U.S. DEPARTMENT OF COMMERCE
National Technical information Service
PB-271 128
Chapters 1-4
Final Report for Low
Pressure Tests of the
CPU-400 Pilot Plant
Combustion Power Co, Inc, Menlo Pork, Calif
Pr«par«d ftr
Industrial Environmental Research Lab, Cincinnati, Ohio
Sep 77
-------
U.S. DEPARTMENT OF COMMERCE
National Technical information Service
PB-271 128
Final Report for Low
Pressure Tests of the
CPU-400 Pilot Plant
Combustion Power Co, Inc, Menlo Park Calif
Prepared Kr
Industrial Environmental Research Lab, Cincinnati, Ohio
Sep 77
-------
FINAL REPORT FOR LOW PRESSURE TLSTS OF
THE CPU-400 PILOT PLANT
by
Combustion Power Comnany, Inc.
1346 Willow Road
Menlo Park, California 940P5
Contract No. 63-03-C054
Program Element No. 10i3314
Project Offi cer
Richard A. Chapman
Energy Systems Environmental Control Division
Industrial Environmental Research Laboratory
Cincinnati, Ohio 45Z68
U. S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
CINCINNATI, OHIO 45268
-------
4. TITLt ANO SUtlTIT' 6
FINAL REPORT FOR LOW PRESSURE TESTS OF THE
CPU-400 PILOT PLANT
TLCHNiCAL lUI'OHT t>ATA
/'li jif icaJ lining fii'iii H'i /'i. ni./ , ''' ' "> i I»H/I
i r NO
EPA-600/2-77-195
7 ALlTMOHISI
Combustion Power Company
Septembe'_ '977 issuing daje
(,. rtHronwiNtj OMI.ANIZ-V i IUN COOL
a.PlHFOHMINGOIK.ANI.'AllON Ml TO HI NO
9 PlRFOnMINGOHGANI^ATIONNAMt ANUAL'UHISS
Combustion Power Company, Inc.
1346 Willow Road
Menlo Park, California 9*025
12. SPONSOriING AGENCY NAML AND AOOTItSS
Industrial Environmental Research Laboratory
Office of Research and Development
U. S. Environmental Protection Agency
Cincinnati, Ohio 45268
J Ml CII'll Nl S ACl I V.U >» '.O
10. V R OC'. II A r.l I II Ml N I NO
V)~coNiTtAC r, on ~"
68-03-0054
\-J. TYFl OF HI.POflT A\O PLHIOD COVLHLD
__ [in^v __ ^__ __ _____
14. SPONSORING AGtNCY COOt
EPA/^00/12
IS. SUPf LtMl NTAMV NOTKS
16. AOSTHACr
This report presents the progress made during the component design nhase of
a program to develop an economical and environmentally safe waste-energy system
known as the CPU-400. It discusses the hardware development and low oressure
testing performed to evaluate CPU-400 operational characteristics in a large
scale pilot plant. This work was oerformed as a orerequisite to final high
pressure testing of the pilot plant facility.
Significant accomplishments made during this contract period included:
the final procurement and testing of hot gas system components such as the full
scale vertical combustor, three particle separators, and ash removal equipment;
the final design and procurement of a computerized orocess control system; the
procurement of a turbo-electric system, including the completion of all design
required for its incorporation into the pilot plant; the completion of a survey
of gas turbines for the prototype plant; and the successful performance of long
duration, low pressure testing of all components integrated into the pilot plant
system. Other work accomplished during the contract period included the quanti-
tative and qualitative analyses of off-gases from the-sol id waste-fired combustor,
the testing of a subscale granular filter, and the improvement of the solid waste
17.
KEV WORDS ANO DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPCN ENDED TERMS
COSATI I Ictd/CilOUp
Refuse Disposal*
Fluidized Bed Processors
Combustion*
Turbines*
Cyclone Separators
Process Control
Shredders
Incinerators
Electric Power
Generation
CPU-400
Solid Waste
Ai r Classification
21/D
10/A
13/H
13/B
ia.'OISTRIOUTION STATEMLNT
RELEASE TO PUBLIC
19. SECURITY CLASS (Tim Ktpotl)
' Unclassified
21. NO. OF PAGL;
406
20. SECURITY CLASS (Thn pagt)
Unclassified
11. PRICE
rX' (\ ) V
CPA Form 1J3L-1 (»-73)
-------
DISCLAIMER
The Industrial Environmental Research Laboratory-Cincinnati
has reviewed this reoort and approved its publication. Approval
does not signify that the contents necessarily reflect the views
and policies of the U. S. Environmental Protection Agency, nor
does mention of trade names or commercial products constitute
endorsement or recommendation for use.
-------
FOREWORD
When energy and material resources are extracted,
processed, converted, and used, the pollutional impact on
our environment and even on our health often requires th.it
new and increasingly more efficient pollution cortr.'l
methods be used. The Industrial Environmental Research
Laboratory-Cincinnati (lERL-Ci) assists in developing and
demonstrating new and improved methodologies that will
meet these need', both efficiently and economically
Since 1967, the EPA has supported the development
of the CPU-400, a system for converting waste directly
into electrical energy. This report describes the
progress made during the component design phase of the
program. This report will be of special interest to
those involved in the design, construction, and testing
of waste-to-energy systems.
David G. St.phan
Di rector
Industrial Environmental Research Laboratory
Cincinnati
11
-------
ABSTRACT
This report presents the progress made during the component de-. iqn
phase of a program to develoo an economical and environmentally sarV
waste-energy system known as tne CPU-40n It discusses the hardvvdre-
development and 1 ow pressure testing performed to evaluate CPU-4GQ
operational cha~acteristics in a large scale pilot plant This work
was performed as a prerequisite to final high pressure testing of the
pi lot plant faci1i ty.
Significant accomplishments made during this contract period
included: the final procurement and testing of hot gas system com-
ponents such as the full scale vertical combustor, three particle
separators, and ash removal equipment; the final design and procure-
ment of a computerized process control system; the procurement of a
turbo-electric system, including the completion of all design required
for its incorporat on into the pilot plant; the completion of a suyvew
of gas turbines for the prototype plant; and the successful performaiv 3
of long duration, lev/ pressure testing of all component* integrated
into the pilot plant system. Other work accomplished dur~nn the con-
tract period included the quantitative and qualitative analyses of
off-gases from the ^olid waste-fired combustor, the testing of a sub-
scale granular filter, and the improvement of the solid waste processing
system throughput rote.
Based on test results, it is concluded that the pilot plant is
ready for turbine integration and frgh pressure testing. Combustion
efficiency has exceeded design requirements and gas sampling analyses
have indicated that present pollution control methods have met or ex-
ceeded all known pollution control requirements.
This report was submitted in fulfillment of Contract Number
68-03-C054 by the Combustion Power Company, Inc., under the sponsorship
of the Environmental Protection Agency, Office of Research and Develop-
ment, National Environmental Research Center, Cincinnati, Solid and
Hazardous Waste Research Laboratory. Work was completed as of
January 1973.
i v
-------
CONTENTS
Page
Abstract ii
List of Figures
-------
CONTENTS (continued)
Page
III Hot Gas System (continued)
Design and Development of the Residue
Demoval System 3-88
Determination cf Need for Residue
Renoval System (TdSk LP-5) 3-88
Determination of Off-Gas Composition of a
Fluid Bed Combustor Fueled with Municipal
Solid Waste (Task PL-3) 3-- !3
Test Setup for Determination of Off-Gas
Composition 3-115
Gas Analyses 3-115
Particle Analyses 3-119
Turbine Blade Material Testing 3-121
IV Controls and Instrumentation 4-1
Design and Development of Gas Analysis
Equipment 4-2
Gas Composition System 4-2
Particle Analysis 4-5
Testing 4-5
Design and Development of Semiautomatic
Controls and Instrumentation 4-8
Location of Controls 4-8
Gas Temperature Control 4-8
Feedline Air Flow Control 4-12
Dump Air and Combustor Air Control 4-12
Fisher Process Controllers 4-12
Mode Switches 4-15
VI 1
-------
CONTENTS (continued)
Page
IV Controls and Instrumentation (continued)
Temperature and Pressure Measurements 4-17
Data Acquisition System 4-17
Design and Development of an Automatic
Control System 4-20
Location of Controls 4-20
Process Control Computer 4-21
Software 4-28
V Integrated System Tests 5-1
Task LP-1 Testing (16-Senes Tests) 5-2
Test Summary 5-2
Test Objectives 5-2
Test Setup and Hardware Description 5-3
Test Results and Analysis 5-10
Post Test Inspection. 5-12
Conclusions 5-15
Task LP-8 Testing (18-Series Tests) 5-16
Test Summary 5-16
Test Objectives 5 17
Test Setup and Hardware Description 5-17
T^sf Result
-------
CONTENTS (continued)
Priqe
V Inteqrated System Tests (continued)
Task LP-11 Testing (2C-Series Tests) 5-45
Test Summary 5-45
Test Objectives 5-46
Test Setup and Hardware Description 5-47
Test Results and Analysis 5-60
Post Test Inspection 5-75
Conclusions 5-79
VI Turbo-Electric System 6-1
Design and Procurement of the
Turbo-Electric System 6-1
Turbo-Generator Selection 6-2
Turbine Description 6-3
Power Generator 6-7
Switch Gear 6-7
Load Banks 6-8
Piping and Control Design Considerations 6-10
Survey of Candidate Turbines for the
CPU-400 Prototype Plant 6-19
VII References 7-1
VIII Appendices 8-1
Appendix A -Solid Waste Handling System 8-2
Introduction 8-2
Shredder/Ai r Classi fication/Pneumatic
Transfer Subsystem 8-3
IX
-------
VIiI I".'MJ..,I 11 n .". -y -,ii.' >j|.,t. Hand' no. .vc.*<"ii i ont-' ffd'
Shreddfi" Input ConvevO'-
Shredder
Sn>ed'ier Out feed Conveyor
Ai r '" I ass i f ier
Reject Fraction Conveyor
Pneumatic Transfer Component*
S-ibs ;s tern '.'ont rol
Atlas Stor-aqe/Combustor Foen 'v/;,,' _'-'
Atlas Storage Unit
Control System (/WTEK)
Transfer and Weioh Conveyo s
Load Splitter/Feed C.hute
feeder Valve(s)
Subsystem Control P-.?*i
Appendix B -Gas Composition and Particle Analyses
of Model Fluid Bed Combustor [xn.iust Gas 8-31
Introduction 8-31
Description of the'Analytical Techniques IKed for
the Collection and Determination of the Off-Gas
Constituents ' 3-31
Nitrogen, Oxygen and Oxides of Carbon 8-32
Hydrogen R-^3
Methane 8-33
Oxides of Nitrogen 8-33
-------
"! ! . r -'id'. ' or h''0 '''.<'! .if,, i ' irti'lr1 An.
','(-'1 Hulii :5ed ; .'" Cur. t - >r ;y|i.j;ist r,d^ (l
. i 1 ' 'ji T rrtx i :lf:
:" ' t*1" ' ' n ' f] ; 0 f'1 !t
'i . 3r ~'':<~n r lour i 'It.1
iiv '''oqt.'n Sul f i 'lo
f'-v, i hcit nil-, >Vn to
!,1 .UPS
So : ' W-istc Fuel ?j">c r-, i-r > on
I- r '!» i n--.--,
["' i . c.j-j iinn o* T?st "i^'". .. i t. ',
fids An.ilv'p';
Flvash "n.
Pef 0'mrendatiori for f,!rj Turbinp iliri^e "'ati>f i
Tes t
Appendix C -Gas f.ompos i tion rind P.ii't. irlf> Anal
of C'hajst Gas from the Full Sr.ihj Vei'tiral C
Int rod'.jr ti nn
las f'omposition Aridlyii^ Dt'Scri ut ion
("arbnn Monoxir!p anil ("arbrn Dioxide
Hydrocarbons
-------
CONTENTS (continued)
Page
VIII Appendix B -Gas Composition and Participate Analyses
of Model Fluid Bed Combustor Exhaust Gas (continued)
Sulfur Dioxide and Oxides of Nitrogen 8-53
Hydrogen Chloride --54
Off-Line Determination of Gas Stream Particle
Loading and Size Determination 8-54
Techniques for Determination of Particle
Size Distribution H-55
General Procedures and Typical Results K-66
XI 1
-------
FIGURES
No. Page
2-1 CPU-400 Pilot Plant 2-5
3-1 Vertical Combustor 3-2
3-2 Vertical Fluid Bed Combustor Assembly 3-3
3-3 Horizontal Fluid Bed Combustor Assembly 3-5
3-4 Horizontal Combustor Internals 3-6
3-5 Combustor Exhaust and Backheat Burner Ports 3-9
3-6 Air Distributors and End Feedpipes 3-10
3-7 Normalized Flow Characteristics 3-12
3-8 Vertical Combustor Air Valve Definition 3-13
3-9 Burner System Schematic 3-14
3-10 Flame Detection Schematic Diagram 3-16
3-11 Flame Detection Interconnection Diagram 3-17
3-12 Typical Oil Gun Assembly 3-19
3-13 Inertial Separator Tube 3-22
3-14 Inertial Seoarator Fractional Efficiency 3-24
3-15 First Stage Separator 3-26
3-16 Second Stage Multicyclone Separator Tube Assembly 3-29
3-17 Six-Inch Diameter Tube Configurations 3-30
3-18 Typical Separator Hopper Vibrator Installation 3-32
3-19 Ash Removal and Handling System 3-33
3-20 Separator Assembly Installed in Horizontal
Combustor System 3-35
3-21 Test Setup Schematic for 18-Series Tests 3-38
XI 11
-------
FIGURES (Continued)
No.
