wrong boards, and everything still worked. The SMART features obviously didnt, but the probes themselves all functioned fine. “
  28. 28. Key Indicators of Sensor PerformancePlantweb pH measurementsprovide a complete view ofthe operational parameters: pH reading raw sensor output temperature reference impedance Glass impedance RTD resistance
  29. 29. Diagnostics - Broken Glass 3K 150 M 0-5   Broken Glass! Reference Electrode Glass Solution Electrode GroundBroken Glass Fault  pH Glass electrode normally has high impedance of 50-500 Megohm  Recommended setting of 10 Megohm will detect even hairline cracks  Glass can be cracked at the tip or further back inside the sensor (and not easily visible) 29
  30. 30. Diagnostics - Coated Sensor 3K 40k 150 M   Coated Sensor! Reference Electrode Glass Solution Electrode GroundCoated Sensor Fault (Ref Z Too High)  pH Reference electrode normally has low impedance of 1-10 KilOhm  Reference coating slowly builds up around the junction  Setting of 20 KilOhm should not generally cause false alarms 30
  31. 31. Diagnostics - Non-Immersed Sensor 60K 1500 M   Dry Sensor! Reference Electrode Glass Solution Electrode Ground Dry Sensor Fault (Glass Z Too High)  pH Glass electrode normally has impedance of 50-500 Megohm  When sensor is dry there is no continuity between the electrode(s) and the solution ground so impedance reading is very high  Recommended setting of 1000 Megohm will not cause false alarms 31
  32. 32. More Advanced Diagnostics  pH parameters slope, reference offset, glass, and reference impedances change little over bulk of operational life.  When parameters start to change, they indicate that more frequent calibrations will be necessary.  Diagnostics are at their most powerful when they can be compared to the original properties of the sensor.  Example: a pH slope of 54 may not indicate a problem, but a sudden drop in slope from 58 to 54 may indicate a 9 month old sensor will not last much longer.  Trending the electrode slope, reference offset, and reference impedance will show the first sign of problems.
  33. 33. SMART and Trending DiagnosticsSMART pH sensorsautomatically record their ^mV/ pH 59 58.9 58.8 Max pH error per calibration cycle Change in Slope 58.7 58.6 58.5 58.4 58.3 58.2 58.1 58 57.9 57.8 57.7 57.6 57.5 57.4 57.3initial conditions and the ^pH 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 0.1 # 0.02% # 0.03% # 0.05% # 0.07% # 0.09% # 0.10% # 0.12% # 0.14% # 0.15% # 0.17% # 0.19% # 0.21% # 0.23% # 0.24% # 0.26% # 0.28% # 0.30% 0.2 # 0.03% # 0.07% # 0.10% # 0.14% # 0.17% # 0.21% # 0.24% # 0.27% # 0.31% # 0.34% # 0.38% # 0.42% # 0.45% # 0.49% # 0.52% # 0.56% # 0.59% 0.3 # 0.05% # 0.10% # 0.15% # 0.20% # 0.26% # 0.31% # 0.36% # 0.41% # 0.46% # 0.52% # 0.57% # 0.62% # 0.68% # 0.73% # 0.78% # 0.84% # 0.89% 0.4 # 0.07% # 0.14% # 0.20% # 0.27% # 0.34% # 0.41% # 0.48% # 0.55% # 0.62% # 0.69% # 0.76% # 0.83% # 0.90% # 0.97% # 1.04% # 1.11% # 