Analysis and Modelling of the Energy Consumption of Chemical Batch Plants

8/10/2019 Analysis and Modelling of the Energy Consumption of Chemical Batch Plants

1/238


2/238

Diss. ETH No. 15532

Analysis and Modelling of the Energy Consumption

of Chemical Batch Plants

A dissertation submitted to the

SWISS FEDERAL INSTITUTE OF TECHNOLOGY

ZURICH

for the degree of

Doctor of Technical Sciences

Presented by

PATRIC S. BIELER

Dipl. Chem.-Ing. ETH

Dipl. NDS ETHZ in Betriebswissenschaften

born July 28, 1975

citizen of Luzern (LU) and Giswil (OW)

accepted on the recommendation of

Prof. Dr. K. Hungerbhler, examiner

Prof. Dr. D. T. Spreng, co-examiner

Prof. Dr. A. Wokaun, co-examiner

Dr. U. Fischer, co-examiner

Zurich 2004


3/238

ISBN 3-906734-39-0


4/238

Dedicated to my parents

i


5/238


6/238

AcknowledgementsThis thesis is based on research performed at the Safety and Environmental

Technology Group of the Swiss Federal Institute of Technology (ETH) in Zu-rich between June 2000 and April 2004. The funding rose by the Swiss FederalOffice of Energy (SFOE) (Project No. 39592) is gratefully acknowledged.

First, I would like to thank my supervisor Prof. Dr. K. Hungerbhler for giv-ing me the possibility to perform my research in his group. Moreover, he kindlysupported my decision of conducting a postgraduate study in industrial scienceduring my dissertation.

Thanks should go to Prof. Dr. A. Wokaun and to Prof. Dr. D. T. Spreng fortheir assistance as co-examiners and for the interesting discussions and chal-lenging questions concerning my work throughout the time of my thesis.

Dr. U. Fischer receives a special thank you for advising and managing my

project. I will never forget all the good discussions not only in the field of mythesis but also concerning topics from all over the world and his marvelloussupport during the whole time of my dissertation.

I want to thank the industrial partner company, that I can not name for con-fidentiality reasons, for enabling me to perform my dissertation in a challengingindustrial environment and supporting me in all the work I had to do, the ques-tions I had to pose and the measurements I had to perform. A great thankshould also go to the many people working at the company. I had a very goodand enriching time in the industrial environment. I will never forget all the in-teresting discussions with the site and the plant manager, the foremen, the op-erators, the people in the service teams, the production chemists, the engineers

and all the people in my environment. Many people supported me with tips ofhow to conduct the measurements and gather the required information for thisI would like thank a lot everybody. A special thank should be addressed to the

people that conducted autonomously several measurements for supporting me inmy work.

A special thank is also going to Dr. Chr. Blickenstorfer, who started the pro-ject of energy modelling of chemical batch plants in 1996 within the Safety andEnvironmental Technology Group and to Dipl. Chem.-Ing. ETH D. Dahinden,who carried out his diploma thesis within this project. He made major contribu-tions to the Excel-programming resulting in the Excel-model of the whole

plant.I do not want to miss to thank all the people from the Safety and Environ-

mental Technology Group at ETH for the good time we had first in the goodold CAB building in the centre of the beautiful little big-city of Zurich and af-terwards in the HCI building on top of Hnggerberg.

My greatest thank, nevertheless, goes to my parents, I. and P. S. Bieler-Schmed who supported me during all the intense time of my studies in Chemi-cal Engineering and Industrial Management and during my dissertation. With-out their support and their care I would not have been able to perform this bigworkload and to succeed in my education.

Finally, yet importantly, I would like to thank my girlfriend M. Oeschger for

her support and her patience during the final phase of my thesis.

iii


7/238


8/238

Engineering is the science of economy, of conserving the energy, kinetic andpotential, provided and stored up by nature for the use of man. It is the businessof engineering to utilize this energy to the best advantage, so that there may bethe least possible waste.

William A. Smith, 1908

v


9/238


10/238

Table of Contents

1 Introduction 11.1 Context and Motivation......................................................................11.2 State of the Art....................................................................................21.3 Thesis Statement and Contribution ....................................................31.4 Thesis Organization............................................................................4

2 Structure of a Batch Plant 7

2.1 The General Structure of Batch Plants ...............................................72.2 The Differences between Batch Plants.............................................10

2.2.1 The Monoproduct Batch Plant..............................................10

2.2.2 The Multiproduct Batch Plant ..............................................102.2.3 The Multipurpose Batch Plant..............................................11

3 Two Approaches for Energy Modelling 13

3.1 The Top-Down Approach.................................................................133.1.1 The Model for the Production Dependent Energy

Consumption.........................................................................133.1.2 The Heating Steam Model....................................................14

3.2 The Bottom-Up Approach................................................................153.2.1 Equations for Heating and Cooling of Substances...............153.2.2 Equations for Electric Equipment ........................................16

3.2.3 Unified Equation for the Bottom-Up Modelling..................174 Top-Down Modelling of Production Plants (TODOMO) 23

4.1 The Basic Equation for the Top-Down Modelling...........................234.2 The Characteristics of the Different Buildings Investigated ............234.3 Analysis of the Different Energy Carriers........................................24

4.3.1 Steam ....................................................................................254.3.2 Electricity .............................................................................314.3.3 Cooling Energy.....................................................................34

4.4 Applicability of the Models..............................................................364.5 Conclusions ......................................................................................38

5 Modelling of Single Unit Operations 41

5.1 Reactors ............................................................................................415.1.1 Description of the Equipment...............................................415.1.2 Measurements.......................................................................425.1.3 Model and Conclusions ........................................................45

5.2 Nutsche Dryer...................................................................................545.2.1 Description of the Equipment...............................................545.2.2 Measurements.......................................................................555.2.3 Model and Conclusions ........................................................55

vii


11/238

5.3 Heat-Chamber.................................................................................. 585.3.1 Description of the Equipment .............................................. 585.3.2 Measurements ...................................................................... 585.3.3 Model and Conclusions........................................................ 58

5.4 Vacuum Pumps ................................................................................ 615.4.1 General Vacuum Pumps ...................................................... 615.4.2 Anti Pollution Vacuum Pumps (APOVAC) ........................ 625.4.3 Steam-Jet Vacuum Pumps ................................................... 65

5.5 Stirrers and Motors .......................................................................... 665.5.1 Description of the Equipment .............................................. 665.5.2 Measurements ...................................................................... 675.5.3 Model and Conclusions........................................................ 71

5.6 Continuous Equipment..................................................................... 725.6.1 Infrastructure........................................................................ 72

5.6.2 Short-Path Distillation Column............................................ 735.6.3 Falling-Film Evaporator....................................................... 765.7 Horizontal Vacuum Rotary Dryer.................................................... 78

5.7.1 Description of the Equipment .............................................. 785.7.2 Measurements ...................................................................... 785.7.3 Model and Conclusions........................................................ 79

5.8 Batch Distillation Column ............................................................... 825.8.1 Description of the Equipment .............................................. 825.8.2 Measurements ...................................................................... 825.8.3 Model and Conclusions........................................................ 83

5.9 Centrifuge ........................................................................................ 86

5.9.1 Description of the Equipment .............................................. 865.9.2 Measurements ...................................................................... 865.9.3 Model and Conclusions........................................................ 86

5.10 Conclusions...................................................................................... 88

6 Bottom-Up Modelling of Multipurpose Batch Plants 91

6.1 Combining the Different Unit Operation Models to a PlantModel (BOTUMO) .......................................................................... 916.1.1 Description of the Program for Modelling Multipurpose

Batch Plants ......................................................................... 916.1.2 Modelling and Report Generation ....................................... 93

6.2 Results of the BOTUMO ................................................................. 956.2.1 Modelling of Different Periods............................................ 956.2.2 Analysis of the Energy Consumption of the Building ....... 1006.2.3 Modelling of Different Aspects of the Reactors and

Nutsche Dryers................................................................... 1046.2.4 The Differences between the Products .............................. 1106.2.5 The Differences between the Apparatus............................ 116

6.3 Sensitivity Analysis of the BOTUMO........................................... 1186.3.1 Time ................................................................................... 1196.3.2 Steam Heat Transfer .......................................................... 1206.3.3 Brine Heat Transfer............................................................ 121

6.3.4 Condensation Enthalpy of Steam....................................... 122

viii


12/238

6.3.5 Stirrer Input to the Reaction Vessels..................................1236.3.6 Stirrer Electricity Consumption..........................................1246.3.7 Circulation Pump................................................................1256.3.8 Vacuum Pumps...................................................................126

