and MAPPS USERS GROUP

Transcription

1 SYSTEMS ANALYSIS PROJECT ADVISORY COMMITTEE and MAPPS USERS GROUP SLIDE MATERIAL October 23-24, 1986

2 AGENDA STATUS PERFORMANCE ATTRIBUTES MECHANICAL PULPING UTILITIES OPTIMIZATION FUTURE DEVELOPMENTS STATUS 23 CLIENTS 4 TRAINING COURSES VERSION 3.0 DUE IN LATE FALL PERFORMANCE ATTRIBUTE STRUCTURE MECHANICAL PULPING NEW UTILITIES NEW MODULES

3 -2- MARKETING AND ALLIED EFFORTS DEMONSTRATION AT TAPPI ANNUAL MEETING CHAIRED SESSIONS AT TAPPI AND AICHE MEETINGS PRESENTED SEVERAL MAPPS SEMINARS PREPARED TWO PAPERS FOR PUBLICATION/PRESENTATION PERFORMANCE ATTRIBUTE MODELING PURPOSE IS TO DEFINE STRUCTURE CARRY NONCONSERVED INFORMATION TESTED WITH MECHANICAL PULPING MODULES SIMILAR TO EXISTING STREAM STRUCTURE

4 -3- SYSTEMS ANALYSIS PROJECT ADVISORY COMMITTEE MEETING OCTOBER 1986 GARY L. JONES PROCESS MODELING GROUP MAPPS DEVELOPMENT TOPICS * MECHANICAL PULPING MODELING SIMULATION * PERFORMANCE ATTRIBUTES * OPTIMIZATION MECHANICAL PULPING OBJECTIVE BY FALL MEETING: 1. DEVELOP MAPPS MODULES AND TO SIMULATE MAIN FEATURES PULPING PROCESSES FLOW SHEET OF MECHANICAL 2. INCORPORATE AND TEST FEASIBILITY OF SIMULATION OF PERFORMANCE ATTRIBUTES /

5 -4- MECHANICAL PULPING BACKGROUND * FAMILY OF HIGH-YIELD PULPING PROCESSES - TMP, CMP, TCMP - SGW, RMP, PGW * MAJOR STEPS - PRETREATMENT - CHIP REFINING OR GRINDING - SECONDARY REFINING - SCREENING - CLEANING - REJECT o REFINING o SCREENING o CLEANING - CONSISTENCY CONTROL o STOCK MIXING o DILUTION o THICKENING - PROPERTY DEVELOPMENT o FIBERS SIZE REDUCTION o SHIVE REMOVAL o CURL SETTING OR REMOVAL o SURFACE AREA DEVELOPMENT o BRIGHTENING

6 -5- MECHANICAL PULPING BACKGROUND s LITERATURE ABOUNDS WITH PROPERTY, QUALITY AND PERFORMANCE MODELS, AND DATA e DIFFICULT TO INCORPORATE ALL FEATURES CONSENSUS POINTS TO FOLLOWING STRUCTURE PROCESS ATTRIBUTES BASIC PERFORMANCE ATTRIBUTES OF FIBERS PERFORMANCE ATTRIBUTES OF PULP AND PAPER X1 - PAT1- X2 PAT2 PROP2 X3 PAT3 xm PAT, (LESS FREQUENTLY MEASURED) PROP n (GENERALLY MEASURED)

7 -6- BACKGROUND PROP'S CORRELATE WITH EACH OTHER PAT's MAY BE INDEPENDENT OR DEPENDENT NOT ALL X's ARE CONTROLLABLE EXAMPLES X's o PAT's o PROP'S PULP SPECIES, REFINING CONDITIONS, SCREENING, CLEANING DESIGN, PRE- TREATMENT AND BLEACHING CONDITIONS SIZE DISTRIBUTION, SHAPES, SPECIFIC SURFACE DEVELOPMENT, TENSILE PRO- PERTIES, CHEMICAL PROPERTIES HANDSHEET PROPERTIES: BULK, TEAR, BURST, TENSILE, ROUGH- NESS, BRIGHTNESS, BREAKING LENGTH PULP PROPERTIES: VISCOSITY, YIELD, WET-WEB STRENGTH, FREENESS MECHANICAL PULPING STATUS o FOUR MECHANICAL PULPING MODULES DEVELOPED AND TESTED o FLOW SHEET DEVELOPED AND TESTED o PERFORMANCE ATTRIBUTES INCORPORATED AS INTEGRAL FEATURE OF MP SIMULATION o FLOW SHEET OPTIMIZATION PERFORMED o SEVERAL PRELIMINARY PULP AND PAPER PROPERTY MODELS INTRODUCED