3-22 First Stage Inertia! Separator Tube Layout for
18-Series Tests 3-40
3-23 Particle Count Analysis Plot, Test 18-1 3-44
3-24 Particu Count Analysis Plot, Test lrt-2 3-45
3-25 Particle Count Analysis Plot, Test 18-6 3-46
3-26 Particle Count Analysis Plot, Test 18-7 (Sample 1) 3-47
3-27 Flyash Particle Count Analysis Plot, Test 18-3 3-48
3-28 Flyash Particle Count Analysis Plot, Test 18-6 3-49
3-29 Test Setup Schematic for 20-Series Tests 3-52
3-30 Six-Inch Diameter Tube Arrangement of Second Stage
Separator 3-55
3-31 Ash Deposits in Vane Section of Second Stage Separator
Tube 3-57
3-32 Third Stage Inertial Separator Assembly 3-58
3-33 Third Stage Cyclone Tube Assembly 3-59
3-34 Third Stage Cyclone Tube 3-62
3-35 Airflow Resistance Test Setup 3-63
3-36 Overall View of Collector and Setup 3-64
3-37 Ambient Airflow Resistance, Third Stage Separator 3-65
3-38 Flow Balance Test Results, Third Stage Separator 3-66
3-39 Dust Test Particle Size Analysis, Third Stjge Separator 3-69
3-40 Cyclone Efficiency vs. Stated Particle Size 3-70
3-41 Particle Loading Distribution, Test Series 20-4A 3-74
3-42 Rod Style Rattlers in 3S-Inch Tube Separator 3-75
xiv
-------
FIGURES (Continued)
No. Page
3-43 Flyash Buildup on Vanes of Third Stage Separator Tubes 3-76
3-44 Granular Filter Design 3-78
3-45 Granular Filter Outlet Louvers 3-81
3-46 Hot Gas Test Setup 3-83
3-47 Flyash Buildup at Inlet Louvers 3-87
3-48 Granular Filter Fractional Efficiency 3-89
3-49 Size Distribution of Material Penetrating and
Entering the Panel Filter 3-90
3-50 Model No. 3 Installation 3-92
3-51 Model No. 3 System Schematic 3-93
3-52 Model No. 3 Solid Waste Feed 3-94
3-53 16 Mesh Beach Sand 3-95
3-54 Temperature History (LP-5 Testing) 3-97
3-55 Aged Bed Material 3-100
3-56 Bed AP (Volume) and Inerts Added 3-103
3-57 Exhaust Gas Impingement Plate Configuration 3-107
3-58 Material Removal System Schematic - Low Pressure
Operation 3-112
3-59 Pilot Plant Material Removal System 3-114
3-60 Test Apparatus Schematic 3-116
3-61 Test Apparatus 3-117
3-62 Sample Collecting System 3-118
4-1 Gas Analyzer Equipment 4-3
4-2 On-line Gas Analysis System Schematic 4-4
xv
-------
FIGURES (Continued)
N£. Page
4-3 Particle Analysis, Off-line Technique 4-6
4-4 Coulter Z Analyzer 4-7
4-5 LP-8 Pilot PI art Control Room 4-9
4-6 Gas Temperature Control Loop (LP-8 and LP-11) 4-11
4-7 Feedline Air Flow Control 4-13
4-8 Dump Air and Combustor Air Control LOOPS 4-14
4-9 Basic Mode Control Circuit Logic (Illustrated for Three
Modes) 4-16
4-10 Data Acquisition System Block Diagram 4-18
4-11 Data Acquisition System 4-19
4-12 Pilot Plant Control Room 4-22
4-13 Generalized Con trol System Schematic 4-23
4-14 Computer System Block Diagram 4-24
4-15 Parameter Section of Cortmon Data Base 4-35
4-16 Task 50 Listing 4-46
5-1 LP-1 Test Setup 5-4
5-2 LP-1 Control Loop Simplified Block Diagram 5-5
5-3 Solid Waste Handling System (LP-1 and LP-8) 5-7
5-4 Shredder/Air Classification/Pneumatic Transfer Subsystem
(LP-1 and LP-8) 5-8
5-5 LP-1 Atlas Storage and Combustor Feed Sybsystem 5-8
5-6 LP-1 Airlock Feeder Valve and Drive Assembly 5-9
5-7 LP-1 Analog Record of Key System Temperatures 5-11
xvi
-------
FIGURES (Continued)
No. Page
5-8 LP-8 Test Setup 5-19
5-9 LP-8 Combustor and Particle Separator Configuration 5-20
5-10 LP-8 Simplified Control Loop Block Diaoram 5-22
5-11 LP-8 Atlas Storage and Combustor Feed Sibsystem 5-26
5-12 LP-8 Second Stage Separator Exhaust Temperature 5-29
5-13 LP-8 Bed Differential Pressure 5-37
5-14 LP-8 Superficial Velocity 5-42
5-15 LP-11 Test Setup 5-48
5-16 LP-11 Combustor and Particle Separator Configuration 5-50
5-17 LP-11 Simplified Control LOOD Block Diagram 5-51
5-18 LP-11 Solid Waste Handling System 5-57
5-19 LP-11 Atla; Storage and Combustor Feed Subsystem 5-58
5-20 30-Inch Feeder Valve Installation 5-59
5-21 LP-11 Third Stage Separator Exhaust Temperature 5-61
5-22 LP-11 Fluid Bed Pressure Drop 5-68
5-23 LP-11 Superficial Velocity 5-76
6-1 Ruston TA-1500 Gas Turbine 6-4
6-2 Turbine Installation into the CPU-400 Pilot Plant 6-6
6-3 Switch Gear and Load Bank 6-9
6-4 Compressor Line Piping with Valves to the Fluid Bed
Combustor 6-11
6-5 Hot and Cold Pressure Piping Runs 6-12
6-6 Generator Shaft Braking System Schematic 6-17
xvi i
-------
FIGURES (Continued)
No. Page
8-1 Solid Waste Handling System Schematic (LP-1 and LP-8
Testing) 8-4
8-2 Solid Waste Handling System Schematic (LP-11 Testing) 8-5
8-3 Shredder and Air Classifier (LP-1 and LP-8) 8-7
8-4 LP-11 Air Classifier 8-10
8-5 Central Control Panel for Shredder/Conveyor/Air
Classifier 8-14
£_6 Photograohic Views of the Atlas Storage Unit
Interior 8-16
8-7 Atlas Outfeed Volume Sensor 8-19
8-8 16-Inch Air-lock Feeder Valve (LP-1 and LP-8) 8-23
8-9 Air-lock Feeder Valve, Modified for Dual 5-1 ich
Feed System 8-23
8-10 LP-8 Air-lock Feeder Valve and Drive Assembly 8-25
8-11 Typical 30-Inch Air-lock Feeder Valve and Drive Assembly 8-25
8-12 LP-1 Solid Waste Metering Control 8-27
8-13 LP-8 and LP-11 Solid Waste Feed Control 8-29
8-14 Sampling System Schematic 8-57
8-15 Non-Isokinetic Sampling 8-58
8-16 Particle Analyzer System 8-61
8-17 Initial Calibration Data Sheet 8-68
8-18 Recalibration Data Sheet 8-71
8-19 Dilution and Volume Factor Summary Data Sheet (Probe C) 8-73
8-20 Dilution and Volume Factor Calculations Data Sheet
(Probe C) 8-74
xvii i
-------
FIGURES (Continued)
Np^_ Page
8-21 Raw Data Particle Count Data Sheet (Probe C) 8-75
8-22 Dilution and Volume Factor Summary Data Sheet (Probe B) 8-76
8-23 Dilution and Volume Factor Calculations Data Sheet
(Probe B) 8-77
8-24 Raw Data Particle Count Data Sheet (Probe B) 8-78
8-25 Dilution and Volume Factor Summary Data Sheet (Probe A) 8-79
8-26 Dilution and Volume Factor Calculations Data Sheet
(Probe A) 3-80
8-27 Raw Data Particle Count Data Sheet (Probe A) 8-81
8-28 Particle Counting Data Reduction Sheet (1) 8-83
8-29 Particle Counting Data Reduction Sheet (2) 8-84
8-30 Preliminary Distribution Curves 8-85-
8-31 Efficiency Curves 8-89
8-32 Comparison with Background 8-90
8-33 Compensated Distribution Curves 8-93
8-34 Loading Curves 8-95
8-35 Loading Analysis Shcci. (Probe C) 8-96
8-36 Loading Analysis Sheet (Probe B) 8-97
8-37 Loading Analysis Sheet (Probe A) 8-^8
xix
-------
TABLES
N£. Page
3-1. Particle Loading 3-23
3-2. 16-Series Second Stage Separator Test Summary 3-36
?-3. 18-Series Second Stage Seoarator Test Summary 3-41
3-4. Sieve Analysis Data for Material Collected by
the Second Stage Separator 3-50
3-5. 20-Series Second Stage Separate'- Test Summary 3-53
3-6. Dust Test Data 3',-Inch Cyclone Cold Flow Tests 3-67
3-7. 20-Series Third Stage Separator Test Summary 3-71
3-8. 30 Mesh Sand Cold Flow 3-82
3-9. 16 Mesh Sand Cold Flow with Louvers 3-82
3-10. Hot Flow Test Data 3-84
3-11. Total Loading Filter Efficiency Data 3-85
3-12. Solid Waste Sample Data 3-98
3-13. Bed Material Size Distribution 3-99
3-14. Incipient Fluidization Velocity 3-99
3-15. Spectrographic Analysis (Semi Quantitative) 3-105
3-16. LP-Solids Mass Accounting 3-109
3-17. Analytical REsults of Combustor Exhaust Gar Analysis 3-120
3-18. Flyash Analysis Summary 3-122
4-1. Common Data Base Definition for LP-11 4-32
4-2. Task Identification Summary 4-37
4-3. Computer Software Time Sequence of Events 4-38
4-4. Task 10 Executive Commands 4.50
5-1. LP-1 Test Summary of Fuel Heating Value Results 5-13
XX
-------
TABLES (continued)
No.. Page
5-2. LP-8 Temperature Sensors 5-23
5-3. LP-8 Pressures and Fuel Flow Sensors 5-24
5-4. LP-8 Fisher Controller Setting 5-26
5-5. LP-8 Combustion Efficiency Resets, Heat Balance
Technique 5-30
5-6. LP-8 Combustion Efficiercy Results, Mass Balance
Technique 5-32
5-7. LP-8 Combustion Efficiency Results, Carbon
Combustion Technique 5-33
5-8. LP-8 Inert/Ash Material Accounting 5-35
5-9. LP-8 Gas Sampling Volumetric Basis 5-39
5-10. LP-P Particle Sampling 5-40
5-11. LP-"1 Temperature Sensors 5-52
5-12. LP-11 Pressure Sensors and Switches 5-53
5-13. LP-11 Fisher Controller Setting 5-55
5-14. , Control Temperature Statistics 5-63
5-15. LP-11 Exhaust Gas Composition (Volumetric Basis) 5-65
5-16. LP-11 Particle Sampling 5-67
5-17. LP-11 Inerts Accounting 5-69
5-18. LP-11 Combustion Efficiency, Heat Balance Technique 5-71
5-19. LP-11 Combustion Efficiency, Carbon Combustion
Technique 5-74
6-1 Turbo-Electric System Candidate Suronary 6-20
8-1. Test Summary Sheet 8-38
xxi
-------
ACKNOWLEDGEMENTS
Sincere appreciation is extended to Professor Arthur M. Squires of
the City College of the Citv University of New York for providing
the basic design of the subscale granular filter.
xxi i
-------
SECTION I
CONCLUSIONS
Shredding and air classification operations on municipal solid waste
produce a fuel form having very satisfactory physical properties for
energy recovery through combustiv. .n a fluidized bed reactor. A solid
waste handling subsystem with reliable components has been developed and
extensively operated.
The vertical fluidized bed combustor has been tested and proven to be a
highly efficient, easily fed, readily controlled reactor of simple de-
sign that is capable of utilizing low quality fuels.
Test results of low pressure operation of the waste-fired vertical flui-
dized bed combustor show an average combustion efficiency of 99 percent.
Sixteen mesh silica beach sand is an acceptable starter bed material.
The inert content of typical solid waste provides a natural bed makeuo
material that leads to a satisfactory steady state bed composition. As
a result, the need for elaborate bed maintenance or ?nti-elutriation
devices is minimized if not eliminated.
Test results to date show no problem with fluidized bed residue buildup.
Relatively large metal and inert particles entering the active bed
exoerience gradual oxidation or attrition to typical bed size oarticles
1-1
-------
and eventually are elutriated to be collected by particle separators.
No bed material aqqlomeratir-> is exnerienced in the U700 F to 1450° F
bed temperature range.
Gas ohase combustion above the fluidized bed has been reduced to accept-
able values without resorting to undesirable remedies such as extensive
internal bed structures, numerous fuel feed points, deeper bed, multi-
stage combustors, etc.
Combustor freeboard and exhaust system deposit problems due to the alu-
minum content of solid waste have been solved.
In low pressure testing such as conducted to date, problems cf long tenn
inertial separator performance degradation have been encountered due to
tube plugging, particularly in the final stage with very fine hot flvash.
Remedial development work will continue at the more promising high pres-
sure conditions associated with integration into the gas turbine cycle.
Exhaust gas sampling instruments indicate that the pollution control
effort required for the gaseous comnonents of the CPU-400 process can be
expected to be minimal.
No evidence of serious hot gas system material corrosion or erosion prob-
lems has been found.
The capability of the manual and computerized control systems to ade-
quately control low pressure pilot plant operation was successfully
demonstrated. No problems are foreseen for control of high pressure
pilot plant operation.
1-2
-------
SECTION II
INTRODUCTION
GENERAL
Combustion Power Inc. of Menlo Park, California har. successfully demon-
strated the completion of component desiyn and low pressure testing for
thp CPU-400 pilot plant. The CPU-400 is a gas turbine-combustion con-
cept that cleanly and cheaply burns municipal solid waste at elevated
pressure for the recovery of energy in the form of electrical power
while recovering valuable secondary materials in the form of metals and
glass. The electrical energy that can be generated for the orojected
1975 per capita solid waste production of 3.8 pounds per day is 14.2
percent of the combined electrical energy consumed by American house-
holds and commercial establishments, or 8.3 percent of the total elec-
trical energy oroduction of the United States that is projected for
975.*
T, e expected increase in solid waste generation, together with the capa-
bility of the CPU-400 to burn fast replenishable combustible materials
such as wood, and the critical need to develop new inexpensive energy
sources in order to lessen this country's dependence on oil, gives the
CPU-400 concept great growth potential as an important ancillary source
of energy generation.
*Values calculated from Hata presented in EPA Publication No. SW-122
2-1
-------
The work described in this report pertains tc tasks defined in contract
No. 68-03-0054 and was performed during the period from June 30, 1971,
to April 30, 1973, with funds provided by the United States Environmental
Protection Agency, Office of Research and Develooment, Industrial Environ-
mental Research Laboratory, Cincinnati, Ohio. It was follow-on work to
contract PH 86-68-198.which was completed on December 31, 1971.
OBJECTIVES
The ultimate objective of the CPU-400 project was to successfully demon-
strate the operation of a prototype plant that will cleanly and effi-
ciently dispose of solid waste at a fraction of the total cost of cur-
rent disposal methods while generating electrical power. The immediate
objective of the work performed during the contract period covered by
this report was to successfully demonstrate the pilot plant component
development and readiness for turbine integration through the perfor-
mance of extensive long duration testing.
ACHIEVEMENTS
Important achievements made during the contract period included: Final
hardware procurement and'low pressure testing of the fluid bed reactor,
three stages of particle separators, ash removal systems, process con-
trol computer, instrumentation, gas analysis equipment, 1/20 scale
model testing, and the procurement and design for integration of the
turbo-electric system including the compressor/turbine/genetator set,
switch gear, and load bank. This work culminated in the successful
2-2
-------
demonstration of the integrated operation of the solid waste handling
system, hot gas system, and automatic process control system at pre-
cisely controlled temperature conditions for 48 hours of continuous op-
eration. This test co. pleted the scheduled low pressure testing and
verified system readiness for initiation of high pressure testing.
During the low pressure, high temperature (1450 F third stage outlet
temperature) testing, multitube cyclone separator tube plugging prob-
lems and collecting hopper ash uiloa-.ng problems were encountered. In-
vestigations with the separator .nodules and full scale units are con-
tinuing to provide a successful solution without compromising separation
efficiency. The future higher pressure testing operation will be bene-
ficial to ash unloading via blowdown, will provide a higher separator
tube differential pressure, and will raise the Reynolds number above the
present level of 2000 under which tube-operational problems can be pre-
dicted. IP addition, an investigation of various separator internal de-
vices has indicated their suitability for keeping individual tubes from
plugging.
Work is presently in progress to integrate the turbo-electric system
with the hot gas system and process controls, and to ready the pilot
plant for high pressure testing in which the energy recovery concept
(the generation of electrical power with an open gas turbine cycle us-
ing solid waste as the only fuel), will be demonstrated. This work
is being performed in accordance with contract PH 68-03-143 and is now
scheduled for completion in 1975.
2-3
-------
PILOT PLANT DESCRIPTION
The present CPU-400 pilot plant (Figure 2-1) consists of four primary
systems: the solid waste handling system, the hot gas system, the
turbo-electric system, and the process control system.
The solid waste handling system prepares the solid waste material for
combustion. This includes shredding, separation of the shredded ma-
terial into combustible and inert components, storing the combustibles
until ready for use, and metering the solid waste to the combustor. Un-
processed municipal waste is iiitially loaded on thp shredder conveyor
by a skip loader. The conveyor modulates the feed to the shredder
based upon electrical loading of the shredder motor. After shredding,
the material is fed to the air classifier where the light, combustible
materials are pneumatically lifted and transported to an Atlas storage
unit while the heavy, inert materials drop out for subsequent separa-
tion and recovery. Metering of the prepared solid waste fuel is accom-
plished through a variable speed, servo controlled outfeed conveyor in
the Atlas storage unit and variable speed transfer and weighing con-
veyors which transfer the processed solid waste to airlock feeder
valves. The airlock feeder valves admit the solid waste to pressured
feed pipes which route the material to the combustor fluidized bed.
The hot gas system consists of the solid waste combustor, 3 particulate
removal stages, ash removal equipment, and interconnecting piping and
valving. After the solid waste is introduced into the combustor it is
burned and the resulting hot gases are then cleaned of susoended solid
2-4
-------
DUST COLUUOH
SHHC nnt. h
ClASSIHtH ^
SOLID WASTE PROCESSING
BAG HOUSE
FILTER
CONTROL ROOM
Figure 2-1 CPU-400 Pilot Plant
-------
material in the three stages of separation. Residue material is re-
moved from the particle separator collection hoppers by pneumatic
transport to external collection points.