6.3.9 APOVAC............................................................................1276.3.10 Short Path Distillation ........................................................128

6.4 Conclusions ....................................................................................129

7 Conclusions and Outlook 131

7.1 Conclusions ....................................................................................1317.2 Outlook...........................................................................................136

Appendix I

A The Model I

A.1 The Assumptions for the BOTUMO ...................................................IA.1.1 The Assumptions for the Single Unit Operation Models ........IA.1.2 The Assumptions for the Plant Model................................... II

A.2 The ExcelModel ............................................................................ IIIA.2.1 The Sheets of the Program ................................................... IIIA.2.2 Description of the Required Input Data.............................. VII

A.3 The Results of the Model ..................................................................XA.3.1 Modelling Results..................................................................XA.3.2 Sensitivity Analysis ...........................................................XXI

B The Measuring Equipment XXIII

B.1 Steam Measurements.................................................................. XXIIIB.1.1 The Equipment ...............................................................XXIIIB.1.2 The Accuracy..................................................................XXIV

B.2 Measurement of Brine ................................................................XXIVB.2.1 The Equipment ...............................................................XXIVB.2.2 The Accuracy................................................................... XXV

B.3 Measurement of Electricity Consumption...............................XXVIIIB.3.1 The Equipment ............................................................ XXVIIIB.3.2 The Accuracy............................................................... XXVIII

C Miscellaneous XXIX

C.1 Distributions of the Times Given in the PSP..............................XXIXC.2 Reflux Conditions.......................................................................XXXIC.3 Investigations on the Cleaning of Vessels............................... XXXIII

D Measurements XXXVII

D.1 Measurements for the TODOMO...........................................XXXVIID.2 Measurements for the BOTUMO................................................XLIII

E Improvement Potentials for the Investigated Plant LV

F Glossary LIX

ix


13/238


14/238

List of Figures

Figure 1-1: Structure of the thesis ........................................................................5

Figure 2-1: Value chain in the chemical industry (shaded: typical batchprocesses) ........................................................................................7

Figure 2-2: Engraving of a 16thcentury gold processing plant (Stitt 2002).........8Figure 2-3: General structure of a batch plant ......................................................9Figure 3-1: The basic concept of the BOTUMO................................................18Figure 3-2: The principle of the BOTUMO .......................................................19Figure 4-1: Consumption of production steam (5 and 15 bar) of the different

buildings as a function of amount of products per month(according to Equation (3-1)) ........................................................26

Figure 4-2: Consumption of heating steam (5 bar) as a function of degree-days per month (according to Equation (3-2)) ..............................29

Figure 4-3: Normalized heating steam consumption (5 bar) as a function ofthe number of degree-days per month (according toEquation (3-3)) ..............................................................................30

Figure 4-4: Electricity consumption (excluding electricity for coolingpurposes) of the investigated buildings as a function of theamount of chemicals produced per month (according toEquation (3-1)) ..............................................................................32

Figure 4-5: Hourly electricity consumption of Building 1 (without electricheating of a specific process) during an ordinary week in 2001 ...33

Figure 4-6: Consumption of cooling energy of the different buildings as a

function of production output per month (according toEquation (3-1)) ..............................................................................35Figure 4-7: Modelled monthly electricity consumption as a function of

capacity usage for those buildings where the model accordingto Equation (3-1) was suitable .......................................................37

Figure 4-8: Flowchart for energy analysis in chemical batch production..........38Figure 5-1: Scheme of a standard batch vessel with its heating/cooling-

system............................................................................................42Figure 5-2: Example of the steam measurements for a 10 m3, glass lined

reaction vessel heated with 5 bar steam ........................................43Figure 5-3: Example of brine measurements for a 10 m3stainless steel vessel.44

Figure 5-4: Measurements of the electric heating of the 4 m3

stainless steelhigh-temperature reaction vessel...................................................45

Figure 5-5: Modelling and measurements of the steam consumption ofreaction vessels..............................................................................47

Figure 5-6: Modelling results of the steam consumption of a 10 m3stainlesssteel reaction vessel (in comparison with measured steamconsumption and reaction time) ....................................................48

Figure 5-7: Measurements of the brine consumption of a 10 m3stainlesssteel vessel (regression according to Equation (3-16))..................50

Figure 5-8: Modelling of the brine consumption of a 10 m3stainless steelvessel (according to Equation (3-16); in comparison with

measured steam consumption) ......................................................51

xi


15/238

Figure 5-9: Modelling of the brine consumption (according toEquation (3-16)) vs. measurements .............................................. 52

Figure 5-10: Scheme of a nutsche dryer with its heating/cooling-system......... 54Figure 5-11: Modelling according to Equation (5-1) and measurements of

the nutsche dryer with simultaneous heating and cooling ............ 56Figure 5-12: Modelling according to Equation (5-1) of the drying of

product 1 in a 10 m2nutsche dryer (in comparison withmeasured steam consumption and drying time)............................ 57

Figure 5-13: Scheme of a heat-chamber ............................................................ 58Figure 5-14: Measured and modelled (according to Equation (5-2)) steam

consumption and experiment duration for the heat-chamber ....... 59Figure 5-15: Measured vs. modelled steam consumption of the heat-chamber

(according to Equation (5-2))........................................................ 60Figure 5-16: Typical measurement of the electricity consumption of a

vacuum pump (here: PN= 16.5 kW) ............................................ 61

Figure 5-17: Measurements of the electricity consumption of the APOVACpumps (PN= 27 kW)..................................................................... 63

Figure 5-18: Measurements of the cooling media consumption of theAPOVAC pumps (calculation of the cooling energyconsumption according to Equation (3-5)) ................................... 64

Figure 5-19: Principle of a steam-jet vacuum pump (El-Dessouky et al.2002) ............................................................................................. 65

Figure 5-20: Power consumption (P) and temperature of the reaction mass(IT) for a 6.3 m3stainless steel vessel with an Intermig stirrer..... 68

Figure 5-21: Power consumption (P) and temperature of the reaction mass

(IT) of a 6.3 m3

glass lined vessel with an Intermig stirrer........... 69Figure 5-22: Power consumption (P), rounds per minute, and temperature ofthe reaction mass (IT) of a 6.3 m3stainless steel vessel with aCross-Blade stirrer ........................................................................ 69

Figure 5-23: Measurements of the relation of power consumption to nominalpower P/PNof different stirrer types............................................ 70

Figure 5-24: Scheme of a short path distillation column................................... 73Figure 5-25: Measurements of the total electricity consumption of the short

path distillation column................................................................. 74Figure 5-26: Measured brine consumption of the short path distillation

column........................................................................................... 75

Figure 5-27: Scheme of a falling-film evaporator ............................................. 76Figure 5-28: Modelled energy consumption of a one-day operation of the

falling-film evaporator (according to Equation (5-1);parameters see Table 5-5) ............................................................. 77

Figure 5-29: Typical horizontal vacuum rotary dryer with agitator beinginstalled into shell (from (Mujumdar 1995)) ................................ 78

Figure 5-30: Steam measurements of a 4 m3horizontal vacuum rotary dryer .. 79Figure 5-31: Measured and calculated steam consumption for two 4 m3

horizontal vacuum rotary dryers (according to Equation (3-16)) . 80

xii


16/238

Figure 5-32: Modelled steam consumption for the 4 m3Dryer 1 dryingProduct 2 (calculated according to Equation (3-16); incomparison with measured steam consumption and dryingtime)...............................................................................................81

Figure 5-33: Scheme of a batch distillation column...........................................82Figure 5-34: Measured and calculated steam consumption for the

investigated batch distillation column (according toEquation (5-3)) ..............................................................................84

Figure 5-35: Modelling results of the batch distillation column (according toEquation (5-3)) in comparison with measured steamconsumption and distillation time .................................................85

Figure 5-36: Scheme of a centrifuge ..................................................................86Figure 6-1: The four layers of the program for modelling the energy

consumption of chemical batch plants ..........................................91

Figure 6-2: The different layers and the structure of the BOTUMO programand their contents...........................................................................92Figure 6-3: Modelling of the specific utility consumption (per t of product)

of the whole building for one day of production according toEquation (3-14) (in comparison with measured consumptionand modelled data according to CPM) ..........................................96

Figure 6-4: Modelling of the specific utility consumption (per t of product)of the investigated building for one month of productionaccording to Equation (3-14) (in comparison with measuredconsumption and modelled data according to CPM) ....................98