8 -7- MECHANICAL PULPING REFINER MODULE SCHEMATIC SPECIFIC POWER PARAMETERS STOCK STREAM CHIPS FIBERS SHIVES FINES WATER STEAM T HYRFN1 T I POWER HEAT REFINED STOCK STREAM FIBERS SHIVES FINES WATER PERFORMANCE LOSS ATTRIBUTE 1 23 PERFORMANCE ATTRIBUTES MECHANICAL PULPING REFINER MODULE INPUT PARAMETERS ACCOUNT FOR - WOOD SPECIES - FIBER PROPERTIES - REFINER DESIGN - PLATE OR STONE PATTERN - PRETREATMENT CONDITIONS

9 -8- MECHANICAL PULPING REFINERS CALCULATE * CHANGE IN FIBER DISTRIBUTIONS * FIBER FLOWS * OUTLET CONSISTENCY AND TEMPERATURE * PERFORMANCE ATTRIBUTES - LENGTH WEIGHT AVERAGE STANDARD DEVIATION - WIDTH NUMBER AND WEIGHT AVERAGE STANDARD DEVIATIONS - SURFACE AREA PARAMETER (STRAND-EDWARDS K-FACTOR - CANADIAN STANDARD FREENESS * DISTRIBUTION 10 VALUES * PULP AND PAPER PROPERTIES - BULK - TEAR FACTOR - BREAKING LENGTH - BURST INDEX - WET-WEB STRENGTH

10 -9- MECHANICAL PULPING SIMULATION STOCK MIXER MIXES STREAMS AND PERFORMANCE ATTRIBUTES (PAT's) PERFORMANCE ATTRIBUTE MIXING FIBER DIMENSIONS AND SPECIFIC SURFACES ASSUMED CONSERVED DURING MIXING GRAVITY MIXED BASED ON CONSERVATION * CSF MODELED ON K-FACTOR (SPECIFIC SURFACE AREA) * ABSORPTION COEFFICIENT Ck WT-AVERAGED o KAPPA NO., FIBER TENSILE PROPERTIES NOT YET COMPUTED * CURL INDEX (LATENCY) UNDER DEVELOPMENT

11 -10- MECHANICAL PULPING FIBER SEPARATION AND CONSISTENCY CONTROL - PRESSURE SCREEN - CENTRICLEANER - THICKENER PARAMETERS FIBER STREAM + PAT'S HYFRAC ACCEPTS > FIBER STREAM + PAT'S REJECTS FIBER STREAM + PAT's THICKENER: REJECTS = WATER + DISSOLVED COMPONENTS ACCEPTS = SPECIFIED CONSISTENCY SCREEN AND CENTRICLEANER: REJECT PROBABILITY, Pij Pij = f(flow SPLIT, Li, Wj, 2 PARAMETERS) CENTRICLEANER: DIRT GOES TO REJECTS

12 -11- BACKGROUND HYDROGEN PEROXIDE BLEACHING * LIGNIN PRESERVING (2-3% LOSS) REACTIONS * OXIDATIVE ACTION OF PEROXIDE ANION, HO 2- * VARIOUS CHROMOPHORES CONTRIBUTE TO COLOR * CARBOXYLIC ACIDS FORMED DECREASE ph * CHROMOPHORES FORMED AT HIGH ph * STABILIZED BY CHELATING AGENTS * IONIC STRENGTH HAS LITTLE EFFECT * BRIGHTNESS INCLUDES ABSORPTION (k) AND SCATTERING (S) * ABSORPTION COEFFICIENT MORE FUNDAMENTAL - VARIES WITH WAVELENGTH - KINETICS BASED ON R 4 57 EQUILIBRIA PEROXIDE BLEACHING OH- + H H0 2 + H 2 0 PEROXIDE H H + + OH- WATER SiO 2 + H Si(OH) H- BUFFER