1.19% 0.5 # 0.08% # 0.17% # 0.26% # 0.34% # 0.43% # 0.51% # 0.60% # 0.69% # 0.77% # 0.86% # 0.95% # 1.04% # 1.13% # 1.22% # 1.30% # 1.39% # 1.48%last 5 calibrations to make 0.6 # 0.10% # 0.20% # 0.31% # 0.41% # 0.51% # 0.62% # 0.72% # 0.82% # 0.93% # 1.03% # 1.14% # 1.25% # 1.35% # 1.46% # 1.57% # 1.67% # 1.78% 0.7 # 0.12% # 0.24% # 0.36% # 0.48% # 0.60% # 0.72% # 0.84% # 0.96% # 1.08% # 1.21% # 1.33% # 1.45% # 1.58% # 1.70% # 1.83% # 1.95% # 2.08% 0.8 # 0.14% # 0.27% # 0.41% # 0.55% # 0.68% # 0.82% # 0.96% # 1.10% # 1.24% # 1.38% # 1.52% # 1.66% # 1.80% # 1.94% # 2.09% # 2.23% # 2.37% 0.9 # 0.15% # 0.31% # 0.46% # 0.61% # 0.77% # 0.92% # 1.08% # 1.24% # 1.39% # 1.55% # 1.71% # 1.87% # 2.03% # 2.19% # 2.35% # 2.51% # 2.67% 1 # 0.17% # 0.34% # 0.51% # 0.68% # 0.85% # 1.03% # 1.20% # 1.37% # 1.55% # 1.72% # 1.90% # 2.08% # 2.25% # 2.43% # 2.61% # 2.79% # 2.97% 1.1 # 0.19% # 0.37% # 0.56% # 0.75% # 0.94% # 1.13% # 1.32% # 1.51% # 1.70% # 1.90% # 2.09% # 2.28% # 2.48% # 2.67% # 2.87% # 3.07% # 3.26%trending easier. 1.2 1.3 1.4 1.5 1.6 # 0.20% # 0.22% # 0.24% # 0.25% # 0.27% # # # # # 0.41% # 0.61% 0.44% # 0.66% 0.48% # 0.72% 0.51% # 0.77% 0.54% # 0.82% # # # # # 0.82% 0.89% 0.96% 1.02% 1.09% # # # # # 1.03% 1.11% 1.20% 1.28% 1.37% # # # # # 1.23% 1.34% 1.44% 1.54% 1.64% # # # # # 1.44% 1.56% 1.68% 1.80% 1.92% # # # # # 1.65% 1.79% 1.92% 2.06% 2.20% # # # # # 1.86% 2.01% 2.17% 2.32% 2.48% # # # # # Beginning Slope - Ending Slope = 2.07% 2.24% 2.41% 2.59% # # # # 2.28% 2.47% Slope Change 2.66% 2.85% # # # # 2.49% 2.70% 2.91% 3.11% # # # # Slope change * Max change in pH = 2.76% # 3.04% # 3.32% # 2.70% 2.93% 3.15% 3.38% 3.60% # # # # # 2.92% 3.16% 3.40% 3.65% 3.89% # # # # # 3.13% 3.39% 3.65% 3.91% 4.17% # # # # # 3.34% 3.62% 3.90% 4.18% 4.46% # # # # # 3.56% 3.86% 4.15% 4.45% 4.75%• Predictive maintenance 1.7 # 0.29% # 0.58% # 0.87% # 1.16% # 1.45% # 1.75% # 2.04% # 2.34% # 2.63% # 2.93% # 3.23% # 3.53% # 3.83% # 4.13% # 4.43% # 4.74% # 5.04% 1.8 # 0.31% # 0.61% # 0.92% # 1.23% # 1.54% # 1.85% # 2.16% # 2.47% # 2.79% # 3.10% # 3.42% # 3.74% # 4.06% # 4.38% # 4.70% # 5.02% # 5.34% 1.9 # 0.32% # 0.65% # 0.97% # 1.30% # 1.62% # 1.95% # 2.28% # 2.61% # 2.94% # 3.28% Max mV deviation per # 3.61% # 3.94% # 4.28% # 4.62% # 4.96% # 5.30% # 5.64% 2 # 0.34% # 0.68% # 1.02% # 1.37% # 1.71% # 2.05% # 2.40% # 2.75% # 3.10% # 3.45% # 3.80% # 4.15% # 4.51% # 4.86% # 5.22% # 5.57% # 5.93% 2.1 # 0.36% # 0.71% # 1.07% # 1.43% # 1.79% # 2.16% # 2.52% # 2.89% # 3.25% # 3.62% calibration cycle # 3.99% # 4.36% # 4.73% # 5.10% # 5.48% # 5.85% # 6.23% 2.2 # 0.37% # 0.75% # 1.12% # 1.50% # 1.88% # 2.26% # 2.64% # 3.02% # 3.41% # 3.79% # 4.18% # 4.57% # 4.96% # 5.35% # 5.74% # 6.13% # 6.53% 2.3 # 0.39% # 0.78% # 1.18% # 1.57% # 1.97% # 2.36% # 2.76% # 3.16% # 3.56% # (Max mV deviation / Beginning 3.97% # 4.37% # 4.78% # 5.18% # 5.59% # 6.00% # 6.41% # 6.82%with reference Impedance 2.4 2.5 2.6 2.7 2.8 # 0.41% # 0.42% # 0.44% # 