Figure 6-5: Absolute modelled steam consumption of the building during

one month according to Equation (3-14) (PSP data)...................101Figure 6-6: Specific modelled steam consumption of the building during onemonth according to Equation (3-14) (PSP data)..........................102

Figure 6-7: Absolute modelled electricity and brine consumption of thebuilding during one month according to Equation (3-14)(PSP data) ....................................................................................103

Figure 6-8: Specific modelled electricity and brine consumption of thebuilding during one month according to Equation (3-14)(PSP data) ....................................................................................104

Figure 6-9: Modelled specific steam consumption of the reactors andnutsche dryers according to Equation (3-25) (PSP data).............106

Figure 6-10: Modelled specific electricity consumption for the reactors andnutsche dryers according to Equation (3-25) (PSP data).............107

Figure 6-11: Modelled specific brine consumption for the reactors andnutsche dryers according to Equation (3-25) (PSP data).............109

Figure 6-12: Specific modelled steam consumption of the different products(A, B,..,N, O) according to Equation (3-20) and number ofsynthesis steps (PSP data; modelling period: W = one week,M = one month) ...........................................................................111

xiii


17/238

Figure 6-13: Specific modelled electricity consumption of the differentproducts (A, B,..,N, O) according to Equation (3-20) andnumber of synthesis steps (PSP data; modelling period:W = one week, M = one month) ................................................. 113

Figure 6-14: Specific modelled brine consumption of the different products(A, B,..,G, I) according to Equation (3-20) and number ofsynthesis steps (PSP data; modelling period: W = one week,M = one month) .......................................................................... 114

Figure 6-15: Modelled specific steam consumption of the apparatus (1,2,..,26, 27) during one month according to Equation (3-16)(PSP data).................................................................................... 116

Figure 6-16: Modelled specific electricity consumption of the apparatus (1,2,..,26, 27) during one month according to Equation (3-16)(PSP data).................................................................................... 117

Figure 6-17: Modelled specific brine consumption of the apparatus(1, 2,..,26, 27) during one month according to Equation (3-16)(PSP data).................................................................................... 118

Figure 6-18: Sensitivity analysis of the batch time twith regard to thespecific utilities according to Equation (3-14) (one month;PSP data)..................................................................................... 119

Figure 6-19: Sensitivity analysis of the steam loss coefficient KStwithregard to the specific steam consumption according toEquation (3-14) (one month; PSP data) ...................................... 120

Figure 6-20: Sensitivity analysis of the brine loss coefficient KCowithregard to brine consumption according to Equation (3-14)

(one month; PSP data) ................................................................ 121Figure 6-21: Sensitivity analysis of the steam condensation enthalpy HVwith regard to steam consumption according toEquation (3-14) (one month; PSP data) ...................................... 122

Figure 6-22: Sensitivity analysis of the stirrer input with regard to utilityconsumption according to Equation (3-14) (one month;PSP data)..................................................................................... 123

Figure 6-23: Sensitivity analysis of the stirrer electricity consumptionwith regard to utility consumption according toEquation (3-14) (one month; PSP data) ...................................... 124

Figure 6-24: Sensitivity analysis of the circulation pump efficiency with

regard to electricity consumption according to Equation (3-14)(one month; PSP data) ................................................................ 125

Figure 6-25: Sensitivity analysis of the vacuum pump efficiency withregard to electricity consumption according to Equation (3-14)(one month; PSP data) ................................................................ 126

Figure 6-26: Sensitivity analysis of the APOVAC pumps efficiency withregard to electricity consumption according to Equation (3-14)(one month; PSP data) ................................................................ 127

Figure 6-27: Sensitivity analysis of the short path distillation motorsefficiency with regard to electricity consumption according

to Equation (3-14) (one month; PSP data).................................. 128

xiv


18/238

Figure 7-1: Analysis of the total modelled steam consumption of theinvestigated plant (period: one month; PSP data; totalconsumption: 1,354 MWh; heat of reaction: -80 MWh, stirrerinput: -23 MWh)..........................................................................133

Figure 7-2: Analysis of the total modelled electricity consumption of theinvestigated plant (period: one month; PSP data; totalconsumption: 315 MWh).............................................................134

Figure 7-3: Analysis of the total modelled brine consumption of theinvestigated plant (period: one month; PSP data; totalconsumption: 100 MWh).............................................................135

Figure A-1: Equations on the input sheetDatenin the Excelmodel ...............IIIFigure A-2: Equations on the sheetAuswertungin the Excelmodel...............IVFigure A-3: Equations on the sheetBerechnungenin the Excelmodel ............VFigure A-4: Equations on the sheetAufheizenin the Excelmodel ..................VIFigure A-5: Equations on the sheet Verdampfenin the Excelmodel...............VIFigure B-1: Measuring principle for the steam measurements ....................XXIIIFigure B-2: Scheme of the Portaflow X ultrasonic flow meter from

Fuji Electric43...........................................................................XXIVFigure B-3: Principle of the brine measurement ........................................... XXVFigure B-4: Picture of a LEM Memobox 800 ........................................... XXVIIIFigure C-1: Detailed investigations on production time distribution for

the production process (Steps 1, 2,.., 50, 51) of Product A intwo 4 m3and one 10 m3glass lined reaction vessels ..............XXIX

Figure C-2: Time measurements of Product J in a 6.3 m3glass-linedreactor ....................................................................................... XXX

Figure C-3: Time measurements of Product G in a 4 m3

glass-linedreactor ....................................................................................... XXXFigure C-4: Frequency of the measured times of reflux condition (i.e.,

30 min of reflux) in a 10 m3stainless steel vessel given inTable C-2 .............................................................................. XXXIII

Figure C-5: Measurements for a dirty 6.3 m3glass-lined reactor .............XXXIVFigure C-6: Measurements for the same clean 6.3 m3glass-lined reactor XXXIVFigure C-7: Modelling and measured values of the dirty and clean 6.3 m3

glass-lined reactor.................................................................. XXXVFigure C-8: Modelling results of the dirty and clean 6.3 m3glass-lined

reactor (in comparison with measured values and

experiment duration) ............................................................XXXVIFigure D-1: Efficiency of standard motors at different levels of power

consumption (BBC 1976).........................................................XLIII

xv


19/238


20/238

List of Tables

Table 2-1: Sections of a generic batch plant ........................................................9Table 4-1: Characteristics of the investigated buildings ....................................24Table 4-2: Summary of the different production energy consumption models

obtained for the different energy forms (m) in the differentbuildings according to Equation (3-1)...........................................28

Table 4-3: Summary of the models for heating steam consumption obtainedfor the different buildings according to the normalisedEquation (3-3)8 ..............................................................................31

Table 5-1: Calculated loss coefficients for the steam consumption of thereaction vessels and nutsche dryers investigated...........................49

Table 5-2: Loss coefficients for the brine measurements of the investigated

reaction vessels..............................................................................52Table 5-3: Different kind of stirrers used in the investigated building ..............67Table 5-4: Base Consumption of the investigated building ...............................73Table 5-5: Parameter values of the falling-film evaporator ...............................77Table 5-6: Standard sizes of centrifuges ............................................................87Table 5-7: Summary of the Equations and Parameters for the SUOM ..............88Table 6-1: Example of a generic PSP and its translation for the data input to

the program....................................................................................94Table 6-2: Investigated periods ..........................................................................95Table 6-3: Relative deviations of the different modelling methods for the

investigated utilities according to Equation (3-14) .......................98

Table 6-4: Comparison of Measurements, TODOMO results according toChapter 4 and BOTUMO results according to Chapter 6.1 forone month of normal production .................................................100

Table 6-5: Summary of the sensitivity analysis of Chapter 6.3 showing thedeviation of the objective functions Emaccording toEquation (3-14) for changes in the parameter values of 20%;modelling period: one month.......................................................130

Table A-1: Required input data for the Excelmodels (sheetDaten) ............. VIITable A-2: Required input data for the Excelmodels (sheetParameter).....VIIITable A-3: Required input data for the Excelmodels (sheet WK)................VIIITable A-4: Required input data for the Excelmodels (sheet VP-DS-Dest.