13 -12- HYDROGEN PEROXIDE BLEACHING MODULE SCHEMATIC Cko VOLUME ph BLEACHING CHEMICALS HYPROX > BLEACHED PULP + PAT's UNBLEACHED PULP + PAT's MODEL BASED ON S. MOLDENIUS, ROYAL INSTITUTE MODULE CALCULATES o CHANGE IN ABSORPTION COEFFICIENT, Ck PEROXIDE CONSUMPTION o ph CHANGE - NEUTRALIZATION o YIELD LOSS o SHIVE REDUCTION o REACTION TIME o BLEACHED PULP PROPERTIES - TENSILE STRENGTH - DENSITY - SCOTT BOND - TEAR INDEX - ROUGHNESS

14 -13- PEROXIDE BLEACHING KINETICS (MOLDENIUS) -dck= k [Ck] a [OH-]b [H ]tc Ck = Cko t = 0 a = 2.2 b = 0.23 c = 0.69 MAY BE CHANGED BY USER k = fl (pohi) A(CONSISTENCY) EXP(-E/RT) OPTIMAL CONSISTENCY: 10-11% OPTIMAL pohi: -11 Cko = f(species, QUALITY) EQUATIONS INTEGRATED OVER PLUG FLOW REACTOR MECHANICAL PULPING FLOW SHEET o PRIMARY REFINER * SECONDARY REFINER * REJECT REFINER * PRIMARY SCREENS * PRIMARY CENTRICLEANERS * REJECT SCREENING * REJECT CENTRICLEANING * CONSISTENCY CONTROL * PEROXIDE BLEACHING - SINGLE STAGE /

15 CENTRICLEANING HYDROGEN PEROXIDE BLEACHING MAPPS FLOWSHEET HIGH YIELD PULPING/REFINING/SCREENING/ CLEANING/BLEACHING PROCESS REJECT REFINING DILUTION REJECT SCREENING

16 Mapping of 100 Internal Fiber Types to 9 Fiber Stream Components wt. fraction array wt. fraction array * chips and fibers do not appear together in the stream fiber flows FINES FIBER 1 FIBER 2 FIBER 3 FIBER 4 FIBER 5 SHIVE 1 SHIVE 2 SHIVE 3 CHIPS *

17 -16- PERFORMANCE ATTRIBUTES SURFACE AREA DEVELOPMENT PARAMETER - RELATED TO FORMATION, BONDING, TENSILE PROPERTIES - MODELED BY STRAND-EDWARDS K-FACTOR CHIP REFINER K = f(consistency, SPECIFIC POWER) SECONDARY REFINER K = f(k o, CONSISTENCY, SPECIFIC POWER) MIXING SPLITTING K BASED ON CONSERVATION OF SPECIFIC SURFACE SPECIFIC SURFACE = 1 C Xi kn(li/2.4) K FREENESS MODEL CSF EXP ATOT (ml) CSf= ^ (ml) ATOT = SPECIFIC SURFACE (m 2 /g) o RELATIONSHIP DUE TO STRAND AND EDWARDS BASED ON DATA BY STATIONWALA AND ATACK o APPLIES TO REFINER, SPLITTING AND MIXING

18 -17- PERFORMANCE ATTRIBUTE SUBSTREAM PATs STREAM NO. TYPE KAPPA NO. W w K-FACTOR CSF Ck CURL SP. GR. TENSILE STRENGTH MODULUS CELL WALL THICKNESS I YIELD FIBER LENGTH DISTRIBUTION FIBER WIDTH DISTRIBUTION SURFACE AREA DEVELOPMENT SURFACE AREA ABSORPTION COEFFICIENT LATENCY FIBER PHYSICAL PROPERTIES