The turbo-electric system, currently being installed, consists of a
gas turbine, a generator, switch gear, and a load bank. The compressor
section of the turbine supplies the cold air for the solid waste com-
bustor fluidization. The resulting hot gases, after being cleaned in
the separators, are used to power the compressor turbine and the power
turbine. The generator is driven by the power turbine and generates
power which is subsequently controlled by the switch gear. The elec-
trical energy output of this pilot plant system will be dissipated in
a combination load and light bank. In the full scale system, the
electrical power will be delivered to a customer for subsequent use in
the municipality. In the low pressure configuration operated to date,
the gas turbine compressor is replaced by a facility blower and exhaust
gases are cooled by water spray. The blower also supplies fuel trans-
port air to the fluidized bed.
The control system interacts with these systems to control their re-
soective outputs in response to commanded set points. This system,
which features analog controllers under the supervisory control of a
digital process computer, also monitors numerous signals to provide
data acquisition, logging, out-of-tolerance alarming, and status dis-
play functions.
2-6
-------
SECTION III
HOT GAS SYSTEM
The Hot Gas System consists of all equipment required to produce and de-
liver to the turbine a controllable, cleaned, hot gas product. Included
are a combustor, particle collection equipment, a material removal sys-
tem, gas analysis equipment, and all connecting pipi.ig, gauges, meters,
and related ancillary equipment. Several different configurations of
chis equipment were tested as new components were added to the system.
t
Comprehensive details of the major components or subsystems designed and
fabricated under this contract are outlined in the four major areas that
constitute this section. These are: "The Design and Development of the
Vertical Combustor and Support Equipment"; "The Design and Development
of Particle Collection Equipment"; "The Design and Development of the
Residue Removal System"; and "The Determination of Off-Gas Composition
of a Fluid Bed Combustor Burning Municipal Solid Waste".
DESIGN AND DEVELOPMENT OF THE VERTICAL COMBUSTOR AND SUPPORT EQUIPMENT
(TASKS LP-12A and LP-12B)
The vertical combustor was designed, built, and tested in the pilot
plant system for use with the Mid-Continent/Ruston TA-1500 gas turbine.
It was the result of exploratory development testing with the horizontal
combustor and provided an updated pilot plant operating philosophy which
was evolved since inception of the original design. A cross-sectional
view of the vertical combustor is shown in Figure 3-1. The vertical
combustor, as installed in its pit within the existing facility, is
shown in Figure 3-2.
3-1
-------
OIL GUN
TYP 6 PL
u1) 8'FOR 5" OR 6
SOLID WASTE
FEED PIPES
FIRST STAGE
SEPARATOR
ACCESS
PORT
I8"AT 315
MASONRY
FLOOR LEVEL
AIR DIFFUSERS
STAGGERED ON
5 3/4 SPACING
FIRE BRICK
(2) 8'BLINDHOLEi
OPPOSITE FEED
PIPE NOZZLES
AIR DlFFUSER (161)
OlSTR PL TC (2)
C FLUID BED 9" TC (6)
D FLUID BED BOTTOM T3 (I)
FEED PIPE HOLD-DOWN
BED PRESSURE (2)
-I8"(l) HIGH PRESSURE AIR INLET AT 0*
(I) LOW PRESSURE AIR INLET AT 45°
(I) MANHOLE AT 315°
Figure 3-1 Vertical Combustor
3-2
-------
Figure 3-2 Vertical Fluid Bed Combustor Assembly
3-3
-------
Improvements Over the Horizontal Combustor
The vertical combustor system was designed to minimize or °liminate cer-
tain problem areas identified during testing with the horizontal combus-
tor.
One of the principal problems of the horizontal combustor, shown in
Figure 3-3,-was elutriation of the bed material from the combustor unit.
The installation of bubble screens to reduce surface explosions of sand,
and various baffles and inertial devices in the combustion chamber,
were ineffective due to the hightemperature c"inker formation and the
violence of sand explosions, which deformed or otherwise caused the
devices to fail. Also, since the freeboard area decreased in size
toward the top of the vessel, sand particles were entrained in the flow
stream and pneumatically transported out of the chamber at a rate higher
than that at which sand was normally introduced with the solid waste
stream. This necessitated the addition of bed material by mechanical
means. Internal configurations evaluated during the testin-j of the
horizontal combustor are depicted in Figure 3-4.
In the vertical combustor design, elutriation is reduced by providing a
high freeboard volume in which the average gas velocity is approximately
6 ft/sec-below the terminal velocity of nominal bed material at turbine
temperature and pressure conditions. In addition, the vertical combus-
tor design reduced the tendency of the feedline air flow to direct sand
particles towards the exhaust port where they could b* entrained into
the exhaust gas stream and carried out of the combustor. The top cylin-
drical section of the combustor, which contains the exhaust penetration,
3-4
-------
L;
221-208
Figure 3-3 Horizontal Fluid Bed Combustor Assembly
-------
Figure 3-4 Horizontal Combustor Internals
3-6
-------
is designed to be removdble with its insulation to permit increase of
the freeboard height with minimum amount of rework.
A second problem encountered in the horizontal bed was that the height
of the static bed was limited to 2 feet due to the proximity of the
exhaust port. In the vertical unit bed depths up to 4 feet can be used
to obtain a higher portion of the heat release within the active part of
the bed, thus reducing heat release in the gas zone above the bed. The
latter improves potential utilization of the chemical characteristics of
b?d materials in suppressing undesirable gas constituents such as SOo
or HC1. Control of temperature differences between bed material and
combustor exhaust could also be improved by the increased bed depths.
A third problem which the vertical combustor minimizes is heat loss.
In the horizontal combustor, where the inner vessel would have been
cooled with the turbine compressor air, a local outer wall temperature
in the 600° F range would be anticipated for the 400° F compressor air at
rated power, with resultant high heat loss through radiation and convec-
tion. The refractory thermal insulation of the vertical combustor
will maintain a nominal skin temperature of 230° F. The refractory in-
sulation and its high heat capacity also minimizes heat loss from the
bed during static conditions (overnight shutdowns, maintenance cycles,
etc), and permits normal system startup without sand heating after
overnight shutdowns.
In the horizontal combustor, the preheat burner had to be located on
the centerline of the. concentric vessel due to differential thermal
3-7
-------
expansions. This resulted in contact between the burner flame and top
layer of the fluid bed material resulting in incomplete local combustion
and fusion of the adjacent sand ^ayer. This problem was resolved by
the vertical combustor, which permits the use of a high capacity burner
for quick sand heatup operation within two hours. It is located in the
top pressure dome of the vessel at a maximum.distance from the sand.
The internal view of the combustor, Figure 3-5, shows the burner block
at the center of the top dome and the exhaust pipe which extends into
the freeboard. This allows complete combustion and mixing of the burner
combustion products and burner excess air prior to hot gas penetration
into the bed, and minimizes burner fouling during fluiciized operation.
Improvements in the fuel and air feed systems were also incorporated in-
to the vertical combustor. In the horizontal unit, feedline penetration
had to be parallel to the combustor centerline; in the vertical unit,
penetration around the entire circumference at any elevation is possible
The round, flat, distributor plate in the vertical unit with the 161
individual sintered metal screen diffusers (Figure 3-6) also improves
distribution of fluidizing air over the dihedral-rectangular plate in
the horizontal bed. Internal leakage in the horizontal unit, which
allowed bypass of the bed by some of the fluidizing air, was entirely
deleted by the welded construction of the vertical combustor.
Vertical Combustor Description
The vertical combustor is a 9^ foot out^de diameter cylindrical carbon
steel pressure shell with dished heads insulated from the hot gases by
refractory insulation. The fluid bed is supported by a flat carbon
3-8
-------
BACKHEAT BURNER BLOCK
Figure 3-5 Corrbustor Exhaust and Backheat Burner Ports
3-9
-------
I
i
01
I
I
o
CT
CX
Tl
rD
IV
a.
-
O>
v>
-------
steel plate also insulated and welded to the pressure shell. Penetrat-
ing the insulated plate are 161 2-inch pipes that connect to sintered
wire stainless steel diffusers and seven 1-inch pipes for thermocouples
or pressure tap installations. Figure 3-7 shows the normalized dif-
fuser pressure/flow characteristics. The shell has penetrations into
the fluid bed for two solid waste feed pipes and six oil injection
points. Two additional nozzles were provided in the pressure shell to
allow future growth to four solid waste feed points.
Located on top of the combustor is the oil burner used to preheat the
bed material under low pressure conditions. This unit, rated at 1,500
cfm, maintains the output gas at 1500° F over a 4:1 turndown ratio. The
remaining 3,300 cfm from the rotary positive displacement blower flows
into the air plenum below the combustor distributor plate. With the
outlets from the third stage separator blanked off, the combustion gases
produced by the burner are thus forced to flow through the sand where
they are mixed with the blower excess air in the plenum below the dis-
tributor plcte and du.nped overboard through the 12-inch duct (valve 005)
that branches off the high pressure air supply line. The piping sche-
matic to the combustor, Figure 3-8, has .been included for valve and
valve nomenclature definition.
The preheat burner, a schematic of which is shown in Figure 3-9, con-
tains a propane secondary gas ignition system which provides a positive
means for igniting the diesel fuel. The propane is ignited by a spark
which is on for approximately 15 seconds. As soon as oil ignitinn is
3-11
-------
0.08
0.06
_ p
I I
g = GRAVITATIONAL CONSTANT
)j = VOLUMETRIC FLOW, cfm
iP = DIFFERENTIAL PRESSURE,
j = INLET PRESSURE, psia
F! = INLET TEMPERATURE, °R
A = ELEMENT AREA, inch2
R = GAS CONSTANT FOR AIR
psi
OJ
i
0.04
0.02
AP
/
A
0.1
0.2
0.3
0.4
0.5
0.6
(Q/A)'
60q(.RT1
x 10
221-209
Figure 3-7 Normalized Flow Characteristics
-------
OIL
PROPANE
co
i
FEEDLINE «1 AIR CONTROL VALVE
FEEDLINE *2 AIR CONTROL VALVE
FEEDLINE A1P SHUTOFF VALVE
BURNER AIP SHUTOFF VALVE
FLUIDIZING AIR CONTROL VALVE
BACK^AT EXHAUST/DUMP AIR CO'JTPOL VALVE
COMPRESSOR SHUTOFF VALVE
FLUID BED BYPASS CONTROL VALVE
(ADDED FOR LP-11)
005
TURBINE
COMPRESSOR
003
221-210
Figure 3-8 Vertical Combustor Air Valve Definition
-------
-------
obtained, the spark is then shut off to cvoid ultraviolet interference
with the flame detection system, illustrated in Figures 3-10 and 3-11.
This system provides the input signal to the on/off valves and the ig-
nition transformer, monitors system parameters, and shuts off the igni-
tion and main burner valves in the absence of a flame sensed by an ul-
traviolet scanner.
The ducting to and from the combustor includes valving to control the
air flow during the various operating modes of the combustor. During
fluidizatlon, a 12-inch butterfly valve (005) seals off the backheat
exhaust duct and the combustor exhaust flows through the three stages
of in-series particle separators. Duiing low pressure tests, the gas
leaving the last particle separator was exhausted directly into the
atmosphere. During high pressure operation, the gas will be ducted
into the turbine.
Fluidizing air for low pressure tests was supplied to the combustor
through a 14-inch line coming from the rotary positive displacement
blower. During high pressure tests the fluidizing air will be supplied
through an 18-inch diameter line coming from the turbine compressor.
Remotely controlled, air-operated valves (007 and 003, respectively) are
located in these line* to isolate the inactive mode of operation. Inlet
and outlet silencers were provided with the blower to reduce noise
levels to acceptable limits.
Located over the combustor and particle separation units are mono»
hoists for use in assembly and disassembly of the two top sections of
3-15
-------
PROTECTOHER
COMBUSTION SAFEGUARD
OJ
I
PROTECTOF1ER
MANUAL OPERATION
ACF
C-CHECK RELAY
F-FLAME RELAY
AIR PRESSURE SWITCH
(OPENS ON LOW PRESSURE)
OIL PRESSURE SWITCH
(OPENS ON HIGH PRESSURE)
OIL PRESSURE SWITCH
(OPENS ON HIGH PRESSURE)
INTERNAL WIRING
EXTERNAL WIRING
WIRING DIAGRAM
r MJV-OUM, - .
6L1- -o^,^-^ -L-^
T UV|"«-rr-|rrr^ -r>^ NOT USED
c[L f r r^ i F1AMF RLLAY m
INTERCONNECTION J D^ J ' COIL IS ENERGIZED
DIAGRAM BELOW G^ WHEN FLAME IS
I n-J Or-' rcTAPI KHFD AND
^
D.C. TEST JACK\ Wjd-__Df.TECTED
PH START
1 "pi 12 =|; 13 ir4 ( CHECK).-. ACF,
! ' 0,|o V 0|io o : IGNITION
CtT1 7 pi 1C7 ? cU JRANS s
"~^ ' 'SPARK --« ,*!; -VfVYV +-
' IGNITOR " ^ ^_^
" } --" (^>-o| (-o-f^^j^i-o MA I N
-----' i 8 "^f ^
^
FLAME ON -^y
(NEON LIGHT)^
i
1 PILOT
-^"VALVE
^
Figure 3-10 Flame Detection Schematic Diagram
-------
115V-60H
rU
AIR PRESSURE SWITCH
(OPENS ON LOW PRESSURE)
OIL PRESSURE SWITCH
(OPENS ON LOW PRESSURE)
OIL PRESSURE SWITCH
(OPENS ON LOW PRESSURE)
*
L2
INTERCONNECTION DIAGRAM
a
;]
D.C. TEST JACK
FLAME
o r 1
v- r i !
ACF
U
*D
(FLAME)
TRANS.
ACF
P.C."I"ULTRA-VIOLET SuANNLR
-OPERATING SIGNAL
FROM P.C."I" OR
FLAME ROD
IGNITION J"_RANS__
SPARK
IGNITOR
MAIN VALVE
PILOT VALVE
221-212
Figure 3-11 Flame Detection Interconnection Diagram
-------
the combustor and for removal of the top sections of the separators
during inspection cycles.
Six oil guns, similar in design to the units used in the horizontal conv
bustor, were incorporated into the vertical combustor to permit fluidi-
zation combustion of auxiliary fuel. These guns are spaced around the
periphery.of the combustor and penetrate 12 inches into the bed. Each
gun (Figure 3-12) includes mixing air and cooling air flow channels to
keep the injection pipes clear of fluid bed material and diesel oil,
which would otherwise crack under heat soak conditions and plug the
piping. Oil is supplied to the guns through a 100 psi system for low
or high pressure operation. Numerous access and work platforms were
installed around the units at points where maintenance, inspection, or
visual observation are to be accomplished.
After the refractory was installed and cured with the sand backheat
burner, a series of checkout tests and !,ot tests were conducted. The
combustor has provided trouble-free performance for over 150 hours of
high temperature operation with solid waste fuel in satisfying contract
test requirements. All design objectives were demonstrated. They in-
cluded: (1) good fluidization under low air flow conditions; (2) fast
2-hour sand heatup cycle; (3) reduced elutriation, thereby minimizing
bed depth decrease during solid waste burning; (4) reduced heat losses
to permit normal fluidized startup following overnight shutdowns;
(5) elimination of bed clinker formation during backheating c.nd of un-
controlled, high magnitude heat release in th^ comhustor freeboard.
3-18
-------
MIX AIR
I
v£>
PACKED FIBER INSULATION
OIL COOLING AIR
1
M
INSULATING
WEAR RESISTANT PFFPACTri°Y
OIL GUN ASSEMBLY
nu,
1?" (TYPICAL)-
/TYPICAL AIR
DIFFUSOP
PPESSiJPE SHELL
Figure 3-12 Typical Oil Gun Assembly
-------
Feeder Valve and Solid Waste Handling System
Processed solid waste fuel Is applied to the combustor by feeder
valve(s) in the Atlas storage and combustor feed subsystem. This sub-
system and the shredder/air classifier/pneumatic transfer subsystem,
which prepares the solid waste for combustion, comprise the solid waste
handling system and were developed under a previous contract.