Kol.)............................................................................................VIIITable A-5: Required input data for the Excelmodels (sheetReaktionen).......IXTable A-6: Required input data for the Excelmodels (sheet Substanzen).......IXTable A-7: Required input data for the Excelmodels (sheet Gerte)..............IXTable A-8: Measurement and modelling of the utility consumption of the

investigated plant............................................................................XTable A-9: Modelling results for the total utility consumption of the

investigated building .....................................................................XITable A-10: Modelled steam consumption of one week of the reactors and

nutsche dryers of the investigated building...................................XI

xvii


21/238

Table A-11: Modelled specific steam consumption of one week of thereactors and nutsche dryers of the investigated building.............XII

Table A-12: Modelled electricity consumption of one week of the reactorsand nutsche dryers of the investigated building...........................XII

Table A-13: Modelled specific electricity consumption of one week of thereactors and nutsche dryers of the investigated building............XIII

Table A-14: Modelled brine consumption of one week of the reactors andnutsche dryers of the investigated building ................................XIII

Table A-15: Modelled specific brine consumption of one week of thereactors and nutsche dryers of the investigated building........... XIV

Table A-16: Modelled steam consumption of one month of the reactors andnutsche dryers of the investigated building ............................... XIV

Table A-17: Modelled specific steam consumption of one month of thereactors and nutsche dryers of the investigated building.............XV

Table A-18: Modelled electricity consumption of one month of the reactorsand nutsche dryers of the investigated building...........................XV

Table A-19: Modelled specific electricity consumption of one month of thereactors and nutsche dryers of the investigated building........... XVI

Table A-20: Modelled brine consumption of one month of the reactors andnutsche dryers of the investigated building ............................... XVI

Table A-21: Modelled specific brine consumption of one month of thereactors and nutsche dryers of the investigated building..........XVII

Table A-22: Percentage of utility consumption of the produced chemicals..XVIITable A-23: Modelled specific steam consumptions of the apparatus [kg/t]

for one month...........................................................................XVIII

Table A-24: Modelled specific electricity consumptions of the apparatus[kWh/t] for one month ............................................................... XIXTable A-25: Modelled specific brine consumptions of the apparatus [kWh/t]

for one month ...............................................................................XXTable A-26: Results of the different sensitivity analysis for one month ........ XXITable B-1: Test of the steam measurement device ...................................... XXIVTable B-2: Parameters for the flow measurements of the two kinds of

brine ........................................................................................ XXVITable B-3: Temperature comparison of the two temperature probes ......... XXVIITable C-1: Time investigations of PSP and measurements ......................... XXXITable C-2: Time measurements for distillation of 300 l of 1-butanol in a

10 m3

stainless steel vessel .................................................... XXXIITable C-3: Steam measurements for the cleaning investigations for a

6.3 m3glass lined reactor.......................................................XXXVTable D-1: Measurements of Building 1 ................................................. XXXVIITable D-2: Measurements of Building 2 ................................................XXXVIIITable D-3: Measurements of Building 3 ...................................................XXXIXTable D-4: Measurements of Building 4 .......................................................... XLTable D-5: Measurements of Building 5 .........................................................XLITable D-6: Measurements of Building 6 ....................................................... XLIITable D-7: Efficiencies of standard motors at different levels of power

consumption (BBC 1976) ........................................................XLIV

xviii


22/238

Table D-8: Measurements of the steam consumption of the reactionvessels........................................................................................XLV

Table D-9: Summary of the brine measurements.......................................XLVIIITable D-10: Batch times for the electric heating in a 4 m3stainless steel

reaction vessel (high temperature)........................................... XLIXTable D-11: Measurements with simultaneous heating and cooling in

10 m2nutsche dryers ............................................................... XLIXTable D-12: Measurements without simultaneous heating and cooling in

10 m2nutsche dryers ...................................................................... LTable D-13: Measured power consumption of different vacuum pumps............ LTable D-14: Steam and cooling water consumption of different steam-jet

vacuum pumps (four stages) according to (GEA.b ).....................LITable D-15: Summary of the Brine Measurements for the APOVAC pumps ... LITable D-16: Infrastructure Measurements of the investigated building............LIITable D-17: Measurements of the steam consumption of a batch distillation

column ..........................................................................................LIITable D-18: Steam measurements (15bar) for the high temperature reactor

(4 m3stainless steel reaction vessel) .......................................... LIIITable F-1: Definitions of the ISAS88.01-1995 standard for batch

production (ISA 1995)................................................................ LIX

xix


23/238


24/238

List of Abbreviations, Symbols, and Indexes

Abbreviations

APOVAC Anti POllution VACuumBOTUMO BOTtom-Up MOdelCPM Company Proprietary MethodF1 Flow meterF2 Flow meterM MonthP1 Pressure meterPR Production RecordPSP Process Step ProcedureSUOM Single Unit Operation Model

T1 Temperature meterT2 Temperature meterTAM Time Average ModelTODOMO TOp-DOwn MOdelTSM Time Slice ModelW Week

Symbols

A Surface Area [m2]ACR Air Change Rate [h-1]B Base consumption of energy [MWh / period]

C Constant [various]c Sound velocity [m / s]cP Heat capacity [kJ / kg / K]DD Degree-Days [C d]DSS Day-specific steam consumption [MWh / C d]E

Energy consumption [kWh / s]F Energy defining factor (0 for electricity, 1 for

steam and brine)[-]

K Loss coefficient [kW / m2/ K]IT Temperature of reaction mass [C] or [K]InT Inlet Temperature [C] or [K]m Mass [kg]OT Temperature of jacket [C] or [K]OuT Outlet Temperature [C] or [K]P Power [kW]PO Production Output [t / period]RR Reflux Ratio [-]S Specific energy consumption [MWh / t]SC Steam consumption [MWh / period]t Time [s / period]T Temperature [K]

H Enthalpy change [kJ / kg]

xxi


25/238

Actual to nominal power consumption of a motor [%]

Efficiency [%] Density [kg / m

3] Kinematic viscosity [m

2/ s]

Indexes

1, 2 Start- & EndpointA ApparatusAir AirAm AmbientB BarrelBC Batch ColumnBr BrakeC Crystallisation

Co CoolingES Evaporated SolventEl ElectricityF FeedFFE Falling Film EvaporatorHC Heat-ChamberHJ Heating JacketI Infrastructurei Chemicals type (different PSP)j Apparatus typek Number of different specifications of a chemical (PSP)

L LossM Meltingm Energy formN Nominaln Number of different specifications of a apparatusND Nutsche DryerO OperationP ProductionPu Pumpq Indicator for different process steps / unit operations of one recipe (PSP)R Reaction

RD Rotary DryerRM Reaction MassRV Reaction VesselS SolventSo SolidSPD Short Path Distillation columnSt SteamSu SuspensionV VaporisationW WaterZ Centrifuge

xxii


26/238

AbstractTwo different approaches for energy analysis and modelling of chemical

batch plants (a top-down model and a bottom-up model) were conducted in thisthesis. Steam, electricity and brine were the investigated utilities. Steam wasused for heating the reactors and the building. Electricity was used by diverseelectric equipment in the building. Brine was used for low-temperature coolingof the reaction vessels and the nutsche dryers (i.e., cooling below a starting tem-

perature of about 30 C).A top-down model (TODOMO) consisting of a linear equation based on the

specific energy consumption per ton of production output and the base con-sumption of the plant was postulated. This TODOMO showed to be applicablefor batch plants of the following kind:

- Monoproduct batch plants

-Multiproduct batch plants with constant production mix

- Multipurpose batch plants in which only similar chemicals are pro-duced

The results showed that the electricity consumption of infrastructure equip-ment was significant and responsible for about 50% of total electricityconsumption. Base consumptions for the steam and the brine system were onlyminor. The specific energy consumption for the different buildings was relatedto the degree of automation and the production processes performed.

For the heating steam, a model only depending on air change rate and de-gree-days was applicable.

For multipurpose batch plants with highly varying production processes andchanging production mix, the TODOMO was not applicable and produced inac-curate results. A bottom-up model (BOTUMO) was postulated for these plants.The model consists of a production dependent part and a production-independent part accounting for the infrastructure consumption. The productiondependent part actually consists of a term related to the chemicals, another termrelated to the equipment, and a time-dependent loss term.

With the help of numerous measurements, different apparatus and unit op-eration models were built. These models use only easily accessible substanceand apparatus information and account for the losses of the different apparatus.The models are therefore designed for being transferable to other batch plantsand products and not limited to one specific plant. The single apparatus modelsshowed that losses for steam and brine consumption are high. The losses werecharacterised by a so-called loss coefficient. This loss coefficient represents theheat transfer coefficient of the outside surface area of the equipment. For steamconsumption, a loss coefficient of about 4.210-2kW / m2/ K was found whilefor brine consumption a loss coefficient of about 1.710-2kW / m2/ K wasfound. More than 50% of the losses of the steam are therefore due to the heat-ing/cooling-system design with its steam traps.