19 Mechanical Pulping Stock Mixing log-nomal distribution -(DL, DW) 1 [PAT 2 [PAT] n

20 Screening and Centricleaning Performance Attribute Calculation

21 -20- PHYSICAL PROPERTY MODELS PROPERTY = f(l 1, L 2, L 3 ) L 1 > 28 MESH 100 < L 2 < 28 MESH L 3 < 100 MESH BASED ON WORK OF GAREAU AND LAW REQUIREMENTS FOR EXPANDED USE OF PERFORMANCE ATTRIBUTES IN MAPPS * DATABASE o "STRUCTURE" o MODELS, RULES - TRANSFORMING - MIXING - SPLITTING o USER INTERFACE

22 Simulating Performance Attributes With MAPPS Expert N Product MAPPS Simulation Pretreatment Additives, Coatings Paper Machine History

23 -22- EFFECT OF SEPARATION PROCESSES ON PULP AND PAPER PROPERTIES PRIMARY SCREEN - TOTAL FLOW SPLIT 0.35 PROPERTIES ENTERING FIBER STREAM ACCEPT STREAM REJECT STREAM CONSISTENCY, % CSF, ml SPECIFIC SURFACE, m 2 /g BULK, cm 3 /g WET-WEB STRENGTH, g TEAR INDEX BURST INDEX BREAKING LENGTH, km

24 -23- EFFECT OF SEPARATION PROCESSES ON FIBER SIZE DISTRIBUTIONS PRIMARY SCREEN - TOTAL FLOW SPLIT 0.35 FIBER SPLIT 0.44 ENTERING FIBER STREAM (FLOWS IN LB/HR) FINES 24 ACCEPT FIBER STREAM 24 REJECT FIBER STREAM 0 FIBER SHIVES O O PERFORMANCE ATTRIBUTES L, mm OL W, mm aw K-factor CSF, ml

25 -24- EFFECT OF SEPARATION PROCESSES ON FIBER SIZE DISTRIBUTIONS PRIMARY CENTRICLEANER - TOTAL FLOW SPLIT 0.35 FIBER SPLIT 0.52 FINES FIBER SHIVES PERFORMANCE ATTRIBUTES ENTERING FIBER STREAM (LB/HR) ACCEPT FIBERS REJECT FIBERS L, mm al K-factor CSF, ml

26 -25- EFFECT OF SEPARATION PROCESSES ON PULP AND PAPER PROPERTIES PRIMARY CENTRICLEANER - TOTAL FLOW SPLIT PULP SPLIT 0.52 ENTERING FIBER ACCEPT PROPERTIES STREAM STREAM CONISISTENCY, % CSF ml SPECIFIC SURFACE AREA, m 2 /g REJECT STREAM BULK, cm 3 /g WET-WEB STRENGTH, g TEAR INDEX BURST BREAKING LENGTH, km

27 -26- EFFECT OF FIBER MIXING ON PULP AND PAPER PROPERTIES MIXING THREE ACCEPT STREAMS REJECTS PRIMARY CENTRICLEANERS CENTRICLEANING NO. 1 NO. 2 MIXTURE TO BLEACHING PERFORMANCE ATTRIBUTES L, mm al i, mm ow K-factor CSF, ml PROPERTIES SPECIFIC SURFACE, m2/g BULK, cm 3 /g WET-WEB STRENGTH, g TEAR INDEX BURST FACTOR BREAKING LENGTH, km

28 -27- CONCLUSIONS MECHANICAL PULPING * MODULES ARE FLEXIBLE AND ROBUST * PROCESSING DIFFERENCES ACCOUNTED FOR BY MODULE PARAMETERS * STRUCTURE IN PLACE FOR DETAILED SIMULATION PERFORMANCE ATTRIBUTES MODELING * PAT MODELING SHOWS PROMISE * NEED MORE STRUCTURAL CHANGES TO MAPPS OR ADDITIONAL TOOLS DOWN THE ROAD * ADDITIONAL PROPERTY MODELING SHOULD BE LEFT TO THE USER MAPPS EXPANDED CAPABILITIES IN PHYSICAL PROPERTIES CALCULATIONS - USE OF A PROPERTY INTERFACE - LIBRARY OF PROPERTY ROUTINES - USE OF UTILITY ROUTINES ("TOOLS") BENEFITS ARE - EXPANDED CAPABILITIES - SIMPLIFY MODULE PROGRAMMING - SIMPLIFY TROUBLESHOOTING