The airlock feeder valves transfer the solid waste from the ambient
pressure zone of the Atlas storage and combustor feed subsystem con-
veyors to the pressurized solid waste feed piping connected to the com-
bustor. During initial testing with the horizontal combustor, (task
LP-1), a single, CPC designed, 16-inch feeder valve and a single 6-inch
diameter feed pipe were used for this purpose. This configuration was
changed for the LP-8 testing with the vertical combustor when two
16-inch feeder valves, each of which fed separate 5-inch diameter feed-
lines, were used. The 16-inch feeder valves and 5-inch feed pipes
were subsequently replaced with two 30-inch rotor feeder valves and
two 6-inch diameter pipes for the LP-11 testing series. These changes
and other changes made to the elements of the solid waste handling
system in order to satisfy the discrete requirements of the LP-1, LP-8,
and LP-11 tests are discussed in Section V. A complete description >
the solid waste handling system is contained in Appendix A.
3-20
-------
DESIGN AND DEVELOPMENT OF PARTICLE COLLECTION EQUIPMENT
The particle collection equipment described in this section consists of
three full-scale separator stages, designed and tested to determine
their capability to clean the combustor exhaust gas prior to its enter-
ing the CPU-400 pilot plant turbine, and a sub-scale granular filter
which was designed and evaluated as a backup cleaning dev;ce. The
first stage separator, designed and tested under contract task LP-12C,
was evaluated using the vertical combustor during LP-8 and LP-11 testing.
The second stage separator, a multi-tube cycl ne assembly containing
6-inch diameter inertial tubes, was designed under a previous contract,
and was evaluated with both the horizontal and vertical combustors
(contract tasks LP-1 and LP-8, respectively). The third stage separator,
a multi-cyclone separator containing 3Vinch diameter inertial tubes,
was designed under contract tasks LP-2 through LP-4, and was evaluated
with the vertical combustor during the LP-11 testing. The granular
filter, designed and tested under contract task SS-2, was evaluated
using the CPC Model 3 combustor.
Multi-Cyclone Tube Modules
The full-scale inertial separator design was based on testing of inertial
cyclone tube modules and separators with in-house model fluid bed com-
bustor systems. These tests were used to define tube configurations,
operating characteristics and performance at operating temperature,
volumetric flow, and ambient pressure conditions.
Reverse flow cyclone tubes are used in the multi-cyclone assemblies.
As shown in Figure 3-13, the dirty gas enters the inlet annular cavity
3-21
-------
CLEANED AIR
INLET GUIDE
VANES
- OUTLET TUBE
INL£T MR
CONE TUBE
PARTICLES
221-03
Figure 3-13 Inertial Separator Tube
3-22
-------
which contains eight spiral-shaped vanes. The vanes imoart rotational
motion to the downward flowing gas which produces a vortex and centri-
fuges particles to the outside wall. The partiJes spinning along the
wall decelerate and fall through the opening in the bottom of the tube
into the common collection hopper. The gas exits the cyclone tube in
an upward motion through the center of the finned tube. Particle sepa-
ration performance measured for the selected tubes during model tes:ing
is summarized in Figure 3-14. Typical total loading measurements with
the final tube designs (from model test run 13-43) are listed in Table
3-1.
Table 3-1. PARTICLE LOADING^
Sample set
i.
2
3
4
5
First stage in
0.12
-
0.14
0.23
0.18
Second stage in
0.067
0.16
0.087
0.11
0.089
Second stage out
0.021
0.047
0.03
0.035
0.024
gr/sc*
The staged cyclone efficiency operating downstream of the first stage
sand separator is sufficiently high to provide turbine life in excess
of 20,000 hours. Life expectancy is based on test data obtained from
Australian tests 2,3 burning pulverized bituminous coal. These tests
were conducted on a Ruston gas turbine, basically the same as the model
procured for use in the CPU-400 pilot plant. The techniques employed
-------
100.0
99.8
6" - 3-1/2" STAGE
MODULES
3 4 5 6 7 8 9 10
X, PARTICLE SIZE - micron
221-213
Figure 3-14 Inertia! Separator Fractional Efficiency
3-24
-------
for exhaust particle sampling, particle analysis, and data reduction
are described in Appendix B.
First Stage Separator (Task LP-12C)
A basic problem in the operation of a fluidized bed combustor fueled
with solid waste is the elutriation of fine sand particles (particu-
larly in the 16-100 mesh range) and the presence of molten airborne
metallic particles, primarily aluminum, in the hot exhaust stream.
These particles have been theorized to be small aluminum spheres pro-
tected from further oxidation by a thin, stable shell of aluminum
oxide. The particles are of sufficient mass to become separated from
the hot gas stream high velocity areas. Upon impact with component sur-
faces, the aluminum oxide shell is destroyed but is immediately re-
formed due to the contact of the aluminum with the oxygen-rich hot gas.
With time, the continual repetition of this process results in the
formation of large, hard clinker type deposits on the component surfaces.
Early inertial module tests indicated that these materials would create
heavy loading, excessive wear, and material deposit and removal prob-
lems in both multicyclone separators, and that some intermediate sepa-
ration device would be necessary to de-entrain the larger par'.icles.
To this end the first stage separator (sand separator) vessel was de-
signed.
The sand separator vessel, shown in Figure 3-15, is a carbon steel
pressure vessel, 8 feet in diameter and 15 feet 4^ inches high. It has
a 3/8-inch thick carbon steel vessel wall that is insulated by 3 inches
3-25
-------
SPINNER DUCT
MINERAL
FIBER
KAOWOOL
INLET AT 225°
INSPECTION AT 90° ->
n
>
V* /I
Ktf Xi
t
1
1
I
OUTLET AT 0°
221-214
Figure 3-15 First Stage Separator
3-26
-------
of Kaowool and 2 Inches of mineral fiber. The Insulation 1s protected
against erosion by a 3/32-Inch stainless steel (310) sheet that 1s
held In place with studs welded to the pressure shell. The upper por-
tion of the vessel is a removable dome that can be easily lifted by a
chain hoist when overhauling or repairing the vessel liner 1s required.
The first stage separator 1s connected directly to the combustor, and
due to the simplicity of Its design, it was not tested separately, but
was Installed on-line and Its performance judged as an Individual para-
meter within the total system.
Routine analysis of performance Indicated that 70 percent of the total
sand and flyash particulate stream was de-entrained in this separator
vessel and was continually removed through a removal port located in
the bottom of the vessel. A stainless steel screen cone prevented plug-
ging of the removal port by large clinkers formed in the spinner duct
through which the entering gas passed. The spinner duct successfully
eliminated the formation of large clinkers at the Initial gas Impact
point which had been previously encountered during pilot plant and
model system tests with flat Impact/deflection plates.
During LP-8 testing, separated material settled at the bottom of this
vessel and was not removed. Post-test Inspection revealed that fusion of
this static material occurred. To alleviate this problem, a continuous
materl&i removal system, discussed In "Design and Development of the Re-
sidue Removal System" under task LP-5, was used to remove material from
the bottom of this unit. Operation of the material removal system and
1-27
-------
the elimination of ash/sand agglomerations were successfully demonstra-
ted during subsequent testing.
Second Stage Separator
The second stage full scale separator assembly (Figure 3-16) consists of
an inertial separator tube holder with a residue collecting hopper, an
insulated carbon steel pressure vessel with inlet and outlet flanges,
a dished, insulated flanged head, and stainless steel liners. The iner-
tial separator tube holder and hopper shown are fabricated from 310
stainless steel. The hot gases from the first stage separator enter the
second stage separator inlet cavity to the 48 6-inch cyclone tubes (24
made of 310 SS and 24 made of 601 Inconel) through an internal circum-
ferential passage. The entering gas then turns and flows through the
spirally finned annular section of each tube. The configuration of the
tubes investigated in the full scale separator are shown in Figure 3-17.
Particulates are removed from the separator hopper through an opening on
the bottom of the hopper cone. Tube shown in Figure 3-17(c) were se-
lected for CPU-400 pilot plant use.
The outlet pipe and conical tube section of each inertial are clamped
Into position against the bulkheads of the tube holder. Packed ceramic
insulation is used to provide seals for the individual tubes. The
housing of the separator tube assembly is held to the cylindrical pres-
sure shell by stainless steel supports. These supports are designed to
permit the radial differential expansion that occurs when the tube as-
sembly temperature exceeds the wall temperature of the insulated carbon
3-28
-------
Figure 3-16 Second Stage Multicyclone Separator Tube Assembly
-------
REDUCED CONE ANGl f
GJ
o
ORIGINAL
(/
-
C
-- 6
r
REDUCED CONE LENGTH i
*) (B)
*
1
' 1
'
CM
O
CM
r
i
,
i
t
r
-^
^
,,ol ' \
35 -j|~« V
1 t
i
,
i
I
r
8 FIN'
A,
r/ 1 , c
\ 2.5
L .
1 1<
i 1 a
1
1
*
t
I
-, 3.75 .-
riCAL VANE TUBE
^.FIGURATION
^ 1.9 -
i.;
Figure 3-17 Six-Inch Diameter Tube Conn-u rat ions
-------
steel housing. This housing has an outside diameter of 8 feet and is
insulated with a 4-inch thick, 8 lb/ft3, high temperature ceramic wool
blanket that is protected from erosion by a thin stainless steel liner.
Stand offs are used to fasten the steel liner to the vessel wall.
As a result of the early testing, two pneumatically operated roller
style vibrators were installed on the hopoer cone (180 degrees apart)
to prevent bridging of separated material in the cone. Bridging of
materials prevents unloading of ash and produces eventual plugging of
the individual tubes. The vibrator installation is shown in Figure
3-18. The vibrator is encapsulated with insulation and the air which
operates the vibrator also provides for its cooling. The air exhausts
from the vibrator housing assembly through a passage provided in the
mounting stand-off and is mixed with the combustion gases in the plenum
of the separator assembly. The vibration is applied to the hopper skin
through a 3-inch channel. The vibrator is able to provide frequencies
between 8000 and 20,000 cyles per minute and forces up to 500 pounds.
The piping to the vibrator is insulated with ceramic fiber rope and
blanket to minimize the temperature of the air to the vibrator.
For continuous removal jf separated material, the ash removal and hand-
ling system shown in Figure 3-19 was connected to the ash discharge port
during all low pressure testing.
Inertia! Separator Testing with the Horizontal Combustor (16-Series Tests)-
These tests were performed to evaluate the operation of the inertial
separator and ash removal system under conditions generated by the hot
off-gases from a low pressure fluid bed combustor burning municipal
3-31
-------
IDENTICAL
NSTALLATION
THIS SIDE
t
2" DIA. STAINLESS
STEEL STANDOFF
MATERIAL
OUTLET
PORT
STAINLESS STCLl. ChAiJ^EL
PACKED CERA''IC
FICER INSULATOR
THIN-WALLED
STAINLESS STEEL
VIBRATOR
THERMOCOUPLE
T/C EXTENSION
> WIRE
- AIR SUPPLY DUCT
INSULATIOri
2L'1-01
Figure 3-18 Typical Separator Hopper Vibrator Installation
3-32
-------
PUFFBACK
ARM DRIVE
EXHAUST
BLOWER
i
CO
FACILITY
WATER
r- INERTIAL
\ SEPARATOR
\VESSEL
BALL VALVE
WITH
POSITIONER
FINNED
TUBE SIGHT
TUBE
VALVE
ASH REMOVAL TEE
FLOW
METER
PUFFBACK
BLOWER
MOVABLE
STORAGE
BIN
Figure 3-19 Ash Removal and Handling System
-------
solid waste. As seen in Figure 3-20, the separator was directly coupled
to the horizontal combustor. Specifics of the tests are described in
Section V.
The particles were sampled by a six-nozzle rake, downstream of the se-
para*or, with.each nozzle sensing an equivalent duct area. The ratio
of duct area to nozzle areas is 5000:1. For these tests, an impact
wear plate was installed at the inlet impact surface of the separator
housing assembly because of the high sand elutriation that existed
during fluidized operation in the hori2ontal combustor system.
The results of the tests are summarized in Table 3-2. Plugging of in-
dividual cones were observed at the end of all runs. These plugs pri-
marily resulted from the excessively high temperature encountered in
the separator and from the inability to remove material from the common
collecting hopper. The high elutriation rates from the combustor and
aluminum deposits on the individual tubes dictated the necessity of the
hot gas system redesign to the present configuration. Particulate data
did indicate the capability of the separator to remove flyash. Also,
higher velocities and larger ash outlets were beneficial in avoiding
pi gging. The ash removal system was capable of handling all materials
coming from the separator ash outlet during all tests.
Testing with the Vertical Combustor (18-Series Tests)-
The final pilot plant system containing the vertical combustor, first
stage sand separator, and second stage multi-element, reverse flow
cyclone tube separator is shown schematically in Figure 3-21. The tubes
3-34
-------
OJ
Ul
FIRST STAGE
INERTIA! SEPARATOR
HORIZONTAL
COMBUSTOR
EXHAUST
STACK
PARTICLE SAMPLING
PROBE
Figure 3-20 Separator Assembly Installed in Horizontal Combustor System
-------
Table 3-2. 16-SERIES SECOND STAGE SEPARATOR TEST SUWARY
Test
No.
16-la'c
16-2* »c
16-3a>c
15-4a'C
Inlet flow
(acfm)
13,600
(285/tube)
13,600
(285/tube)
15,800
(340/tube)
13,600
(285/tube)
Inlet temp.
1580 nom.
1820 max.
1575 nom.
1780 max.
1550 nom.
1840 max.
1550 nom.
1750 max.
Fluidized
operating
tjme (min)
174
386
175
292
\.
Material
collected
(Ibs)
1700 from
baghouse
3800 from
baghouse
2830 from
baghouse
1300 from
inertia!
hopper
.
Outlet particle
loading (gr/scf)
Not sampled
Sample #1: 0.12
#2: 0.08
13: 0.10
#4: 0.28
Not sampled
0.08
Comments
Exhaust appeared clean.
Exhaust appeared clean.
Particle loading indi-
cated tube plugging to-
ward end of test run.
Cones of 15 tubes plugged.
Cones of 6 tubes par-
tially plugged. Some
of the plugs in the
cone ends were hard ag-
glomerated ash deposits.
Collecting hopper was
ful 1 up to wi th in C in.
of tube ends . Part i al
plugs appeared to IIP
eroding .
Cones of 5 tubes were en-
tirely plugged, partial
plugs in cone tubes of
31 cyclones. Many of
these partial plugs ap-
peared to be eroding.
Partial plugs were hard,
light brown ash agglom-
erations.
CO
I
GO
cr>
-------
Table 3-2 (continued). 16-SERIES SECOND STAGE SEPARATOR TEST SUMMARY
Test
No.
16-5b'c
16-6b'd
Inlet flow
(acfm)
13,600
(379/tube)
13,600
(378/tube)
Inlet temp.
(°F)
1500 to
1650
1500 nom.
1750 max.
Fluidized
operating
time (min)
38
133
Material
collected
(Ibs)
.
^
Outlet particle
loading (gr/scf)
Not sampled
Not sampled
Comments
Post-test inspection
showed that cones of 2
cyclone tubes were
plugged and 5 were par-
tially plugged. Test was
aborted because of rombus
tor problems.
3 cones were partially
plugged.
CO
OJ
a 48 active tubes were used in tests 16-1 through 16-4.
b
36 active tubes were used in tests 16-5 and 16-6.
c See detail (a), Figure 3-17, for tube configuration used in tests 16-1 through 16-5.
See detail (b) , Figure 3-17, for tube configuration used in test 16-6.
-------
CO
^J
00
SAND
HEATING
BURNER
SYSTEM
STACK
F
PS
PF
SS
A = MEASUREMENT LOCATION
T = TEMPERATURE
FLOhl
PRESSURE SWITCH
VELOCITY HEAD
STATUS INDICATORS
POSITION INDICATORS
PARTICLE SAMPLING
PROBE
-TO
BAGHOUSE
(T) COMBUSTOR
(T) FIRST STAGE SEPARATOR
(T) SECOND STAGE SEPARATOR
(7) AIR DISCHARGE SILENCER
(T) AIR COMPRESSOR
(t) AIR INTAKE SILENCER
221-47
Figure 3-21 Test Setup Schematic for IS-Se-ies Tests
-------
in the separator assembly were modified to the (c) configuration shown
in Figure 3-17. Only 24 active cyclone tubes were utilized; their
arrangement is shown in Figure 3-22. The outlets of the non-active
tubes were removed to provide better flow manifolding to the active
tubes. The resulting holes in the outlet bulkhead were plugged with
special sealing adaptors. A baffle plate was welded on top of the
torroidal manifold at the inlet area to avoid possible high flow and
high particle loading in tubes adjacent to the inlet connection.