With the help of the above-mentioned equations, an Excelmodel was builtfor the modelling of whole production plants according to the BOTUMO.

Modelling of the whole production plant was performed for one and two days,

xxiii


27/238

one week and one month. The production data were taken from either the pro-duction record (PR) or the process step procedure (PSP). The modelling re-sulted in a high accuracy for the longer periods (PSP data is used as input).

Analyses of the modelling results for one month showed that the apparatus

group reactors and nutsche dryersis the most important energy consumer in thebuilding (apart from infrastructure consumption in the case of electricity). Moredetailed analyses of the energy consumption of this apparatus group showed,that about 30 to 40% of steam energy are lost and thus large optimisation poten-tials are revealed. For the electricity consumption, it is shown that the small cir-culation pumps of the heating/cooling-system of the reactors and nutsche dryersrequire about 25% of the total electricity consumption of this apparatus group(i.e., ca. 50% of the consumption of the stirrers) if no electric heating is per-formed. Electric heating is used for one single high-temperature reactor. Theconsumption of this heating circuit is larger than the consumption of all otherstirrers in the building (over 25 stirrers).

A sensitivity analysis showed that the batch time has the largest influence onthe energy consumption. Variations of 50% in the batch time resulted inchanges in energy consumption of about 30%.

Different saving potentials, ranging from elimination of reflux conditions toinvention of a new heating/cooling-system for a generic batch reactor, wereidentified.

The applicability of the BOTUMO is shown for short and long modellingperiods using different types of input data. Transferability and applicability toother buildings and chemicals need to be investigated in further case studies.

xxiv


28/238

ZusammenfassungIn dieser Doktorarbeit wurden zwei verschiedene Arten der energetischen

Analyse und Modellierung (ein top-down Modell und ein bottom-up Modell)von chemischen Batch Produktionsanlagen entwickelt. Die untersuchten Ener-gien waren Dampf, Elektrizitt und Sole. Dampf wurde sowohl zur Erhitzungder Reaktoren und Nutschentrockner (und ihres Inhaltes), als auch zur Gebu-deheizung benutzt. Die Elektrizitt wurde von den verschiedenen elektrischenApparaturen der Gebude verbraucht. Die Sole wurde zu Tieftemperatur-Khlzwecken verwendet (d.h. Khlen unterhalb einer Starttemperatur von ca.30 C).

Ein top-down Modell (TODOMO), bestehend aus einer linearen Gleichung,basierend auf dem spezifischen Energieverbrauch pro Tonne Produktionsaus-stoss und dem Grundverbrauch des Gebudes, wurde vorgeschlagen. Dieses

TODOMO ermglichte die energetische Modellierung von folgenden Typenvon Batch Produktionsanlagen:

- Monoprodukt Batch Betriebe- Mehrprodukt Batch Betriebe mit konstantem Produktemix- Mehrzweck Batch Betriebe in denen ausschliesslich hnliche Che-

mikalien produziert werden

Die Resultate zeigten einen signifikanten Elektrizittsverbrauch der Infra-strukturanlagen auf (ca. 50% des totalen Stromverbrauches). Der Grundver-

brauch fr Dampf und Sole war nur gering. Der spezifische Energieverbrauchder untersuchten Gebude zeigte einen klaren Zusammenhang mit dem Automa-

tionsgrad der Produktionsgebude und den produzierten Chemikalien.Fr den Heizdampfverbrauch des Gebudes wurde ein Modell entwickelt,

welches nur vom Luftwechsel innerhalb des Gebudes und von den Heizgradta-gen abhngig ist.

Fr Mehrzweck Batch Betriebe mit stark unterschiedlichen Produktionspro-zessen und schwankendem Produktemix war das TODOMO nicht anwendbarund ergab ungenaue Resultate. Fr diese Betriebe wurde ein bottom-up Modell(BOTUMO) postuliert und entwickelt. Das Modell besteht aus einem produkti-onsabhngigen Teil und einem batchzeitunabhngigen Grundverbrauchsteil. Der

produktionsabhngige Teil besteht aus einem von den Chemikalienspezifikatio-nen abhngigen Term, einem von den Apparatespezifikationen abhngigenTerm und einem zeitabhngigen Verlutsterm.

Durch diverse Messungen konnten Einzelapparate- und Einzeloperations-modelle entwickelt werden. Diese Modelle bentigen ausschliesslich einfach zu

bestimmende Substanz- und Apparatedaten und modellieren zudem die Verlusteder verschiedenen Apparate. Die Modelle wurden so entwickelt, dass sie sicheinfach auf andere Betriebe und Chemikalien bertragen lassen und nicht aufeinen spezifischen Betrieb beschrnkt sind. Bereits aus den Einzelapparate-modellen ging hervor, dass die Verluste fr Dampf- und Soleverbrauch signifi-kant waren. Die Verluste wurden durch einen Verlustkoeffizienten charakteri-siert. Dieser Verlustkoeffizient beschreibt den Wrmebergangskoeffizienten an

der Aussenflche eines Apparates. Fr den Dampfverbrauch wurde ein Ver-

xxv


29/238

lustkoeffizient von 4.210-2kW / m2/ K und fr den Soleverbrauch ein solchervon 1.710-2kW / m2/ K gefunden. Hieraus kann geschlossen werden, dass ber50% des Verlustes beim Dampf auf das Heiz/Khlsystem mit seinen Kondensa-tableitern zurckzufhren sind.

Zur Modellierung des Energieverbrauches ganzer Produktionsgebude mitHilfe des BOTUMO wurden die oben erwhnten Gleichungen in ein ExcelModell integriert. Dieses Modell wurde zur Modellierung des Energieverbrau-ches des ganzen Produktionsgebudes fr einen und zwei Tage, eine Woche,sowie einen Monat verwendet. Die Modellrechnungen zeigten sehr gute Ge-nauigkeiten fr die Modellierung von lngeren Perioden (mit Hilfe der PSP Da-ten).

Analysen ber die Periode von einem Monat zeigten, dass die Apparate-gruppeReaktoren und Nutschentrocknerden wichtigsten Energieverbraucher imuntersuchten Gebude darstellt (neben dem Infrastrukturverbrauch bei der Elek-

trizitt). Detailliertere Analysen dieser Apparategruppe zeigten, dass ca. 30-40%des Dampfverbrauches fr Verluste aufgewendet werden musste. Dies weist aufgrosse Optimierungspotenziale hin. Beim Elektrizittsverbrauch konnte gezeigtwerden, dass die kleinen Umwlzpumpen der Heiz-/Khlsysteme der Reaktorenund Nutschentrockner ca. 25% des gesamten Elektrizittsverbrauches dieserApparategruppe bentigen (d.h. ca. 50% des Verbrauches der Rhrwerke),wenn die elektrische Heizung nicht luft (sonst entsprechend weniger). Dieelektrische Heizung wird fr einen einzelnen Hochtemperaturreaktor bentigt.Der Verbrauch dieser Heizung ist grsser, als der Verbrauch aller im Betriebvorhandenen Rhrwerke (ber 25).

Eine Sensitivittsanalyse wurde durchgefhrt und zeigte, dass von allen un-

tersuchten Parametern, die Batchzeit den grssten Einfluss auf den gesamtenGebudeenergieverbrauch hat. Eine Variation der Batchzeit um 50% resultier-te in einer Vernderung des Gesamtenergieverbrauches von 30%.

Verschiedene Einsparpotenziale wurden gefunden. Diese reichen von derElimination von Rcklaufbedingungen bis zu einem vllig neuen Design fr dieklassischen Heiz/Khlsysteme.

Die Anwendbarkeit des BOTUMO wurde sowohl fr kurze als auch fr lan-ge Zeitabschnitte gezeigt. Die verschiedenen Zeitabschnitte wurden mittels un-terschiedlicher Genauigkeiten der Eingabedaten modelliert. Die bertragbarkeitauf andere Produktionsgebude und Chemikalien muss in zustzlichen Fallstu-dien untersucht werden.

xxvi


30/238

Introduction

1 Introduction

1.1 Context and Motivation

About 50% of industrial processes (Stoltze et al. 1995) and chemicalproduction (Phillips et al. 1997) worldwide are batch processes.