29 THE PROPERTY INTERFACE PROCESS MODEL 1 1 PROPERTY INTERFACE (PROPRT) "I PROPERTY ROUTINE Program Flow Data Flow

30 -29- PROPERTY ROUTINES THERM1 STEAM MASSFR MOLEFR PMOLWT DENGAS DENLIQ VISLIQ THERM1 * STREAM THERMODYNAMIC PROPERTIES ALL STREAMS - TOTAL FLOW - HEAT CAPACITY - ENTHALPY - EXERGY - TEMPERATURE WATER (STEAM) STREAMS - SPECIFIC ENTHALPY - QUALITY - PRESSURE STEAM * PROPERTIES OF WATER/STEAM - PRESSURE - QUALITY - TEMPERATURE - SPECIFIC VOLUME - SPECIFIC ENTHALPY - SPECIFIC ENTROPY - HEAT CAPACITY

31 -30- o COMPUTE GASEOUS - IDEAL DENGAS DENSITY OF A STREAM GAS LAWS DENLIQ o COMPUTE DENSITY OF A LIQUID STREAM VISLIQ o COMPUTE VISCOSITY OF A LIQUID STREAM o CONVERT TO MASS MASSFR STREAM COMPONENT FLOWS FRACTIONS OF TOTAL FLOW o CONVERT TO MOLE MOLEFR STREAM COMPONENT FLOWS FRACTIONS PMOLWT RETURN MOLECULAR WEIGHT OF A STREAM COMPONENT

32 -31- UTILITY ROUTINES (TOOLS) PROPRT PHASE UCONSY STRMAN UCOPYI UCOPYO PHASE * CLASSIFY EVERY COMPONENT OF A GIVEN STREAM TYPE AS: - DISSOLVED/SUSPENDED - NONVOLATILE/VOLATILE - NONFIBROUS/FIBROUS - INORGANIC/ORGANIC

33 -32- UCONSY * COMPUTE PERCENT DISSOLVED SOLIDS AND PERCENT CONSISTENCY FOR A GIVEN STREAM STRMAN * COMPUTE ELEMENTAL ANALYSIS OF GIVEN STREAM BASED ON 15 ELEMENTS UCOPYI SN SI SO PROPIN UCOPYO PROPOT > SN SO

34 -33- OPTIMIZATION OBJECTIVE DETERMINE FEASIBILITY OF PERFORMING OPTIMIZATION WITH MAPPS - USE A "DIFFICULT" PROBLEM - COMPARE RESULTS TO SIMULTANEOUS APPROACH (SACDA-MASSBAL) BACKGROUND OPTIMIZATION OPTIMIZATION: "IMPROVEMENT" FINDING A LOCAL OR GLOBAL "BEST" SOLUTION * APPLICATIONS - DATA RECONCILIATION (TYPE 1) - OPTIMIZING PROCESS CONDITIONS (TYPE 2) - PROCESS DESIGN (TYPE 3) * METHOD DETERMINE PROCESS VARIABLES SUCH THAT MINIMIZE Y OR [MAX (-Y)] Y = OBJECTIVE FUNCTION G(VARIABLES) > 0 INEQUALITY CONSTRAINTS H(VARIABLES) = 0 EQUALITY CONSTRAINTS LOWER BOUND < VARIABLES < UPPER BOUND