The 18-serles tests were utilized to meet the requirements of the 24-
hour duration process controllability test of Task LP-8. Seven tests
were conducted in this series. Test particulars and results are sum-
marizeo in Table 3-3.
Exhaust particulate size distribution measurement data ;ken during
runs lfl-1, 18-2, 18-6, and the first sample of run 18-7 (prior to tube
plugging) are shown in Figures 3-23 through j °5. In addition to the
standard sieve analysis for the flyash in the as.i bin, a Coulter counter
particle distribution analysis was done on ash collected in Tests 18-3
and 18-6. These data are summarized in Figures 3-27 and 3-28. Sieve
analysis results for the above two samples are shown in Table 3-4.
The testing analysis indicated that although longer trouble-free opera-
tion of the separator was attained, the collecting hopper unloading and
individual cyclone tube ash outlet plugging problem still sx^ed. As
a result, design modifications directed toward positive emptying of the
collecting hopper were instigated in order to establish whether the
3-39
-------
\
\
ACTIVE TUBES
INACTIVE TUBES
-
ooo
uooxo
\
\
I ''/ /
0°' °0°
googo
uoou
221 -48
Figure 3-22 First Stage Inertial Separator Tube Layout for 18-Series Tests
3-40
-------
Table 3-3. 18-SERIES SECOND STAGE SEPARATOR TEST SUMMARY
Test
No.
18-1
18-2
18-3
18-4
Inlet flow
(acfm)
15,100
(630/tube)
15,100
(630/tube)
15,100
(630/tube)
13,400
(559/tube)
Inlet temp.
(°F)
1500 nom.
1815 max.
1550 nom.
1700 max.
1550 nom.
1780 max.
1350 nom.
1490 max.
Fluidized
operating
time (min)
282
176
176
161
Material
collected
(Ibs)
162 ash
bin
170 iner-
tial hop-
per.
_
_
Outlet particle
loading (gr/scf)
Sample 11: 0.104
#2: 0.063
0.073
_
_
Comments
Tubes were clean except for
minor fine flyash buildup
in 2 tubes. The fdin col-
lecting hoooer did not un
load completely. Remain-
ing in hopper were 170 Ib
Not inspected, but the iner-
tial assembly hopper did
not entirely empty during
the test. This material
was removed with the exist
ing pneumatic removal sys-
tem prior to fluidized op-
eration of 18-3.
This test was run to verify
gain settings on automatic
control system stability .
This test was run to verify
the controllability and
automatic control system
dynamics for the bed tem-
perature control .
-------
Table 3-3 (continued). 18-SERIES SECOND STAGE SEPARATOR TEST SUMMARY
Test
No.
18-5
18-6
18-7
Inlet flow
(acfm)
13,700
(572/tube)
15,400
(641/tube)
15,100 for
29 h 40 m1n
11,800 for
6 h 20 min
Inlet temp.
(°F)
1350 nom.
1460 max.
1530 nom.
1600 max.
1520 nom.
1585 max.
Fluidized
operating
time (min)
152
111
36 hr
Material
collected
(Ibs)
_
963 ash
bin & 35
sep. hop.
These wts.
are totals
for runs
18-2 thru
18-6.
2224 ash
bin & 783
sep. hop-
per.
Outlet particle
loading (gr/scf)
.
0.077
Sample #1: 0.057
#2: 0.14
#3: 0.35
#4: 0.33
#5: 0.4C
#6: 0.31
#7: 0.18
#8: 0.62
Comments
The primary function of this
test was to evaluate the
solid waste feed control
system.
2 cone tubes had minor ash
deposits on cone end. All
tubes had thin ash depos-
its in cleaned gas outlet
tube.
The official test time for
the 24 hr test run started
12 hrs after initial flu-
idization. Data collected
prior to this time was not
processed. Particulate
data was taken in 3 hr. in-
tervals, starting at 15 hrs
after initial fl ji dization.
Post-tfc.t inspection re-
veal fd that the inertia!
tijfap assembly hopper did
not unload at the rate ma-
terial was received. This
resulted in
OJ
I
-------
Table 3-3 (continued). 18-SERIES SECOND STAGE SEPARATOR TEST SUfHARY
Test
No.
18-7
cont.
Inlet flow
(acfm)
Inlet temp.
(°F)
Fluidized
operating
time (m1n)
Material
collected
(Ibs)
Outlet particle
loading (gr/scf)
Comments
material backup into the
cones of the individual
tubes. A passage was
formed through the center
of the ash in the cone pro
viding a passage for un-
loading some of the mater-
ial.
Backing up of material into
Individual tubes started
to occur between the 1st
and 2nd samples. 10 cone
tubes were completely
plugged. The ash-free
appearance of their out-
lets indicated complete
particle pass-through. 14
tubes had ash deposits in
the cleaned gas outlets,
but continued separating
particles .
u>
I
-------
UJ 10? =
1/1
Q
E 106
8
i_>
ee.
«t
CL.
-t - 4=rt
SAMPLE TAKEN FROM INERTIAL SEPARATOR EXHAUST
6 8 10 12
X, PARTICLE SIZE - micron
Figure 3-23 Particle Count Analysis Plot, Test 18-1
20
221-49
3-44
-------
10*
108
10'
o
UJ
o
10C
CC.
O
<
0.
I r , ,,.,,, t-w
1 SAMPLE TAKEN FROM INERTIAL SEPARATOR EXHAUST
i i-i
-1---
103
102
1 2
6 8 10 12
X, PARTICLE SIZE -micron
14
Figure 3-24 Particle Count Analysis Plot, Test 18-2
20
221-50
3-45
-------
UJ 107
5 ,06
<
UJ
Of
SAMPLE TAKEN FROM INERTIAL SEPARATOR EXHAUST
3 in5
1 2
6 8 10 12
X, PARTICLE SIZE -micron
20
221-51
Figure 3-25 Particle Count Analysis Plot, Test 18-6
3-46
-------
109
108
UJ 107
1/1
o
10e
-------
10s
108
-j 10'
(Sl
o
10C
3 io5
10"
SAMPLE TAKEN FROM
BAGHOUSE
---- t t
T- t I
* - r
MM
t I + T
- -» + - -
-
f t « t
f t »
-»- 4 + r *
* t * t *
' t f
1 2
6 8 10 1? 14
X, PARTICLE SIZE -micron
20
221-67
Figure 3-27 Flyash Particle Count Analysis Plot, Test 18-3
3-48
-------
10"
Ul 107
in
a
3
VJ
105
10*
103
102
1
j
fe^sl"
1
----- -
(
- ^
V
_. __ ..
i
4
, 1
- t --T-
-
h H
- '
'
. -
j
J
1
._.
H
t ~*~
1
1
5s,
!=:-
*
- - 4
3
*
'
" f "
1
N
r-
_..
-
_
f_ ._. .
-
L ~~
i
\ '
i
- -
t t "~"
:- --t-_
t
ET:
tr I
>- t
>- t-
'
-
>-
r
^
^--f- r
^ - -t -- t- i
"^*4*^j
- -» 4
i
Li ...
+ -
- * -i
f
.. ,
<
»
f:
K--
-
^ -
*
; i
!r -r- :
i
t
--"
4
_ j
l
t
I SAMPLE TAKEN ROW
T BAGHOU5E
h t
» I
*
» f *-
*" t t
^ r r
+ *-
t - *
i » '
t- -i
i
!* t
' : I
i t *
t » *
i- t
P 1 j t
- { : ' i
"t" I I
- r t- -|
t
t- t v
1
i - »-
- - -t lr
'
I
i
t
"t
+
- t
i
J - ..-:.:
i
i
»
'
t
*
i
; - - -
1
1
-j- -- - - -
*. . _
I_
P
1 2
6 8 10 12
X. PARTICLE SIZE - micron
20
221-68
Figure 3-28 Flyash Particle Count Analysis Plot, Test 18-6
3-49
-------
Table 3-4. SIEVE ANALYSIS DATA FOR
MATERIAL COLLECTED BY THE SECOND STAGE SEPARATOR
Sieve size
(U.S. Standard)
18
20
25
30
50
100
100+
Total sample weight
Particle size,
18-3 Sample
0.6
0.3
0.2
0.2
1.8
2.6
94.3
8T2.5 gm
Percent
distribution,
18-6 Sample
0.1
0.1
0.1
0.1
0.6
1.4
97.6
406.6 gm
3-50
-------
tube plugging resulted from separated deposits of ash in the cones of
the cyclone, or if it was associated with material backup from the
collecting cone into the individual cyclones.
Testing with the Vertical Combustor (20-Series Tests)-
The second stage separator was operated during the 20-series test ir ex-
cess of 110 hours. Durinq those tests, the functional ooeration of the
assembly was to be verified. The test setup is shown schematically in
Figure 3-29. and the results are summarized in Table 3-5. The tube
configuration used during these tests is depicted in Figure 3-17(c).
The tube arrangement is shown in Figure 3-30.
During the initial testing it was found that the ash removal line at
the hopper outlet plugged and the periodic plugging of individual tube
cones occurred. The ash outlet of the collecting hopper was modified
after the 20-4A test resulting in successful ash unloading during
subsequent tests.
After the hopper unloading problem was solved, it was found that plug-
ging of individual tubes still occurred and thus the plugging was not
entirely associated with initial backup into cones due to the hopper
problem or due to absence of bleed flow. Following a series of model
tests in which rattlers were used to wipe the cones continuously, chain
and rod rattlers were incorporated into the full scale configuration
and tested. No deposits were found in the tubes where this chain type
of device was installed.
3-51
-------
SAND
HEATING
BURNER
SYSTEM
T
f
PS
PF
SS
STACK
PARTICLE SAMPLING
PROBE
MEASUREMENT LOCATION
rEKPEKATIIRE
FLOW
PRESSURE SWITCH
VtLOCI TV HEAD
STATUS INDICATORS
SOLID
WASTE
OIL
^MIXING AIR
[TYPICAL
[ 6 PLACES
f AA
PANE
PROPANE
OIL
TANK
COMBUSTOR
FIRST STAGt SFPARATOR
SrCOND STAGE SEPARATOR
AIR DISCHARGE SILENCER
AIR COMPRESSOR
AIR INTAKE SILENCER
THIRD STAGE SEPARATOR
221-98
Figure 3-29 Test Setup Schematic for 20-Series Tests
-------
Table 3-5. 2C-SERIES SECOND STAGE SEPARATOR TEST SUMMARY
Test
No.
20-2A
& -2P
20-2C
20-3
20-4A
Nominal
flowrate
through
assembly
(lb/min)
350
350
350
350
Nominal
Inlet
temp.
(°F)
1500
1490
1486
1440
Fluldlred
operating
time on
sol Id waste
(hrs)
3
3.5
12.2
33
Active
tube
pattern
(Figure
ref.)
3-22
3-2?
3-22
3-30
Comments
Airline connections to vibrator body failed on
both vibrators. Separated ash removal system
failed to maintain ash hopper clear of materi-
al. Airlines to the vibrators were redesigned
and reworked to Incorporate flexllnes.
Vibrators operated satisfactorily during the en-
tire test. Ash removal tee was olugged and
ash accumulated in the hopper. Kaowool and
small ash agglomerations were found in the
tee. After cleaning, the ash removal system
performed satisfactorily.
Two of the 24 separator tube cones were plugged.
The ash removal tee plugged. Tube pattern
chanoe for subsequent tests implemented.
Ash removal from hopper was unsuccessful despite
efforts to keep the tee open. Post-test In-
spections showed that the hopper was full
(460 Ib) of separated material. The level of
material in the hopper was in the proximity
of the individual cone outlets. All cone out-
lets were found to be plugged.
-------
Table 3-5 (continued). 20-SERIES SECOND STAGE SEPARATOR TEST SUMMARY
Test
No.
20-4B
20-5
Nominal
flowrate
through
assembly
(lb/min)
300
350
Nominal
inlet
temp.
(°F)
1430
1475
Fluidized
operating
time on
solid waste
(hrs)
50.6
8.2
Active
tube
pattern
(Figure
ref.)
3-30
3-30
Comments
Prior to this test the ash outlet was modified
by increasing the hopper outlet diameter to
5.8 inches and by adding a secondary collec-
tion hopper to the side of the tee. A ramrod
was incorporated into the se.up to push de-
posits into this secondary hopper.
Post-test inspection showed that thp hopper was
free of ash and two c /clone tube outlets were
nluqqfj. The lower flowrate was obtained by
overboard dump of 50 lb/min of compressor air.
tlechanical rattlers were installed in 10 tubes.
7 were chain rattlers and 3 were 'od rattlers.
They were pivoted at the exhaust plane of the
vane tubes.
Post-test inspection showed two tubes (without
rattlers) were plugged; the tubes with the
rod tvpe of rattler had ash deposits at their
outlets. The open area was approximately if 5
percent of its initial value.
-------
ACT!Vt Tlllli-,
UI/'U.TIVATCO TUBfS
221-215
Figure 3-30 Six-Inch Diameter Tube Arranaement of Second Staae Separator
3-55
-------
At the conclusion of the 20-series tests, the 6-inch inertial separator
was disassembled and inspected. The result of the inspection indicated
that the seals between the dirty gas aisle and clean gas aisle performed
well, with little evidence of leakage. There was a moderate buildup of
aluminum on the outer diameter surface of the inertial tube exhaust
piping, which is exposed to the gas as it enters from the first stage
separator. Figure 3-31 shows a buildup of flyash agglomeration in the
same area of the inertial separator which was precipitated by the aero-
dynamic stagnation areas developed in the area of the vanes. Consulta-
tion has been initiated with a view to effecting a solution for the ash
buildup in the area of the vanes.
Inspection of the shingles, which make up the inner liner of the vessel,
showed no sign of deterioration from either corrosion or erosion. The
wall temperatures indicated that the Kaowool insulation was satisfactory
in holding the vessel wall to a reasonable (130° F to 180° F) range.
Third Stage Separator
The third stage inertial separator assembly is shown in Figures 3-32
and 3-33. Its design, fabrication, and testing were performed under
Tasr.s LP-2, LP-3, and LP-4, respectively. The separator assembly con-
sists of a carbon steel pressure shell that is insulated by a 2-inch
thick layer of mineral wool and a 3-inch layer of high temperature
Kaowool. The insulation is protected from erosion by the high
velocity exhaust gases by a 310 stainless steel liner that consists of
two dished end closures and a cylindrical center section. The insula-
tion is held in place by studs welded to the pressure vessel wall.
3-56
-------
Figure 3-31 Ash Deposits in Vane Section of Second Staqe Seoaretor Tube
3-57
-------
TUBE
SUB-ASSEMBLY
STAINLESS STEEL
LINER ' .ii r;.)NM
FIBER
1
t
t- *
FROM
FIRST
STAGE
SEPARATOR
1
1 '
1
Ki-.j.
. ; MOUNTING PAD
^ J * \ r 1 /' ! !6('LAC(s;
N ^ ^ f
- ;
j '
. : . ^
" 1
-BELLOUS
ASSEMBLY ~-
MOUNTING PADS i
',4 PLACED L_
1 ,
!-
OUTLET
s (TYPICAL
3 PLACES)
v / e 7-
\ / 1 '
BELLOWS
».tt, I
ASH TRANSPORT
AIR
BAGHOUSE
221-02
Figure 3-32 Third Stage Inertial Separator Assembly
>-58
-------
Figure 3-33 Third Staue Cyclone Tube Assembly
3-59
-------
A bellows assembly was used in the dirty gas inlet port liner to pre-
vent leakage of particles past the inertial tube assembly and to pro-
i
vide the flexibility necessary to absorb the differential expansion
associated with the heating and cooling of the inner components. Two
of the three 26-inch diameter flanges in the lower cylindrical section
provided ras exhaust into the overboard stack during low pressure
operation and into the compressor turbine during high pressure opera-
tion. The third flange provides a hot gas inlet from the turbine com-
bustor. During operation, the ports not required are blanked off.