0.46% # 0.48% # # # # # 0.82% # 1.23% 0.85% # 1.28% 0.88% # 1.33% 0.92% # 1.38% 0.95% # 1.43% # # # # # 1.64% 1.71% 1.77% 1.84% 1.91% # # # # # 2.05% 2.14% 2.22% 2.31% 2.39% # # # # # 2.47% 2.57% 2.67% 2.77% 2.88% # # # # # 2.88% 3.00% 3.12% 3.24% 3.36% # # # # # 3.30% 3.44% 3.57% 3.71% 3.85% # # # # # 3.72% 3.87% 4.03% 4.18% 4.34% # # # # # 4.14% 4.31% 4.48% 4.66% 4.83% # # # # # 4.56% 4.75% 4.94% 5.13% 5.32% # # # # # 4.98% Slope) = Max pH error 5.19% 5.40% 5.61% 5.81% # # # # # 5.41% 5.63% 5.86% 6.08% 6.31% # # # # # 5.83% 6.08% 6.32% 6.56% 6.81% # # # # # 6.26% 6.52% 6.78% 7.04% 7.30% # # # # # 6.69% 6.97% 7.25% 7.53% 7.80% # # # # # 7.12% 7.42% 7.71% 8.01% 8.31%trending 2.9 # 0.49% # 0.99% # 1.48% # 1.98% # 2.48% # 2.98% # 3.48% # 3.99% # 4.49% # 5.00% # 5.51% # 6.02% # 6.53% # 7.05% # 7.57% # 8.08% # 8.60% 3 # 0.51% # 1.02% # 1.53% # 2.05% # 2.56% # 3.08% # 3.60% # 4.12% # 4.65% # 5.17% # 5.70% # 6.23% # 6.76% # 7.29% # 7.83% # 8.36% # 8.90% 3.1 # 0.53% # 1.05% # 1.58% # 2.12% # 2.65% # 3.18% # 3.72% # 4.26% # 4.80% # 5.34% # 5.89% # 6.44% # 6.98% # 7.53% # 8.09% # 8.64% # 9.20% 3.2 # 0.54% # 1.09% # 1.64% # 2.18% # 2.74% # 3.29% # 3.84% # 4.40% # 4.96% # 5.52% # 6.08% # 6.64% # 7.21% # 7.78% # 8.35% # 8.92% # 9.49% 3.3 # 0.56% # 1.12% # 1.69% # 2.25% # 2.82% # 3.39% # 3.96% # 4.54% # 5.11% # 5.69% # 6.27% # 6.85% # 7.44% # 8.02% # 8.61% # 9.20% # 9.79%• Determine optimum 3.4 # 0.58% # 1.16% # 1.74% # 2.32% # 2.91% # 3.49% # 4.08% # 4.67% # 5.27% # 5.86% # 6.46% # 7.06% # 7.66% # 8.26% # 8.87% # 9.48% # 10.09% 3.5 # 0.59% # 1.19% # 1.79% # 2.39% # 2.99% # 3.60% # 4.20% # 4.81% # 5.42% # 59mV/pH - 55mV/pH = 4mV/pH * 6pH = 6.03% # 6.65% # 7.27% # 7.89% # 8.51% # 9.13% # 9.76% # 10.38% 3.6 # 0.61% # 1.22% # 1.84% # 2.46% # 3.08% # 3.70% # 4.32% # 4.95% # 5.58% # 6.21% # 6.84% # 7.47% # 8.11% # 8.75% # 9.39% # 10.03% # 10.68% 3.7 # 0.63% # 1.26% # 1.89% # 2.53% # 3.16% # 3.80% # 4.44% # 5.09% # 5.73% # 6.38% 24mV / 59mV/pH = 0.41pH # 7.03% # 7.68% # 8.34% # 8.99% # 9.65% # 10.31% # 10.98% 3.8 # 0.65% # 1.29% # 1.94% # 2.59% # 3.25% # 3.90% # 4.56% # 5.22% # 5.89% # 6.55% # 7.22% # 7.89% # 8.56% # 9.24% # 9.91% # 10.59% # 11.27% 3.9 # 0.66% # 1.33% # 1.99% # 2.66% # 3.33% # 4.01% # 4.68% # 5.36% # 6.04% # 6.72% # 7.41% # 8.10% # 8.79% # 9.48% # 10.17% # 10.87% # 11.57% 4 # 0.68% # 1.36% # 2.04% # 2.73% # 3.42% # 4.11% # 4.80% # 5.50% # 6.20% # 6.90% # 7.60% # 8.30% # 9.01% # 9.72% # 10.43% # 11.15% # 11.87%calibration frequency and Change in Process = 1% = 2% = 3% Beginning Slope - Ending Slope = Slope Change = 4% = 5%predict probe life with pH Slope Change * Maximum Change in Process pH = Maximum pH Deviation For a typical application ranging from 4 toslope trending 10 pH the error from assuming 59 slope instead of 55 could be 0.41 pH units >>> need to recalibrate