35 -34- OPTIMIZATION BACKGROUND * OPTIM PACKAGE - DEVELOPED BY KEN SAFFRAN AT IPC, TESTED OPTIM WITH EARLY VERSION OF MAPPS BLACK LIQUOR OXIDATION AQ AND KRAFT PULPING AND RECOVERY COUNTER-CURRENT BROWN STOCK WASHING OBJECTIVE FUNCTION = OPERATING PROFIT BACKGROUND OPTIMIZATION * SHOWED MAPPS + OPTIM WOULD WORK * DEVELOPED AND EVALUATED SEVERAL INTERFACES - DIRECT INTERFACE CHOSEN AS "BEST" * DETERMINED "BEST" OPTIMIZATION ALGORITHMS OF MANY - GRIFFITH-STEWART - SEQUENTIAL LINEAR PROGRAMMING (SLP) - OPT - REDUCED GRADIENT SEARCH STATUS o SELECTED TEST PROBLEM (TYPE 1) - MULTIPLE EFFECT EVAPORATOR SYSTEM INCLUDING SPLIT EFFECTS - DATA RECONCILED WITH MASSBAL - MEASUREMENTS AND FLOW SHEET * DEVELOPED CONVERGED MAPPS FLOW SHEET - SINGLE EFFECT EVAPORATOR MODULE EVAP02 - HEATER, HEAT EXCHANGERS, FLASHES, SPLITTERS, AND MIXERS o RAN SENSITIVITY STUDIES o SET UP OPTIMIZATION PACKAGE

36 -35- * SPECIFY OBJECTIVE FUNCTION OF REGRESSION FROM XM i = MEASUREMENTS X i = MAPPS VALUES W i = WEIGHTING FACTORS * DETERMINE X i TO MIN Y * SATISFY ALL CONSTRAINTS BY MAPPS - HEAT AND MATERIAL BALANCE - PHYSICAL CONSTRAINTS METHODOLOGY OPTIMIZATION * START WITH CONVERGED FLOW SHEET DATA OF 1/19/84 (RELATIVELY CLEAN TUBES) * OPTIMIZE AGAINST DATA OF 11/13/84 (PRIOR TO BOIL OUT) o POOR STARTING POINT ALSO CHOSEN * ADJUST WEIGHTING FACTORS

37 -36- OPTIMIZATION RESULTS e OPTIMIZATION MADE SUCCESSFULLY o RESULTS IN GOOD AGREEMENT WITH SACDA * MINOR DIFFERENCES DUE TO FOLLOWING BPR CORRELATION STEAM SUPERHEAT LOSS BL COMPOSITION HORTON CONCENTRATORS WEIGHTING FACTORS VARIABLES (INDEPENDENT AND DEPENDENT) VARY MAPPS IPC DATA AT MODULE INLET HEAT CAPACITY EVAP02 + HEATER ASSUMED 36 COMPONENT FLOWS AT MODULE OUTLET SINGLE EFFECT NOT KNOWN 50 MASSBAL TOTAL FLOW + CONCENTRATION RUN TIMES - EVAPORATOR CASE 161 SIMULATIONS 23 MINUTES PROCESSOR TIME

38 -37- OPTIMIZATION APPLICATION TO DESIGN OF MECHANICAL PULPING PROCESSES * OBJECTIVE FUNCTION - DETERMINE SPECIFIC POWER IN THREE REFINERS TO - MINIMIZE TOTAL SHIVES IN BLEACHED PULP * SHIVES REDUCED FROM 22.9% TO 0.76% BY ADJUSTING SPECIFIC POWER FOR THREE REFINERS - LONG FIBER SPECIES CASE CONCLUSIONS OPTIMIZATION o OPTIMIZATION IS FEASIBLE WITH MAPPS * CODING IS SIMPLE * RUN TIMES NOT EXCESSIVE * PROVIDES POWERFUL NEW FEATURE FOR MAPPS PACKAGE * NEED TO DEVELOP OPTIMIZATION ALGORITHMS IN-HOUSE /

39 -38- FUTURE DEVELOPMENTS PHYSICAL PROPERTY MODELS DEVELOP OPTIMIZATION CODE CONTINUE MODULE DEVELOPMENT FUTURE DEVELOPMENTS (UMAPPS) USER FRIENDLY INTERFACE 640 K BARRIER USER GROUP DEFINED MODULES FUTURE DEVELOPMENTS (STUDENTS) EVAPORATOR MODULES CONVERGENCE ACCELERATION CALCULATION ORDERING PHYSICAL PROPERTIES BOILING POINT RISE CALCULATION NONIDEAL GAS PROPERTIES ELECTROLYTE SOLUTION BEHAVIOR