The inner surface of the carbon steel pressure shell is protected from
corrosive acid attack due to water condensation in the exhaust gas en-
vironment with a coating of bisonite phenoflex, the identical coating
system used in the vertical combustor and first stage separator. The
outer surface of the vessel is coated with high temperature paint.
The inertial separator tube subassemoly contains 100 3^-inch diameter,
310 SS tubes whose design i-s based on the testing of a three-tube
module. The tubes are arranged in 20 sets of four-tube modules which
form radial rows around the outer periphery. Mounted in the inner sur-
face are four three-tube modules and four two-tube nodules. Fiberfrax
rope packings seal the cone assemblies and outlet tube assemblies in
their respective bulkheads. The outer housing of the assembly and the
inner housing containing the tubes are eccentric and this forms an
annular passage that provides good flow distribution by minimizing
3-60
-------
velocity variations around the periphery. Six mounting pads are pro-
vided between the pressure vessel and tube housing assembly to reduce
local loading and stress concentrations in the thin section tube housing.
The cones of the inertial tubes discharge the separated particles into
a common collecting hopper from which the material is pneumatically
transported into the baghouse assembly through a 1-inch diameter pipe.
The cyclone tube configuration is defined in Figure 3-34. As on the
second stage separator hopper, two vibrators are installed on the un-
loading cone to prevent material bridging.
Cold Flow Testing-
Cold flow tests were conducted by the Donaldson Company in Minneapolis,
Minnesota with the setup shown schematically in Figure 3-35 and pic-
torially in Figure 3-36. The air flow resistances across the inftial
inlet and the separator assembly are shown in Figure 3-37. Flow distri-
bution measurements were made by measuring the air flow rate (velocity
head) exiting through each cyclone tube. These results are summarized
in Figure 3-38.
During the partic'e collection efficiency tests, the air flow was set
and maintained at 15,000 cfm. Standard Ar: fine test dust was injec.ed
Into the unit at a rate of 0.025 gm/ft of air flow (375 gm/min) through
a 30-gallon barrel dust feeder. A total of 98 pounds of dust was fed
in a 2 hour period. The test data is cited in Table 3-6.
Upon completion of the test, a weight balance was performed by weighing
the amount of dust fed (14,492 gm),dust on the absolute filter (8093 gm),
3-61
-------
A I
T
54
i
in
o
o
2.1 ID
VANE SECTION
(8 VANES)
\ ~ J
I 2.2 -
221-08
Fiqure 3-34 Third Staqe Cyclone Tube
3-62
-------
MA'lOMf TI k
RLOWFR
MANOMF Tl RS
\
FLOW
MCTLK
I'OlfU
I OR DDL T 1HI,
AND I Nil T I"1.' :
MAriOHl TIP
(I -(I
ABSOLUTE
ruitK
HOLUIK
DilLTING
Mf.NIIM
\
X
s^
\st
OllTl.iT 11 SI
PLENUM
/ D \
.
MANOMETLR
Figure 3-35 Airflow Resistance Test Setuo
-------
Figure 3-36 Overall View of Collector and Setup
3-64
-------
-30.0
,-fr . i
Figure 3-37 Ambient Airflow Ro; istanco. Thir-i '-trine
3-65
-------
OJ
I
^ w (a) .(fr -XVW^
VJ.V (T^\ fit\ vl) (Ti) x,-\ NA j
5 © ©
16 - 1.21
MEAN - .99
16 - .77
*- 0 Ji ' i M It'' * * piiiii|
: i. ^ s ip is ;
ill.;.
5 10 15 20 2,5 30 35 40 45 50 55 60 *5 70 7i 80 85 90 95 100 JUBE NUMBER
'- -'- ' J- - -J- .-'-' ' ' IN ; OUT RELATIVE LOCATION
':'.- :] : : (INLET/OUTLET)
IK OUT ; US .. OUT
:our :
.:: i
i
IN
OUT
: IN . ;" our
i: .-
TEST POINT LOCATIONS
Figure 3-38 Flow Balance Test Results, Third Stage Separator
221-25
-------
Table 3-6. DUST TEST DATA 3V-INCH CYCLONE COLD FLOW TESTS
Test
time
(min)
0.0
10
20
22
22
30
32
32
40
50
60
65
65
70
80
90
100
100
110
120
OH f ice
upstream
(In. Hg)
2.2
3.0
4.4
4.5
2.2
3.5
3.7
2.2
3.0
3.8
4.2
4.5
2.2
2.9
3.4
4.0
4.4
2.2
3.5
4.1
Dust &
feeder
(lb)
202.0
196.5
182.5
280.0
280.0
173.0
170.0
170.0
163.0
154.8
146.0
142.5
142.5
137.8
130.7
123.5
116.6
116.6
110.4
104.0
Dust
(lb)
0
5.5
19.5
22.0
22.0
0.0
7.0
10.0
32.0
0.0
7.0
15.2
23.4
27.5
59.5
0.0
4.7
11.8
19.0
25.9
85.4
0.0
6.2
12.6
98.0
Fed
(gm)
0
2,497
8,853
9.988
9,988
0
3,178
4.540
14,528
0
3,178
6,901
10,624
12.485
27.013
0
2,134
5.357
8,626
11.759
38,772
0
2,815
5.720
44,492
Dust
collected
on absolute
(gm)
0
1.852
1,852
848
2,700
2.251
4,951
2.267
7,218
1.119
8,337
Efficiency
( percent )
E
81.5
81.5
81.3
81.4
81.9
81.7
81.4
81.6
80.4
31.4
3-67
-------
and dust in the collection chamber or cone (34,379 gm). A discrepancy
of 1772 gms of dust was unaccounted for after weighing the absolute
filter and comparing it to what was collected in the hopper. Approxi-
mately 300 gm had collected at the separator inlet and the tube baffles.
The remaining dust was deposited in the duct work or drawn through the
absolute filters. The unaccounted for dust was assumed to have pene-
trated the assembly.
After the dust test, the dust samples were analyzed. The part cle size
analyses results are shown in Figure 3-39. The particle cutoff size
(50 percent penetration) for this test was 2.1 microns. The o/erall
efficiency, by percent weight, was 77.9 percent. There were no signifi-
cant dust deposits on the individual tubes during these tests. A plot
of cyclone efficiency versus stated particle size is shown in F1 ure
3-40.
Hot Operation During the 20-Series Tests.
For the 20-series testing the third stage separator was installed as
the final hot gas cleanup stage (see Figure 3-29). The ash unloading
system used for the 3'j-inch tube inertial separator is similar to that
of the second stage separator assembly.
Table 3-7 summarizes the testing and results of the 20-series test
whose prime objectives were to pass LP-11 test requirements. Although
the Initial tests demonstrated the capability of the third stage sepa-
rator to reduce the ash loading in the exhaust to the levels predicted
3-68
-------
99.5
99.0
98.0
15.0
1 I I I I I | I
A '
0 FEED
O CONE
A ABSOLUTE
10.0
PARTICLE DIAMETER -microns
100.0
221-26
Figure 3-39 Dust Test Particle Size Analysis, Third Stage Separator
3-69
-------
100
90
80
70
>-
60
30
20
10
10
PARTICLE
I-icrons)
Figure 3-40 Cyclone Efficiency vs. Stated Particle Size
-------
Table 3-7. 20-SERIES THIRD STAGE SEPARATOR TEST SUMMARY
Test
No.
20-2A
& 2B
20-2C
20-3
Nominal
flowrate
(Ib/min)
350
350
350
Nominal
inlet
temp.
(F5)
1460
1496
1400
No. of
active
tubes
100
100
100
Operating
time
(hrs)
3-3/4
3'i
12.2
Outlet loading (gr/scf)
>2u >5u Total
0.0114
0.0116
0.0173
.
_
0.0001
0.0001
0.0002
.
_
0.157
0.0397
0.0272
0.0391
0.0475
0.056
0.028
0.046
0.046
0.026
Comments
Vibrator airlines failed &
were reworked as those in
the 2nd stage separator.
Assembly was not disassem-
bled after the test ^lex-
line was installed in the
vibrator system as it
appeared to solve the duel
fai lure problem.
This test series consisted
of 8 short duration runs
Post-test inspection re-
vealed 6 plugged tubes
and minor ash deposits in
the remaining 94 tubes, t
a result of this inspectic
the 20 outer ring tubes
we*-e deactivated for the i
mainder of the test serie;
As
-------
Table 3-7 (continued). 20-SERIES THIRD STAGE SEPARATOR TEST SUMMARY
Test
No.
20-4A
20-4B
20-5
Nominal
flowrate
(Ib/mln)
350
350
350
Nominal
Inlet
temp.
(F*)
1440
1433
1430
No. of
active
tubes
80
(tubes
81-100
deacti-
vated)
80
80
(71
with
rat-
tlers)
Operating
time
(hrs)
33 S 1
0 15
9 30
50-2/3 P 3
9 24
0 48
8.2 9 2
0 8
Outlet loading (gr/scf)
>2u >5u Total
0.042
0.21
0.47
0.014
0.19
0.15
.
-
0.0066
0.12
0.33
0.0039
0..0
0.05
_
-
0.044
0.58
-
0.081
0.50
0.40
0.1
0.31
Comments
As Indicated by the loading
data, the ash outlet
cones of the 80 tubes be-
came plugged. The common
ash hopper outlet was
plugged and ash was re-
tained In the hopper. The
assembly was cleaned and
the bleed flow Increased
from 1.5 Ib/m1n to 3 lb/
m1n for the next test.
Post-test Inspection re-
vealed successful rwoys]
of ash from the hopper
during the entire test
period. The cones of all
3»j-inch cyclone tubes
were plugged by ash de-
posits.
Post-test inspection showed
all tubes with rattler
free of ash deposits. 7
tubes without rattlers
were plugged, and 2 tubes
without rattlers were
open.
I
^1
fS)
-------
by the model testing, individual plugging of the cyclone tube cones oc-
curred during all long operating periods. The effect of tube plugging
on particle size distribution is depicted in Figure 3-41. Test series
20-5 was conducted to evaluate the effectiveness of the wiper type of
rattles for eliminating the ash deposits in the cones. These wipers
consisted of rods that are pivoted at the inlet plane to the vane tube.
A picture of the tubes after the test, Figure 3-42, clearly shows the
beneficial effect of these wipers.
At the conclusion of the 20-Series tests, the 3H-inch tube inertial
separator assembly was disassembled and inspected. The inspection
showed the tube seals between the clean and dirty gas aisles functioned
well with no evidence of leakage. There was a slight buildup of alumina
on the outer diameter surface of the tube exhaust pipes which is in the
dirty gas aisle. The vane area of the inertial tubes was heavily loaded
with ash agglomerations as shown in Figure 3-43. A dimensional check
of the 3Vinch diameter tube vanes revealed that, at the trailing edge
of a majority of the vanes, the angle which the vane makes with the
axial direction deviated from print by 9 degrees, resulting in a flow
separation condition below the trailing edge. The edges of some of the
vanes were bent to the point where the trailing edge was almost perpendi-
cular to the axial direction of flow. The vessel assembly showed no evi-
dence of deterioration of the inner liner by either corrosion or erosion.
Wall temperatures were in the range of 150° F to 250° F during testing
and showed that the Kaowool insulation performed satisfactorily. All
vanes were examined and adjusted to an angle of 35 degrees and retested.
3-73
-------
o
1/1
o
IxlO6
8
6
4
2
IxlO5
6
4
2
Uljf
6
4
2
IxlQ3
8
6
4
2
IxlO2
8
6
4
2
IxlO1
Loading not corrected for 02
. _
.
I
|
*
|
INITIAL
LOADING
|
^>«J
iAMPLE
»* -
V.
\
or C0?
FIN
LOA
%
>
concentration
H SAM
DING
v
S
\
>Ll
\
1
_JI
pS
*r
i |
,
KL
\
\
\
\
' "
^
\
\
\
1
+
L
|
3 4 S 6 7 8 9 10
PARTICLE SIZE, x - (micron)
20
30 40 50
221-216
Figure 3-41 Particle Loading Distribution, Test Series 20-4A
3-74
-------
OJ
I
-4
cn
Figure 3-42 Rod Style Rattlers In 3VInch Tube Separator
-------
u>
I
on
Figure 3-43 Flyash Buildup on Vanes of Third Stage Seoarator Tubes
-------
The test results showed significant reduction in the anount of ash de-
posited at the trailing edge.
Design and Development of the Subscale Granular Filter
The subscale granular filter was fabricated and tested as a backup
cleaning <:. *!ce for the inertia! separators. The basic system design
was recommended by Professor Arthur Squires, of the City College of the
City University of New York, whose expertise was consulted for concept
and design evaluations of granular filter devices. The design used for
preliminary evaluation of this device with high temperature solid waste
flyash is shown in Figure 3-44.
The active filter area is 12 inches wide and 75 inches high. At a de-
sign flow of 75 cfm the face velocity across the panel is 12 feet per
minute. The selection of this velocity was based on past erimental
studies of granular filters.
The filter assembly consists of a manifold module which contains the
inlet, outlet, and granular material feed section, an active filter
panel consisting of 120 inlet louvers and 200 outlet louvers, material
collection hoppers at the bottom of the inlet, outlet, and granular
material cavities, and auxiliary electrical heaters along the outside
surfaces of the gas passages to make up for heat losses and maintain
/
the assembly at the gas inlet temperature.
The louvers of the filter were held together with tie rods. Their rela-
tive position was maintained with sleeve support spacers. The ends of
3-77
-------
.06 TYP
u>
i
^i
oo
INLET
MANIFOLD
CLEAN
GRANULAR
MATERIAL
HOPPER
PARTICLE
SAMPLING
RAKES
GRANULAR MATERIAL
"FEED
_ OUTLET
MANIFOLD
DIRTY
LOUVERS
HEATER
WIRE
CERAMIC FIBER
BLANKET INSULATION
ELECTRICAL
INSULATING
STAND-OFFS
CLEAN SIDE
LOUVERS
SEPERATED
MATERIAL
OUTFEED
CLEAN AISLE
OUTFEEl
MATERIAL
REGENERATION
OUTFEED
GRA.JULAR FILTER ASSEMBLY
CLEAN LOUVER
SUPPORTS
SLEEVE ID -= .270
00 = .405
ROD OD = .25
INLET LOUVER
SUPPORTS
SLEEVE ID = .18
00 * .31
ROD OD = .16
TYPICAL PANEL CONFIGURATION
221-217
Figure 3-44 Granular Filter Design
-------
the louvers were set into channels. The granular material itself
formed an effective seal between the ends of the louvers and the channel
walls. The element support and end seal allowed relative growth be-
tween the thin louvers and the vessel side walls due to differential
expansion during high temperature operation.
The design of the active element is ideally suited for enlargement of
the panel size which is planned for future testing with a puffback fil-
ter cleaning system and a material regeneration system.
The inlet, dirty side, louvers were positioned at a 56 degree angle in
order to make the sand surface exposed to the incoming gas normal to
the louvers. This optimizes the puffback cleaning operation because
the filter cake is formed on a minimum surface plane.
The principle of operation of the device is very simple. Hot, dirty gas
enters the filter through the inlet manifold and flows between the lou-
vers into the granular material (16 mesh sand was used). The path
followed by the gas as it flows through the sand results in the ash
being deposited on the sand surface. The cleaned gas then exits through
the outlet manifold. Puffback may be used to momentarily reverse the
flow, causing a layer of ash cake and particle-laden sand to be blown
out to be collected at the bottom of the assembly. Clean sand enters
through the top to replace that which leaves during puffback.
Granular Filter Testing-
The granular filter was tested using both cold air and hot solid waste
exhaust gas. The cold gas tests were used to make general checks of
3-79
-------
filter differential pressure, flyash loading characteristics, and sand
size and retention ability. Testing was carried out at a filter inlet
temperature of 1400° F. The effect of high temperatures on filter bed
material, the effect of pressure fluctuations on the filter bed, and
the total loading filtering efficiency were test objectives.
Cold Gas Testing - The cold gas testing was performed using two granular
filter configurations. The first configuration (Figure 3-45) used 30
mesh sand. The angle of repose of the 30 mesh sand was such that vibra-
tion during filling caused a loss of sand from the outlet louvers. A
flow test was conducted resulting in a continual loss of sand from the
outlet side of the filter for all flow conditions tested. Test results
are cited 1n Table 3-8.