Energy consumption of production processes contributes significantly tooverall resource use. The fewer resources the production of a substance (orfunctional unit) uses, the more environmentally friendly the process is (assum-ing that all other parameters remain constant). Moreover, about 75% of man-made air pollution is caused by energy use (Wang and Feng 2000). Therefore,minimization of energy consumption is listed as the sixth principle of greenchemistry (Anastas and Warner 1998).

The chemical industry is a large, and in certain sectors, intensive user of en-

ergy. For example, U.S. chemical industry accounted for about 20% of totalmanufacturing primary energy consumption in 1994 (i.e., about 5.4 EJ) as statedin (DOE 2000; Worrell et al. 2000). This value is even greater, if oil and gasfeedstock were included. The US chemical industry sets in their Vision 2020the clear target to reduce energy consumption of chemical production and toimprove energy efficiency (ACS 1996; Eissen et al. 2002).

Energy consumption of plants engaged in continuous chemical productionshas been investigated extensively in the past by pinch technology (Linnhoff1993). Similar methods for batch production are not yet well established. Fur-thermore, such studies are usually limited to heat-integration (Bouhenchir et al.2001; Kemp and Macdonald 1988) and therefore rely on available storage ca-

pacity or constant production schedules. Other studies account for time-varyingtemperatures (Vaselenak et al. 1986) and rescheduling (Vaselenak et al. 1987).The use of these methods in batch production is limited because most of themare considered as too complicated, lengthy, demanding and complex to be of

practical interest for most of the cases encountered (Stoltze et al. 1995). Thefact that energy costs amount to about 5 to 10% of total production costs forcommon chemicals produced in batch operation (Vaklieva-Bancheva et al.1996) limits the efforts undertaken in achieving high energy efficiency. A help-ful overview of energy consumption and management in batch production is

provided in (Grant 1996).

Reliable statements on energy efficiency and improvement potentials ofproduction processes need standardized parameters characterizing energy con-sumption. It is only reasonable to set energy targets if the relation between theactual and the minimal practical energy consumption is known. In multiproductand in multipurpose batch plants, this energy consumption has to be allocated todifferent products and unit operations. Focus may then be put on the greatestsaving potential of the largest energy consumers. This prevents a wasting of thelimited resources for re-engineering by using them for the most effective saving

potentials.A survey on the chemical industry in the U.K. showed that, on average for

different chemical branches, the most energy is used for process heating (40%),

1


31/238

State of the Art

with distillation (13%), drying (10%) and compression (10%) being the othermajor energy-consuming unit operations (Anonymous 1986b).

Energy models for multiproduct and multipurpose batch plants are lackingin industry. It is known that energy consumption is, to some extent, related to

production output, but exactly where energy is used is not known. Whether thedependence on production output is strong or whether the base load consump-tion of a building is dominating is not known. Energy consumption models on

building level are needed for providing consumption forecasts to the energysupplier and for calculating total production costs.

1.2 State of the Art

Many papers, models and theories of the past and present research havedealt with energy modelling of continuous processes as stated in (Linnhoff1993; Worrell et al. 2001; Zalba et al. 2003) or heat exchanger networks

(Furman and Sahinidis 2002; Gundersen and Naess 1988; Jezowski 1994a; Je-zowski 1994b; Zhao et al. 1998). Batch production is hereby most of the timeneglected or the models are considered as too complex for industrial use (Stoltzeet al. 1995). Nevertheless, much literature is available on scheduling of batch

plants, which allows a more efficient use of energy by reducing waiting andchangeover times (see e.g., (Caldern et al. 2000; Papageorgiou et al. 1994;Reklaitis et al. 1997; Sahinidis et al. 1989; Suhami and Mah 1982; Verwater-Lukszo 1996; Vin and Ierapetritou 2000)).

A novel approach named as Time Average Model (TAM) or Time SliceModel(TSM) is introduced by (Linnhoff et al. 1988) and further used by severalauthors (e.g., (Krummenacher 1997; Stoltze et al. 1995; Zhao et al. 1998)).

Both the TAM and the TSM adapt the concept of pinch analysis introduced by(Linnhoff et al. 1982) to batch processes. The TAM assumes that all batch op-erations can be performed at any time and in any order, so that no account isgiven to scheduling or time availability of energy flows. The time dependentconsumption of a batch reactor is averaged over the whole batch time for one

process resulting in a mean consumption for the whole process. In other words,time is completely ignored as a constraint and the energy source and sink values

become averaged over a chosen period. This results in a model similar to con-tinuous processes that can be handled by pinch analysis. This model is easy-to-use but has, nevertheless, not much in common with the real behaviour of batch

production and is therefore of no significant practical use. The TSM, on theother hand, does incorporate assumptions about time, e.g., cycle times and timeavailability. Time is then sliced into periods during which process energyflows can be analysed and a separate model is calculated for each slice. Foreach of these slices, energy consumption is again analysed as an average con-sumption over the whole time of the slice. Both the TAM and the TSM, never-theless, have no wide acceptability in industry. Furthermore, they have not beenapplied to different energy carriers (only examples for steam are available) anddifferent products and processes in one unified model.

Some authors mention that significant savings of energy cost (and consump-tion) in batch plants of up to 25% are possible (Allen and Shonnard 2002;

Ashton 1993; Benz 2003; Krummenacher et al. 2002; Phillips et al. 1997;

2


32/238

Introduction

Rumazo et al. 2000). (Jimnez-Gonzlez and Overcash 2000) state, that espe-cially energy challenging in early process phases reduces the level of emissionsduring the whole lifecycle of the product. In this paper, energy lifecycle infor-mation is developed to support the decision-making process.

Besides these detailed papers mentioned above the basic concept of energyaudit is essential for performing an energy analysis of a whole production plant.The concept of energy analysis is widely discussed in literature; some examplesmay be found in (Bhatt 2000a; Bhatt 2000b; Ganji 1999; Haman 2000; Hoshide1995; Robert and Markus 1994). (Blickenstorfer 1999) provides a good over-view of literature dealing with energy analysis.

No models are available in the literature to compute the energy consumptionof batch processes, accounting for the consumption caused by the chemical

process itself, the consumption due to the equipment and especially the losses ofthe different systems. This will be investigated and analysed in this thesis (seethe next chapter).

1.3 Thesis Statement and Contribution

Energy consumption plays an important role in todays business since mostof the processes are not possible without an appropriate energy source (Krsten1996). Allocation to different processes and products is, nevertheless, often not

possible for batch production. As stated in the preceding chapter, energy con-sumption contributes quite significant to production costs and to environmentalhazard in the producing industry. Nevertheless, accurate and ready-to-use toolsfor predicting or modelling the energy consumption of chemical batch plants aremissing. Goals for energy savings or targets for focusing on improvement po-

tentials are most of the time set according to common (engineering) sense orpolitical targets. This is, contrary to continuous production processes, wheredetailed models for energy consumption and integration methods are available,an unsatisfying situation. Moreover, legislation needs tools to predict the en-ergy saving potentials of plants to meet the goals set (see e.g.,(Eidgenossenschaft 1999)) and the Kyoto protocol (see (http://unfccc.int/resource/docs/convkp/kpeng.html ) for the text of the protocol and (Rsonyi2002; Thne and Fahl 1998; Wrsten 2003) for some comments). The goals setin CO2-legislation as mentioned in (Eidgenossenschaft 1999), lead to voluntarysavings and agreements of objectives with industry as mentioned in (BFE2001a; BFE 2001b) and in (BFE 2002). To succeed in these agreements of po-tential savings, detailed models for energy consumption are required. Withoutsuch models, it would not be possible to control whether or not the goals areachieved.

For all these reasons, easy to use tools should be available for energy model-ling of chemical batch production plants. A former thesis by (Blickenstorfer1999) showed the possibility of energy modelling on building level for a spe-cific kind of batch production (top-down approach for one kind of batch produc-tion plant as discussed in Chapters 3.1 and 4below). Applicability of this ap-

proach to other buildings will be investigated in this thesis.In this thesis, easy-to-use and adaptable single unit operation models

(SUOM) on apparatus level are developed. The new approach of the thesis of-

3


33/238

Thesis Organization

fers the possibility to model the energy consumption of a complete productionplant with a detailed bottom-up model based on the SUOM with the help of eas-ily accessible data. The required data consists of apparatus specifications,

building infrastructure consumption, specifications of the chemicals and the

production processes as well as operation times from the process step procedure(PSP). With the help of this model, it is possible to gather information on theenergy consumption of a specific batch plant with a minimal of surplus meas-urements and data requirements. The data may be aggregated for different lev-els of analysis, as the user likes.