  34. 34. Using Diagnostics Instruments ship with the diagnostics turned off When enabled, default setpoints will generally be ok Few false alarms when correctly configured Some problems may not be detectable with online diagnostics When in doubt, check with buffers
  35. 35. Section 2: Modeling and Control  Virtual plant and embedded process models  Online identification of titration curve  Minimization of project capital cost  Cascade pH control  Batch pH control  Linear reagent demand control  Elimination of split range control  Model predictive control 35
  36. 36. Embedded Process Model for pH 36
  37. 37. Titration Curves can Vary Weak Acid and Strong Base Strong Acid and Weak Base pka = 10 pka = 4 Slope moderated near each pKa Weak Acid and Weak Base Multiple Weak Acids and Weak Bases pKa and curve change with pka = 10 temperature! pka = 9 pka = 5 pka = 4 pka = 3 37
  38. 38. Nonlinearity can cost big moneypH measurement error may look smaller on the flatter portion of a titrationcurve but the associated reagent delivery error is larger 10 pH 4 Reagent to Feed Reagent Flow Ratio Optimum Savings Original set point set point Oscillations could be due to non-ideal mixing, control valve stick-slip. or pressure fluctuations 38
  39. 39. Titration Curve Matched to Plant pH Slope 39
  40. 40. Modeled pH Control System AY signal pH set point characterizer 1-3 Signal characterizers linearize loop via reagent demand control AC 1-1 LC LT 1-5 1-5 signal splitter characterizer AY Feed 1-4 AY 1-2 NaOH Acid To other Tank middle signal selector FT FT Tank 1-1 1-2 AY 1-1 AT AT AT Eductors 1-1 1-2 1-3 Static Mixer From other Tank To other Tank Downstream system 40
  41. 41. Conventional vs. Reagent Demand One of many spikes of recirculation pH spikes from stick-slip of water valve Influent pH Tank 1 pH for Reagent Demand Control Tank 1 pH for Conventional pH Control Start of Step 4 (Slow Rinses) Start of Step 2 (Regeneration) 41
  42. 42. Traditional System for Minimum VariabilityThe period of oscillation (4 x process dead time) and filter time(process residence time) is proportional to volume. To prevent Reagentresonance of oscillations, different vessel volumes are used. Major overlooked Reagent Reagent problem is reagent Deliver delay from dip tube design FeedSmall first tank provides a faster responseand oscillation that is more effectively filtered Big footprintby the larger tanks downstream and high cost! 42
  43. 43. Traditional System forMinimum Reagent Use Reagent The period of oscillation (total loop dead time) must differ by more than factor of 5 to prevent resonance (amplification of oscillations) Feed Reagent Reagent Big footprint and high cost! The large first tank offers more cross neutralization of influents 43
  44. 44. Tight pH Control withMinimum Capital IL#1 – Interlock that prevents back fill of reagent piping when control valve closes IL#2 – Interlock that shuts off effluent flow until vessel pH is projected to be within control band Eductor High Recirculation Flow Reagent Any Old Tank Signal Characterizer LC LT 1-3 1-3 *IL#2 f(x) FT 1-1Effluent AC 1-1 FC 1-2 AT 1-1 *IL#1 Influent FT 1-2 10 to 20 pipe diameters 44