As a result of the flow test, two changes were instituted: the replace-
ment of the 30 mesh sand with 16 mesh sand, and the addition of 1/8-inch
high x %-inch deep bars welded along the outlet louvers to form a ledge
against which the sand could support itself. Cold testing was then re-
run. Results are cited in Table 3-9. The changes made the filter less
sensitive to vibration, and later cold flow testing at 75 acfm demon-
strated a filtering efficiency of 99.29 percent which proved acceptable
to the point where hot gas testing could be Initiated.
Hot Gas Testing - The hot gas test setup was as shown in Figure 3-46.
Three tests produced the results cited in Tables 3-10 and 3-11. During
the time between test 15-4 and 15-ll(a), the filter was cleaned of fly.
ash and the sand was removed and replaced with new sand.
3-8C
-------
(A) INITIAL FLAT DESIGN
(B) FINAL DESIGN WITH
LIP ON EXIT PLANE
221-196
Figure -1-45 Granular Filter Outlet Louvers
3-81
-------
Table 3-8. 30 MESH SAND COLD FLOW
Flow
(acfm)
75
120
Filter AR
(inches H-O)
0.23
0.30
0.37
0.45
0.46
C.62
0.80
0.10
Comment
Continuous loss of sand past the lou-
vers resulted in blowthrough at the
top of the panel as hopper capacity
was spent at the end of testing.
Table 3-9. 16 MESH SAND COLD FLOW WITH LOUVERS
Flow
(acfm)
47
69
91
108
125
142
160
173
173
194
Filter AP
(inches H20)
0.14
0.23
0.25
0.47
0.60
0.74
0.90
1.01
1.06
1.46
Comment
Water from the compressed air system
began to appear at the filter out-
let.
3-82
-------
OJ
I
oo
Figure 3-46 Hot R*s Test Setuo
-------
Table 3-10. HOT FLOW TEST DATA
Elapsed time
(min)
Test 15-4 Flow = 7*> acfm
190 Diesel oil
205 Solid waste
220 Solid waste
260 Solid waste
275 Solid waste
290 Solid waste
310 Solid waste
325 Solid waste
Test 15-ll(a) Flow = 75 acfm
258 Diesel oil
282 Solid waste
287 Solid waste
307 Solid waste
317 Solid waste
329 Solid waste
352 Solid waste
Test 15-ll(b) Flow = 75 acfm
104 Diesel oil
123 Diesel oil
148 Solid waste
155 Solid waste
180 Solid waste
Filter inlet temp. ( H
Inlet
1160
1225
1340
1355
1360
1365
1380
1460
1105
1200
1265
1405
1400
1400
1405
985
1040
1140
1235
1290
Top
765
850
890
890
900
920
980
1040
740
885
860
965
985
825
770
750
800
900
835
730
Filter outlet temp. (°F)
Top
505
560
650
760
740
750
770
780
460
595
625
7550
775
850
810
475
565
650
660
730
Outlet
900
900
900
900
900
900
900
900
555
680
70S
860
905
1095
1200
550
635
940
775
1050
Filter AP
(inches H20)
1.2
1.6
3.5
3.5
4.5
5.0+
5.0+
5.0+
.
1.6
1.7
1.7
1.6
1.0
0.6
1.1
1.2
1.6
0.7
0.0
OJ
I
oo
-------
Table 3-11. TOTAL LOADING FILTER EFFICIENCY DATA
Sample
No.
15-4
1
2
1
2
3
1
2
Total loading
(gr/scf)
In
0.0336
0.8788
0.077
1.0571
0.8353
0.683
1.8883
Out
0.002522
0-.05144
0.0027
0.2570
0.7552
0.0007
0.7684
Efficiency
92.5
94.15
96.5
75.7
9.59
98.98
59.3
Elapsed
test time
(min)
178 - 192
305 - 325
256 - 268
294 - 307
357 - 369
128 - 140
164 - 178
Comments
Diesel oil operation
Solid waste operation
Diesel oil operation
Solid waste operation
Solid waste operation after
breakthrough
Diesel oil operation
Solid waste operation after
breakthrough
CD
(-1
-------
After tests 15-4 and 15-ll(b) inspection revealed a heavy flyash cake
buildup except where breakthrough occurred in the top 15 percent of the
filter panel. A gradual taper in cake thickness was evident which con-
tinued uniformly to the bottom of the filter where in random spots,
there was little evidence of ash deposits as depicted in Figure 3-47(a).
Final inspection of the filter panels after completion of the testing
disclosed a warpage of the top eight louvers on the outlet panel. The
louver spacing increased from an original 0.172 inches to as much as
0.5 Inches for the top louver. Warpage also occurred on the top two
louvers on the inlet panel. The spacing increased from an original
0.57 inches to as much as 0.72 inches. Further inspection of the inlet
panel pointed up a structural failure in the two outer edge louver tie-
down bars just below the second louver from the top. The warpage of
the louvers can be eliminated by providing retaining stops on the tie
bars above the last louver to prevent relative motion between the louvers.
From the data taken from the tests 15-11(a) and 15-ll(b), it was evident
that flyash buildup occurred to the point where the cake could no longer
support the pressure differential across the panel and a breakthrough
occurred at the top of the panel where the louvers were spaced at a
greater distance than those below. Figure 3-47(b) shows evidence of
the breakthrough where an absence of flyash cake is noticeable.
The hot gas test runs were initiated using diesel oil, and the panels
heated to an outlet temperature of 500° F to 600° F. The panel heaters
were connected to a power source which supplied 0.7 to 1.8 kW during
3-86
-------
(A) WITHOUT BREAKTHROUGH
(B) WITH BREAKTHROUGH
221-198
Figure 3-47 Flvash Builduo at Inlet Louvers
1-87
-------
the dlesel oil heatup and up to 4.8 kW during solid waste operation.
Filter Inlet temperatures reached 1400° F with no sand agglomeration
noticeable on either the inlet or outlet face of the sand. However,
despite the power input of 4.8 kW, the outlet temperatures never met the
specified 100 F less than inlet temperature requirement cited in the
Test Plan.
The 15-4 test Indicated that a panel differential pressure of 5 inches
of water is attainable on an as-designed panel without breakthrough or
loss of sand due to combustor induced pressure oscillations. This test
demonstrated that the granular filter 1s capable of building up a filter
cake which is Insensitive to pressure fluctuations produced by a fluidiz-
ing bed.
The total loading filter efficiencies are listed in Table 3-11. The
loadings for tests 15-ll(a) and (b) reflect the sharp difference in out-
let flyash loading after breakthrough occurs. The majority of the solid
waste operation was typified by sample 2 of test 15-4. Figures 3-48
and 3-49 are submitted as an example of particle removal efficiency and
particle loading distribution as a function of particle size.
DESIGN AND DEVELOPMENT OF THE RESIDUE REMOVAL SYSTEM
Determination of Need for Residue Removal System (Task ^?-5)
A total of 247 hours of testing was conducted to determine the require-
ment for residue removal as the result of solid waste combustion within
a fluldlzed bed. The quantity, size, growth or attrition, the nature
3-88
-------
100.0
99.8
>»
u
I
u
«l
t;
*
i
o
ui
i
2
o
lf>
6 7 8 9 10 11 12 13 14 15
X. PARTICLE SIZE -micron
20
221-219
Figure 3-48 Granular Filter Fractional Efficiency
3-89
-------
2 3 4 5 6 \ 8 9 10 20
X. PARTICLE SIZE - micron
Figure 3-49 Size Distribution of Material
Penetrating and Entering the Panel Filter
30
40 SO
221-218
3-90
-------
of the residue, handling considerations, and an evaluation of the ex-
haust system separators constituted the scope of these tests. The
entire testing sequence was performed with the CPC designed Model No. 3
combustor. This vessel (see Figure 3-50) has an internal diameter of
20 inches and an outside diameter of 32 inches, with 134 feet between
the air distributor and the exhaust pipe. The inner liner was composed
of 16 gauge, 310 stainless steel; the outer shell is made of 11 gauge
carbon steel. The 6-inch innular space between the internal and out-
side diameter was packed with Kaowool insulation to reduce heat loss.
Solid waste was introduced into the fluidized bed through a 2-inch
feed pipe which entered the bed just above the air distributor. The
i
solid waste was fed to the feedpipe pneumatically from a 5-inch
feeder valve. The average feedrate was controlled by the desired bed
temperature setting in the cent.-"1! system which fixed the on/off
periods of feed. Provisions also existed for introducing fuel oil
into the bed. A schematic of the Model No. 3 combustor test setup and
a photograph of the solid waste feed system are shown in Figure 3-51
and 3-52, respectively. Details of data and measurements required dur-
ing the tests are summarized in the following paragraphs.
Fluid Bed Combustor Bed Test Conditions-
For the 240 hours of testing, during which solid waste was used as fuel,
the initial fluidized bed was composed of 16 mesh beach sand. The
size distribution of the starter sand is shown in Figure 3-53. For
the first 120 hours of testing the bed level was allowed to float.
During the next 60 hour test period, the bed was maintained at 2 feet.
3-91
-------
Figure 3-50 Model No. 3 Installation
3-92
-------
£'7k
' '6^:
PW **
-;^..
_.._.
t-ao
A,s
^S, *^
?E*
A-*O
4
-^
\ '
Figure 3-51 Model No. 3 System Schematic
-------
Figure 3-52 Model No. 3 Solid Waste Feed
3-94
-------
I
IA-221-61
Figure 3-53 16 Mesh Beach Sand
3-95
-------
For the final 60 hour test period, the bed level was held at 3 feet.
During a subsequent 7-hour test, during which dlesel oil was used as
fuel, the bed level was again maintained at 2 feet. During all tests,
the air flow through the bed was maintained at 26:5 scfm to provide a
nominal superficial velocity of 7 ft/sec at bed temperature. During
the entire test the bed temperature was maintained at 1425 ± 10° F,
except for some momentary delays caused by plugging 1n the feedllne.
This caused the bed temperature to drop below 1400° F, and the test
clock to be stopped. However, these problem were easily solved and
the testing resumed without further Incident. A typical temperature
record of the bed and freeboard 1s shown 1n figure 3-54. Of Interest
1s that the freeboard temperature 1s normally higher than the bed tem-
perature and generally decreases with Increased bed depth.
Bed Characterisecs-
SoHd Waste Consumed - A total of 32,074 pounds of solid waste was
consumed. Twenty-three samples of the material consumed were taken
for limited analysis of moisture content, total inerts, size distribu-
tion of Inerts, and metal content. The results of this analysis are
cited 1n Table 3-12.
Bed Size Distribution - A simplified size distribution of 16 mesh
starter beach sand and the size distribution after 120, 180, and 240
hours of bed operation 1s given in Table 3-13. As Indicated by Table
3-13, the steady state bed was reached prior to 120 hours of operation.
The aged bed material 1s shown in Figure 3-55.
3-96
-------
>£>
MODEL TEST 13-21 SYSTEM TEMPERATURES
AVERAGE BED TEMPERATURES
A ... (T/T1+T/L 2+T/L3+T/l4)/4
B ... (T/L5+T/L7)/2
C ... T/L10 "."
AVERAGE DISTRIBUTOR PLATE TEMPERATURE
FREEBOARD TEMPERATURE
'SOLID WASTE FEED TIME (HRS);
221-220
Figure 3-54 Temperature History (LP-5 Testing)
-------
Table 3-12. SOLID WASTE SAMPLE DATA
Test
time
(hrs)
0 - 120
120 - 180
Solid
waste
consumed
(Ibs)
15,595
Moisture
content
U)
20.9
28.9
Inert
total
(%)
11.27
6.06
Inert
metal
U)
0.25
0.09
Inert
distri-
bution
U)
22.11
33.72
44.17
25.07
lfi,47Q -n ?fi
41.67
180 - 240
23.4
10.02
0.08
Size
(U.S.
Standard
mesh)
>16
16>30
<30
>16
16>30
<30
3-98
-------
Table 3-13. BED MATERIAL SIZE DISTRIBUTION
Size
(U.S.
Standard
mesh)
>16
16>30
<30
Beach sand
( 16 mesh starter)
(X)
25.37
72.76
1.87
120 hour
W
34.61
53.88
11.51
180 hour
(%)
34.21
53.06
'?.73
240 hour
(*)
35.89
52.58
11.53
Table 3-14. INCIPIENT FLUIDIZAT10N VELOCITY
Test time
(hrs)
0
120
180
240
Incipient
fluidization
velocity
(ft/sec)
1.68
1.54
1.43
1.43
3-99
-------
TA-221-62
Figure 3-55 Aqed Bed Material
3-100
-------
At the conclusion of the 240 hour test the bed material was passed
through a 'j-inch screen. Approximately 5/8 pounds of the 610 pounds
of bed material sieved was greater than ^-inch 1n size. A comparable
percentage of large particles was found in the samples removed from the
operating bed Indicating that these items are evenly distributed
throughout the bed. The large particles consisted of rocks, ceramic
pieces, clinker pieces, bolts and miscellaneous items. This small
quantity (5/8 pounds) compared to the 32,000 pounds of solid waste
consumed is not expected to cause a problem.
A 7-hour attrition test using diesel oil as fuel was performed in an
effort to establish the effect of bed operation on bed media size
changes. The test was conducted at a bed temperature of 1400 F and
a superficial velocity of 7 ft/sec. Size distribution and weights
were analyzed prior to and following the test. Three pounds of the
183 pounds of bed material greater than the 16 mesh size were worn
away during this test.
Metals - Based on 22 solid waste samples, 86 pounds of metal were fed
into the combustor during the 240 hour test. Fifty-five pounds of
metal were removed as samples and for height maintenance during bed
operation. Nineteen pounds were present at the end of testing.
Therefore, only 12 pounds of metal were consumed during the test. A
metal attrition rate of 0.003 pounds oxidized per pound present per
hour was calculated from this Information. The 7-hour bed material
attrition test described above was also used to verify this value.
3-101
-------
In that test, the metal attrition rate was 0.005 pounds oxidized per
pound present per hour. This value is comparable to the calculation
made during the 240 hour test.
Minimum Fluidizing Velocity - An important parameter in bed material
evaluation is minimum fluidizing velocity. Table 3-14 displays the
manner in which this parameter varied with time. The data, taken at
ambient conditions, indicates that the population of fine particles
(Table 3-13) is the dominant factor in bed activity, which improved
with time during the test.
Bed Depth - During the first 120 hours of testing, the depth was
allowed to establish its own level. The depth was, therefore, deter-
mined by a balance between inerts introduced with the solid waste and
elutriation. Figure 3-56 shows the bed variation (AP) and solid waste
inert percentage as a function of test time. As indicated in this
figure, the lowest level of solid waste inerts recorded during the
test was 6 percent. The 6 percent solid waste inert level was the
lowest experienced during the contract oeriod and probably occurred
because of excessive handling of the solid waste during the early part
of the test. However, a low solid waste inert level is a possibility
and consideration should be given to shallow-bed operation and bed
material addition.
Between 120 and 180 hours the bed was maintained at the 2 foot level
(static depth) as determined by bed pressure drop. To accomplish this,
494 pounds of bed material was drained from the operating bed. For
3-102
-------
o
OJ
20
1 I I
INERT AVERAGE (11.27 )
TEST 13-20 120 HOUR RESIDUE TEST
2 FEBRUARY, 1972
ELAPSLJ TIME - 'iO
Figure 3-56 Bed AP (Volume) and Inerts Added
-------
the final 60 hours, the bed depth ,vas held at 3 feet. Durinq this
period, 496 pounds of bed material was removed in order to maintain
the desired level.
Bed Density - Density of the starter bed was 100 pounds per cubic foot.
At the end of the first 120 hours the density had stabilized at 89
pounds per cubic foot.
Elutriation - In the fluidlzed bed combustion process, the lighter
particles are constantly being carried away in the exhaust gases.
The flyash particles (<150u) are readily elutriated. Other inert
particles introduced into the fluidized bed in the solid waste (of
which the steady state bed is composed), are elutriated in relation
to size, freeboard height, gas velocity, and bed size.
In the model test, elutration for the 2 foot and 3 foot bed depths
was 840 and 1189 pounds of material, respectively, over the two 60-
hour periods. Also, the particle size of the elutriated material
captured by the first stage separator was larger for the 3 foot bed:
13.3 percent >30 mesh compared to 4.8 percent >30 mesh for the 2 foot
bed.