The applicability, usability, and accuracy of such models will be investi-gated in this thesis. The models should be simple enough to be useable in daily

production and accurate enough to analyse the energy consumption of a produc-tion plant in detail. Such models would help legislation and particularly the

production chemists and plant management to analyse and in a second step op-timise the energy consumption of their production plants.

1.4 Thesis Organization

Figure 1-1 shows the structure of the thesis with its different chapters. Thechapters are organized to provide first the theoretical background and show the

postulated models for modelling of the energy consumption of chemical batchplants and show afterwards the measurements, modelling results and outcomesof the investigations of this thesis.

Chapter 1summarises the literature in the field of the thesis, gives the the-sis statement and shows how the thesis results integrate in the literature knownso far. In Chapter 2, the structure of batch plants is described. The different

types of batch plants and their characteristics relevant for the investigations ofthis thesis are discussed. The two models (i.e., top-down and bottom-up model)are introduced in Chapter 3. The generic modelling equations for the two ap-

proaches are given and shortly discussed. The general principles, definitions,and usage of the single unit operation models are presented as well as the differ-ent possibilities and levels of energy modelling and analysis of whole produc-tion plants according to the bottom-up approach. The measurements, modellingresults, and model applicability for the top-down model are presented in Chap-ter 4. The characteristics of the different investigated buildings are discussed interms of their influence on the applicability of a top-down model to these plants.In Chapters 5 and 6, the bottom-up model is discussed in detail. Chapter 5introduces the specific single unit operation models and shows the measure-ments performed for investigating the required parameters for the models pre-sented in Chapter 3. Chapter 6 combines the specific single unit operationmodels developed in Chapter 5 and the infrastructure consumption to a model ofthe whole plant according to the general equations presented in Chapter 3. Themodel is discussed and the different levels of analysis are presented according tothe equations given in Chapter 3. Sensitivity analyses of the different model-

parameters are given and discussed as well. In Chapter 7,the different resultsof the thesis are summarised and a short outlook on future work and open ques-tions is given.

4


34/238

Introduction

1 Introduction1 Introduction

2 Description of the Structure of Batch Plants2 Description of the Structure of Batch Plants

3 Description of the Top-Down and the Bottom-Up Model3 Description of the Top-Down and the Bottom-Up Model

4 Top-Down Modelling4 Top-Down Modelling

Bottom-Up ModellingBottom-Up Modelling

5 Single Unit Operations5 Single Unit Operations

6 Whole Plants6 Whole Plants

7 Conclusions & Outlook7 Conclusions & Outlook

AppendixAppendix

Figure 1-1: Structure of the thesis

The Appendixpresents the model assumptions, the actual Excelmodel ofthe whole plant and the modelling results in Chapter A. Chapter B in theAppendixpresents the measuring equipment and discusses the accuracy of thedifferent measurements performed. In Chapter C in the Appendix,the results

and measurements for different special investigations performed during the the-sis are given. These investigations include the distribution of the batch times,the investigations concerning the reflux conditions, and investigations on theinfluence on the cleaning of a vessel on its energy consumption are presented.Chapter Din the Appendixpresents the results of the measurements performedon apparatus and building level during the thesis. Chapter Ein the Appendixsummarises the main improvement potentials for a batch plant and gives achecklist for daily use where energy may be saved in daily operation. In Chap-ter Fin the Appendix,a small glossary of the terminology of batch operations is

provided.

5


35/238


36/238

Structure of a Batch Plant

2 Structure of a Batch Plant

2.1 The General Structure of Batch Plants

An introduction to the terminology of batch production is provided in (ISA1995) and in (Blickenstorfer 1999). The most important definitions for the pur-

pose of this thesis and of batch production in general are given in Table F-1.A batch plant cannot be operated by itself. Many different processes, plants

and operations have to be performed before a raw material enters the plant andafter a substance (product) leaves the plant. A generic value chain of a chemical

production is depicted in Figure 2-1. Basic chemicals like crude oil are ex-tracted from nature, transformed and upgraded to intermediate chemicals thatare the required raw materials for the pharmaceutical and fine chemical indus-try. These intermediate chemicals are most of the time produced with continu-

ous processes in large amounts. Fine chemicals on the other hand, are high-value, low-tonnage products. These products are often produced in batch proc-esses to maintain the flexibility and efficiency of low production amounts. Fora general overview of the chemicals produced in batch production, see(Parakrama 1985) or (Anonymous 2001). The same is true for the upgrading(i.e., formulating and mixing) of the fine chemicals. This is often done with thehelp of batch processes as well. The final industrial application and the end us-ers often use batch processes too for their purposes. Therefore, batch processesare of high interest. Because of the difficulties related to the modelling of batch

processes and the high prices often achieved on the market (compared to thetotal production costs), energy optimisation was only a minor issue so far. To-day, prices of the fine chemicals are decreasing, production and raw materialcosts are increasing (i.e., decreasing margins). Moreover, environmental legis-lation gets stricter and energy consumption is sanctioned (see e.g., (Burkhardt2002; Eidgenossenschaft 1999; Ewers 2000; Gundersen 1991; Rsonyi 2002;Wrsten 2003)). Therefore, the importance of minimising energy use is increas-ing. Moreover, modelling is required to declare and check the voluntary agree-ments of objectives for energy-savings in industry as mentioned in (BFE 2001a;BFE 2001b).

Basic

Chemicals

Intermediate

Chemicals

Fine

Chemicals

- Formulations

- Mixes

Industrial

Applications

Natural

Products

End

Users

Figure 2-1:Value chain in the chemical industry (shaded: typical batch proc-

esses)

The shape of a batch reactor has little changed for the last 500 years. The

stirred tank has remained the same from the alchemists time until today, al-

though new concepts are available and propagated today (e.g., micro-reactors

etc. as described in (Hller and Renken 2000; Stitt 2002)). The uncanny resem-

blance between a 16thcentury gold plant depicted in Figure 2-2 and a modern

fine chemicals plant, with both being dominated by the stirred tank reactor, has

been noted by (Stankeiwicz and Moulijn 2000).

7


37/238

The General Structure of Batch Plants

Figure 2-2:Engraving of a 16thcen-tury gold processing plant (Stitt 2002)

In (Villermaux 1995) another ex-ample of this fact is stated. He notesthat the technology of the Concorde

has almost nothing to do with that ofthe Wright brothers or Bleriot, and thatthey would probably not be able to flyit. By contrast, technical drawings ofchemical processing apparatus, such as

batch reactors, taken from patents filedin the 1880s are remarkably similar tothose still in use and being installedtoday. Whether or not new conceptsfor batch reactors are required from anenergetic point of view will be investi-gated in this thesis.

A batch plant (i.e., area, see Table F-1)usually consists of several parts, asdepicted in Figure 2-3 and Table 2-1. The batch productionequipment repre-sents the heart of the batch plant (i.e., batch reactors, batch dryers, nutsche fil-ters, etc.). In this equipment, the process input is transformed to the processoutput (i.e., the actual value is added to the product).

Another part of a batch plant consists of so calledspecial equipment. This isequipment with special features, not common to the usual batch reactor likehigh-temperature devices, continuous equipment such as distillation columns forsolvent recovery or continuous drying equipment, or equipment for filling and

packaging. This equipment is, in contrast to the batch production equipment,

very different from plant to plant depending on the kind of process output of theplant.Theproduction infrastructureis required for specific processes. Equipment

like circulation pumps, vacuum pumps, etc. could fall in this category. Theseapparatus are not operated continuously for the whole building but specific forone or the other process.

The final part of a batch plant is represented by the building infrastructure.This infrastructure consists of heating and ventilation systems, general vacuumsystems, waste-air treatment, etc. All equipment units that cannot be allocatedto one specific process and that are therefore operated continuously or stepwiseare considered as building infrastructure for the purpose of this thesis.

8


38/238


39/238

The Differences between Batch Plants

The different utilities (e.g., steam, electricity) required in a production build-ing are most of the time produced externally in a central facility for a completesite. Typically, cooling media production is an exception from this rule of cen-tralized production. The term cooling medium, as used in this thesis, stands for

ice or low-temperature fluids like brine (i.e., cooling water is not investigatedbecause of the lack of measurements). Cooling media production is mostlydone directly in the specific plant because decentralized production of coolingmedia is efficient and transportation losses would be significant in centralized

production.The recovery and storage of spent solvents is done either within the batch

plant or by a contractor. Large equipment is required for this purpose. Thisequipment is considered independently (decoupled from batch-production). Op-timisation of the regeneration operation can thus be done independently as well.