  45. 45. Linear Reagent Demand Control Signal characterizer converts PV and SP from pH to % Reagent Demand – PV is abscissa of the titration curve scaled 0 to 100% reagent demand – Piecewise segment fit normally used to go from ordinate to abscissa of curve – Fieldbus block offers 21 custom space X,Y pairs (X is pH and Y is % demand) – Closer spacing of X,Y pairs in control region provides most needed compensation – If neural network or polynomial fit used, beware of bumps and wild extrapolation Special configuration is needed to provide operations with interface to: – See loop PV in pH and signal to final element – Enter loop SP in pH – Change mode to manual and change manual output Set point on steep part of curve shows biggest improvements from: – Reduction in limit cycle amplitude seen from pH nonlinearity – Decrease in limit cycle frequency from final element resolution (e.g. stick-slip) – Decrease in crossing of split range point – Reduced reaction to measurement noise – Shorter startup time (loop sees real distance to set point and is not detuned) – Simplified tuning (process gain no longer depends upon titration curve slope) – Restored process time constant (slower pH excursion from disturbance) 45
  46. 46. Cascade Control to ReduceDownstream Offset Linear Reagent Demand Controller Flow Feedforward FT 1-1 RSP FC AC Trim of Inline 1-1 Sum Set Point 1-1 Reagent AT f(x) Filter f(x) 1-1 FT Static Mixer PV signal 1-2 SP signal Characterizer characterizer Feed Coriolis Mass 10 to 20 Flow Meter pipe diameters M AC 1-2 Any Old Tank Enhanced PID Controller AT 1-2 46
  47. 47. Full Throttle Batch pH Control Batch pH End Point Predicted pH Reagent Cutoff Sum Rate of Projected Past Change DpH New pH DpH DpH/Dt Sub Div Mul Old pH Delay Dt Total System Dead Time Batch Reactor Filter AT 1-1 10 to 20 pipe diameters Section 3-5 in New Directions in Bioprocess Modeling and Control shows how this strategy is used as a head start for a PID controller 47
  48. 48. Linear Reagent Demand Batch pH Control FQ FT Secondary pH 1-1 1-1 PI Controller AC FC 1-1 1-1Influent #1 AT Online Curve 1-1 Identification Static Mixer 10 to 20 pipe FT diameters 1-2 Influent #2 AC f(x) 1-1 Batch Reactor Signal Master Reagent Demand Characterizer Adaptive PID Controller AT Uses Online 1-1 Titration Curve 10 to 20 Reduces injection and mixing delays and enables some cross pipe neutralization of swings between acidic and basic influent. It is diameters suitable for continuous control as well as fed-batch operation. 48
  49. 49. Conventional Fine andCoarse Valve Control Large Small (Coarse) (Fine) ZC CV 1-1 Integral only Controller (CV is Implied Fine Control Valve Position) Neutralizer AC ZC speed of response must 1-1 be slow and tuning is difficult Must add feedforward for fast and large influent disturbance PID Controller AT 1-1 49

×