2.2 The Differences between Batch PlantsIn batch production, different kinds of batch plants can be differentiated:

- The monoproduct batch plant- The multiproduct batch plant- The multipurpose batch plant

The characteristics of these different plants will be discussed shortly in thefollowing sections.

2.2.1 The Monoproduct Batch Plant

A monoproduct batch plant is a plant that is designed especially for the pro-duction of one specific chemical. It is a dedicated plant with fix installation.The path of an amount of raw material through the plant is clearly defined. Noor minimal manual operation is usually required since automation is elaboratedand recipes are seldom changing (if cheap labour is available, degree of automa-tion may be low as well). In an automated plant, data availability is most of thetime good. Because of the constant production steps, focus is given to optimisa-tion of the production process (e.g., energy savings by heat integration(Krummenacher 1997; Krummenacher and Favrat 2001)).

2.2.2 The Multiproduct Batch Plant

A multiproduct batch plant is a plant where different chemicals are producedthroughout the year, but the same production steps are mostly performed in thesame equipment (see (Rippin 1992) as well). Fixed lines of batch reactors are

producing different products (probably different products on one line at differ-ent times of a year). The amounts of the different products may vary with salesrequirement. Therefore, production mix may not stay constant and may have aninfluence on scheduling and utility requirements. Each line in a multiproduct

batch plant may be considered as a (small) monoproduct batch plant (i.e., fixedmaterial pathways, potential for specific optimisation) for each production pe-riod.

10


40/238

Structure of a Batch Plant

2.2.3 The Multipurpose Batch Plant

A multipurpose batch plant, on the other hand, produces different chemicalslike the multiproduct batch plant, but in each equipment unit, different produc-

tion steps might be performed (i.e., such plants are characterized by high flexi-bility; (Rippin 1992)). The units are most of the time independent from eachother and connected via (flexible) pipes. This allows a construction of a produc-tion path for the purposes of one specific chemical, each time this chemical may

be produced in the plant in a different way (i.e., in different reaction vessels).The pathway of a chemical in the monoproduct and the multiproduct batch plantis most of the time from top to bottom for reasons of ease of transportation (i.e.,gravitation is helping to transport the chemicals). In multiproduct batch plants,this is probably considered as well, but not necessarily, because this would re-strict the flexibility of the plant.

No or few dedicated equipment can be found in a multipurpose chemical

batch plant. This implies that all the equipment items are capable to perform allpossible unit operations and limits the optimisation potential.

The infrastructure part of the multipurpose batch plant may also differ fromthe infrastructure of the other two kinds of batch plants. Because of the multi-

purpose characteristic of these plants, the infrastructure is not optimised for onespecific use. It is tried to operate as few infrastructure equipments as possible(cost savings) but to install the equipment as flexible as possible. This allows

producing many different products. If a new product with new infrastructurerequirements is introduced to the plant, the new infrastructure equipment has to

be integrated in the former concept. This opens doors for oversizing and ineffi-ciencies in a specific production campaign but ensures the flexibility of the

plant.

11


41/238


42/238

Two Approaches for Energy Modelling

3 Two Approaches for Energy Modelling

For the modelling of energy consumption, two basic models are traditionally

found in literature (see e.g., (Aebischer et al. 1988; Kbler 1986)): a Top-Down-model (TODOMO) based on measurements of a complete system and aBottom-Up-model (BOTUMO) based on diverse measurements of single

parts of a system and summation of the single energy consumers (as stated e.g.,in (Werbos 1990)). Both models were adapted, elaborated, and investigated inthis thesis for the specific purposes of the chemical industry.

The purpose of the two models in this thesis is to model and allocate energyconsumption of (chemical) batch plants. The time horizon will be no shorterthan one day. This limitation was set, because the short-term modelling wouldrequire clumsy integral equations that would need many input parameters usu-ally unavailable in production business. Moreover, the important period for a

production plant is one week or even one month. For those periods, accountingof the production output is available and contractors bill the energy consump-tion based on the consumption during such periods.

The following two subchapters introduce the modelling concepts of theTODOMO and the BOTUMO together with the equations of these models.

3.1 The Top-Down Approach

3.1.1 The Model for the Production Dependent Energy Consump-

tion

For each utility, a model that computes the energy consumption of a build-ing as a function of the specific consumption per ton of product output and thebase consumption was postulated. The equation for the TODOMO is repre-sented by Equation (3-1).

Em= SmPO + Bm (3-1)

where Emis the overall consumption of a specific energy form in a specifiedperiod (i.e., longer than one day, mostly one month) in kWh per period, Sm isthe specific consumption of one energy form per ton of products in kWh / t, POis the production output on a weight basis during the period specified (including

all products and intermediates leaving the plant, excluding solvents and aggre-gates) in t per period, and Bmis the so-called base consumption of the buildingof a specific energy form in kWh per period. The base consumption is the con-sumption of a warm production building that is ready to start production but inwhich no production process is actually running (i.e., base consumption meas-ures infrastructure consumption and infrastructure losses).

Two different possibilities exist for the determination of the base consump-tion. Each building undergoes a period of revisions at least once a year. Duringthis period, maintenance activities are undertaken and production is shut down.Therefore, it is possible to measure the consumption of the warm (ready to pro-duce but not yet producing) and the cold (only safety equipment is running)

13


43/238

The Top-Down Approach

building. Losses of the whole system have to be analysed in this way. A sec-ond possibility is the direct measurement of the consumption of the specific in-frastructure equipment itself since it is known which apparatus is on stream dur-ing shutdown or production.

Such linear models were also postulated by (Blickenstorfer 1999). Modelsof this kind are only applicable to monoproduct or multiproduct batch plants ormultipurpose batch plants with similar products as will be discussed in Chap-ter 4.

For multipurpose batch plants with large differences between their productsand changing production mix, linear TODOMO are not applicable as will beshown in Chapter 4. For these buildings, that are the main research topic of thisthesis, a new BOTUMO is postulated and discussed in the Chapter 3.2.

3.1.2 The Heating Steam Model

Production plants are heated by heating the fresh air entering the building.This is (unlike to residential buildings, where radiators are used most of thetime) done by heat exchangers with condensing steam. This (comfort) heatingsteam is measured separately. A linear model only depending on degree-days(see (http://www.eia.doe.gov/neic/infosheets/degreedays.htm )) was first postu-lated according to Equation (3-2)but found to be not applicable as discussed inChapter 4.3.1.

SC= DSS DD + B (3-2)

where SCis the steam consumption in MWh / month, DSS is the degree-

day specific steam consumption in MWh / C / d, DDis the number of degree-days in Cd / month and B is the base consumption of heating steam inMWh / month, which is unique for each building.

Since the air change rate of production buildings is significantly higher thanfor residential buildings (safety reasons), the model was adapted to account forthe air change rate. The adapted model was found to be applicable for the heat-ing steam consumption of batch plants (see Chapter 4.3.1)and is depicted in thefollowing equation:

SC= 0.32 ACRDD + B (3-3)

whereACRis the air change rate of a building in h-1.If no production infrastructure uses heating steam and if the main pipe of

heating steam is closed during summer, the base consumption is almost equal tozero. Otherwise, the base consumption has to be measured or estimated before

predictions of heating steam consumption can be made, as discussed in Chap-ter 4.3.1.

14


44/238

Two Approaches for Energy Modelling

3.2 The Bottom-Up Approach

The basic equations for the BOTUMO, describing the concepts of calculat-

ing the energy consumptions for heating and cooling procedures (Chapter 3.2.1)and calculating the energy consumption of the electric equipment (Chap-ter 3.2.2)are presented here. These basic equations are combined in differentways for the different unit operation models on single apparatus level presentedin Chapter 3.2.3 and 5. The single unit operation models are summarised to re-sult in a model of a whole plant (see Chapter 3.2.3 as well).

3.2.1 Equations for Heating and Cooling of Substances

In any book dealing with heat transfer and physical chemistry (e.g., (Atkins1990) or (Wedler 1987) or (Perry et al. 1997)), the basic equations for the heat-ing and cooling of substances can be found. The heating or cooling of a sub-stance without phase change can be calculated by Equation (3-4).

(3-4) =2

1

T

T

PdTcmH

where H is the enthalpy change in k

Analysis and Modelling of the Energy Consumption of Chemical Batch Plants

Documents

Transcript of Analysis and Modelling of the Energy Consumption of Chemical Batch Plants