Chemical Analysis - Samples of the original bed material, aged bed
material, flyash, and exhaust pipe clinker were subjected to spectro-
graphic analyses. The results of this analysis are summarized in
Table 3-15.
3-104
-------
Table 3-15. SPECTROGRAPHIC ANALYSIS (SEMI-QUANTITATIVE)3
Element
Iron
Copper
Nickel
Chromi urn
Aluminum
Lead
Tin
Zinc
Cobalt
Manganese
Molybdenum
Si licon
Silver
Vanadium
Magnesium
Calcium
Titanium
Sodium
Potassium
Strontium
Zirconium
Boron
Barium
16 Mesh
beach sand
1.00
0.0003
0.001b
0.02
4.50
0.001
o.ooib
o.io b
0.001b
0.01
0.001b
Major0
O.C001b
0.002
0.15
0.35
0.05
2.00
3.00
0.03
0.007
0.001
0.15
120 hr
bed
2.00
0.12
0.002
0.08
3.50
0.08
0.004
0.10
O.OOl6
0.08
b
.0.001
Majorc
0.001
0.004
0.30
10.00
0.15
6.00
2.00
0.02
0.02
0.05
0.18
240 hr
bed
1.60
0.20
0.001
0.25
3.00
0.20
0.02
0.10
o.ooib
0.03
0.001b
Majorc
0.002
0.005
0.70
10.00
0.20
5.00
2.00
0.02
0.01
0.03
0.20
Flyash
after
240 hrs
3.00
0.10
0.01
0.07
10.00
0.10
0.03
0.15
0.001b
0.25
b
0.002
Major0
0.002
0.006
1.20
MajorC
0.30
5.00
2.00
0.02
0.01
0.02
0.10
Exhaust
pipe
clinker
3.00
0.15
0.04
0.10
Major
0.10
0.06
0.20
b
0.001
0.25
b
0.001
10.00
0.002
0.01
0.80
6.00
0.25
2.00
1.00
0.01
0.004
0.01
0.10
All values expressed in percent
3
Less than indicated value
More than 10 ».- "cent
3-105
-------
Generally, bed material had stabilized by 120 hours and a slight in-
crease in trace elements was visible with time. A pronounced increase
(from 0.35 to 10 percent) in calcium was observed. The exhaust pipe
clinker was chemically very similar to the flyash except the aluminum
content was much higher and the silicon content was lower.
Exhaust System-
The secondary objective of the 240-hour test was the evaluation of the
exhaust system including material separators. It was initially found
that a very hard gray clinker formation occurred at the first impinge-
ment point (sharp turn) in the exhaust pipe. The composition of this
formation, shown in Table 3-15, was principally aluminum with lesser
amounts of silicon present as the secondary element. The formation rate
was approximately 0.25 pounds per 100 pounds of solid waste consumed.
The first device used in breaking up this formation was a hollow im-
pingement plate which was periodically shocked by water circulation.
This configuration, shown schematically in Figure 3-57, successfully
flaked the clinker off, but soon failed structurally as a result of
thermal shock.
After 49*5 hours of test operation, the clinker handling concept was
changed. The impingement plate was inclined at an angle of 45 degrees
to the incoming gas, and a settling chamber and a sand/ash removal
line were added to the test configuration. This configuration operated
satisfactorily for the duration of the test. The inclined plate re-'
moved 93 percent of all hard clinker generated while the remaining
3-106
-------
o
J
COOLING WATER
TO K!PI%r.ME'JT
/ PLALE
0 TO 49 5 HOURS
FROH
COMUUSTOR
PLATE
\
49.5 TO 24O HOURS
I'lERTIAL
SEPARATOR
Figure 3-57 Exhaust Gas Impingement Plate Configuration
-------
7 percent was deposited downstream in the exhaust and inertial
separator.
Mass Accounflng-
A summary of the materials handled during the 240 hour test is shown
in Table 3-16. Except for those noted as having been calculated, all
weights cited in the table were taken from a beam balance scale. The
solid waste ash content was calculated from incinerator refuse and
residue data presented during the 1968 National Incinerator Conference
and was adjusted for moisture and inerts content. The flyash entrained
in the exhaust gas was calculated from particle loading data. Table
3-16 shows that the total amount of calculated inerts added (including
ash) was 5111 pounds. The measured inerts and ash weighed 5846 pounds.
This is a discrepancy of 12.6 percent and is probably the result of a
nonrepresentative solid waste inert content value. This value is sus-
pect because of the excessive handling required in model combustor
operation. This causes the inert fraction to settle in the storage
vessels, and results in a nonrepresentative grab sample. The measured
inerts plus ash percent of 5846/32074 x 100 = 18.25 percent (of which
one-third is ash), is in accord with previous findings. Table 3-16
also shows that 1977 pounds of ash, (defined as all particulates
smaller than 100 mesh) passed through the sand separator, and that
3787 pounds of sand were removed from the bed and sand separator.
3-108
-------
Table 3-16. LP-5 SOLIDS MASS ACCOUNTING
Parameter
Solid waste
consumed (Ib)
Ash content (Ib)
Feed rate (Ib/min)
Average moisture (%)
Average inerts (%)
Total inerts (Ib
calculated)
Total (ash i inerts)
Bed sand removed (Ib)
Sand exhausted from
combustor (Ib)
Flyash exhausted
from combustor (Ib)
Hard clinker
formation (Ib)
Total
0 - 120 hrs
15.199
990 (8 6.35ro3
2.16
20.9
11.27
1,758
339
1,208
1,263
(380)b
42
2,852
120 - 240 hrs
16,479
1054 9 6.40%3
2.18
26.2
8.04
1,309
925
1,315
714 h
(205)b
40
3,199
Total
2,044
3,067
5,111
1,264
2,523
1,977
82
5,846
a Ash content is defined as oarticulates smaller than 100 mesh.
The number in parentheses represents the ash penetrating the
inertial separator. The values are based on particle loading
measurements.
3-109
-------
Development of the Residue Removal System
During the LP-8 test series, 1t was found that the Inerts inflow to
the verf'cal combustor equaled the bed material elutrlatlon and, there-
fore, no bed material removal was required. The data obtained from
the model testing Indicated that a significant amount of metal should
accumulate during pilot plant operation; but actual operation of the
vertical combustor 1n excess of 100 hours yielded metal levels of 0.5
percent - significantly below the anticipated steady state metal level
of 10 percent. This la^ge difference 1n metal content was thought to
be associated with Improvements 1n solid waste shredding/air classi-
fying equipment used during the LP-8 testing. The steady state level
was determined by using a metal attrition rate of 0.004 pounds per
hour per pound of metal present, a 0.16 percent content of high melt-
Ing point metal in the solid waste feed, and a flow rate of 30 pounds
per minute.
However, 1t was determined that a material removal system would be
needed for the first stage separator. The primary function of this
device was to remove sticky particles and the larger Inerts from the
gas stream. The sticky particles, such as molten aluminum, are de-
posited and are converted to aluminum oxide 1n the spinner duct of
this vessel. The Inert materials elutriating from the combustor pro-
vide the scrubbing action necessary to eliminate any significant
buildup of iiaterlal. The material accumulating In the bottom of this
separator (sand, flaked deposits, ash, etc.) must be continually
3-110
-------
removed or 1t will fuse Into a large clinker. This occurred when
material was al.owed to accumulate 1n the bottom of the pilot plant
first stage separator during a 24-hour test. The material removal
system developed to alleviate this problem 1s shown 1n Figure 3-58
and was first used during the LP-11 testing.
During the LP-11 testing, the material removal system was operated for
approximately 100 hours and successfully demonstrated the continuous
removal of the residue material and Its transportation to an external
collection point. The sand/ash mixture was pneumatically transported
to a sand bin through a 2-1nch diameter pipe. The fines were pneu-
matically transported and collected in the existing medium temperature
baghouse system. The flow Umlter plate at the end of the pipe in the
sand bin was used to control and to adjust the gas flow rate in the
transport pipe.
A similar setup will be used for direct fluid bed combustor material
removal "uring future high pressure operations 1f the bed depth does
not stabilize at a suitable level. However, minor modifications will
be made to the system to permit residue unloading during high pressure
operation. These modifications consist of the incorporation of a
double-ball valve air-lock volume with continuous bleed in the pipe
section below the vessel outlet, and the removal of the transport line
ball valve. The double-ball valve arrangement with Intermediate
volume bleed will allow cyclic unloading from the pressurized system
without exposing the ball valves to high velocity conditions.
3-111
-------
00
I
SPINNER DUCT - -
FIRST STAGE
SEPARATOR
AIR
FLOW LIMITER
PLATE
LARGE PORE
SCREEN
2"dia BALL VALVE
WITH POSITIONER
TEMPERATURE
CT.TROL
PUT-BACK
BLOWER
SEPARATOR ASH
INLETS
1st STAGE
-^- -2nd STAGE
AIR LOCK ROTARY
UNLOADING VALVE
WATER
221-221
Figure 3-58 Material Removal System Schematic - Low Pressure Operation
-------
The pipes above and below the valves will be water-jacketed to facili-
tate cooling. The ash removal and residue removal system that will
be intw^rated into the pilot plant for high pressure testing is shown
in Figure 3-59.
DETERMINATION OF OFF-GAS COMPOSITION OF A FLUID BED COM8USTOR FUELED
WITH MUNICIPAL SOLID WASTE (TASK PL-3)
Testing was conducted to establish the chemical qualitative and quanti-
tative composition of the off-gases and fumes from a fluid bed combustor
burning municipal solid waste. The purpose of the testing was to es-
tablish the principal corrosive constituents of gases and flyash parti-
culates and their concentrations to permit simulation of the fluid bed
exhaust during the turbine blade corrosion/deposition testing which
was later conducted by The Westinghouse Research Laboratories. The
gas analysis testing consisted of operating a fluid bed combustor with
municipal solid waste fuel utilizing parallel exhaust analysis methods.
One method was a continuous on-line sampling of the exhaust for gas
composition with a mass spectrometer. The other method consisted of
two identical sequences of collecting gas samples and particulate
samples utilizing standard dry and wet chemistry methods developed for
pollution analysis. Repetition of the sample collection sequence was
scheduled for protection against possible sample loss or contamination
during the analysis.
3-113
-------
FROM TURBINE
COMPRESSOR
TO SOLID WASTE
COMBUSTOR PLENUM
THIRD STAGE
SEPARATOR
FIRST STAGE
SEPARATOR
SECOND STAGE
SEPARATOR
RESIDUE
TRANSPORT
AIR SOURCE
RESIDUE SEPARATOR TRANSFER TEES
RESIDUE
COLLECTION AND
STORAGE
HOPPERS
BAG HOUSE
FILTER
221-222
Figure 3-59 Pilot Plant Material Removal System
-------
Test Setup for Determination of Off-Gas Composition
Testing was performed using the CPC-designed Model No. 1 fluid bed
combustor in the test setup shown 1n Figures 3-60 and 3-61. The Model
No. 1 combustor, a 1-foot by 2-foot low pressure fluid bed combustor,
burning shredded and air classified municipal solid waste, was used
to generate the off-gases and flyash particulates. The combustor was
operated at 1400° F with the off-gases at a nominal 1600° F in the ex-
haust duct. The fluid bed was 16 mesh silica sand, the static sand -
height 2 feet, and the superficial velocity 5.2 ft/sec. No techniques
nor additives were added to suppress any constituent of the exhaust gas.
The exhaust gas sample was collected downstream from a modified
Donaldson Company inertial separator. The sample collection and dis-
tribution system 1s shown schematically in Figure 3-62. The sampling
system was constructed from 300 series stainless steel components and
was designed for continuous gas sampling with the mass spectrograph,
gas sample collection for later laboratory analysis, and for primary
particulate collection with an in-line filter. A secondary particle
collection point was incorporated Into the overboard vent of the resi-
due collection tank below the inertial separator assembly.
Gas Analyses
The wet sampling and chemical analyses were conducted by the Trapelo-
West Laboratories, Inc. The continuous on-Hne gas sampling was per-
formed by Uthe Technology, Inc., using a Uthe quadropole mass spec-
trometer type of gas analyzer. The laboratory results of the gas
3-115
-------
u>
I
Q
<>3
2«
QS
SOLID FUEL TRANSPORT AIR
BURNER GAS FLOW
FLUIOIZING AIR
OIL GUN COOLING AIR
OIL GUN MIX AIR
BURNER AIR
OIL GUN FLOW
T/C 113
INERTIAL
SEPARATOR
COOLING
WATER
EXHAUST
SECONDARY PARTICLE
COLLECTION BAG
RESIDUE
COLLECTION
TANK
SAMPLING
PORT
OVERBOARD
INERTIAL BLEED
SHOP
AIR 0,L
TANK
221-223
Figure 3-60 Test Anparatus Schematic
-------
Flqure 3-61 Test Apparatus
3-117
-------
L~J
I
00
m
JC
l/2"dia SS LINES
INSULATED AND
ELECTRICALLY HEATED
LINES
SAMPLE
COLLECTION -
POINT
H
TWO STAGE LIQUID SCRUBBER
WITH ICE BAT" COOLING
PRIMARY PARTICULATE
COLLECTION FILTER
90 mm
MILLIPORE ,
0
TWO 1/min
BATTERY POWERED
MSA PUMP
FLOW
L l tr\ <
LDER |
1
|
1
T
T/C AND
1 METER
METER
f
1
SETTLING TANK
AND
CONDENSATE TRAP ( vp
EXHAUST
ELECTRIC MOTOR
DRIVEN VACUUM PUMP
(High Flow)
-TEST POINT
MASS
SPECTROMETER
WITH INTERNAL
SAMPLE PUMPING
SYSTEM
^1-224
Figure 3-62 Sample Collecting
-------
analyses are summarized in Table 3-17. The techniques used in obtain-
ing the data are described in Appendix B.
As shown in Table 3-17, the analytical results reveal six major con-
stituents for the combustor off-gas, namely nitrogen (T^), oxygen (Op),
carbon dioxide (CCU), carbon monoxide (CO), hydrocarbons (HC), and
water vapor (h^O). All other constituents appear in such small quanti-
ties that their concentration can better be expressed in parts per
million (ppm) units rather than in percentages. The amounts of N2« Oo,
C02., and ^ are typical for combustion exhaust gas while the concentra-
tions of unburned hydrocarbons and CO are unusually high and low, respec-
tively. The low levels of nitrogen oxides (NOX) and High unburned hy-
drocarbon levels obtained under the 300 percent excess oxygen conditions
were surprising when compared to available literature on solid waste
incineration. Later tests gave repeated NOX concentrations between 100
and 200 ppm and HC less than 20 ppm.
Particle Analyses
The flyash collected at the exit of the inertial separator residue col-
lection tank (the secondary participate collection point) was analyzed
by the Trapelo-West Laboratory as well as L>y another outside laboratory,
the Metallurgical _abs., Inc. of San Francisco. Trapelo-West determined
-2
the anions chloride (Cl~), fluoride (F~), sulfate (50^ ). and phosphorus
pentoxide (PpOc"). and also ran emission spectroscopic analyses for the
primary and secondary particulute collection. Metallurgical Labf., Inc.
determined the anions and concentrdtioi of the major constituents of the
flyash from the secondary particulate collection point by wet chemical
3-U9
-------
Table 3-17. ANALYTICAL RESULTS OF COMBUSTOR EXHAUST GAS ANALYSIS
Gas
C0?
c
CO
°2
4.
N?
c
NO
NO
NO,
4,
sox
HC1
HF
NH,
j
H?S
(.
P2°5
t J
CH.
HC
H-
H?0
£
11
»2
HI
n
#1
12
11
#2
#1
#2
#1
#2
1
*2
#1
#2
#1
#2
*1
*2
n
#2
*i
*2
#1
f2
«1
#2
#1
n
*i
12
11
#2
. Trapelo data
5.4%
5.7%
0.2%
0.
-------
methods while the minor constituents were derived from an emission spec-
trum. The results of the flyash analyses are summarized in Table 3-18.
It must be noted that the percentages listed are converted values for
metals oxides. The techniques used in obtaining the flyash analyses
data are described in Appendix B.
Turbine Blade Material Testing
The results of the analyses of the combustor exhaust gas and flyash were
forwarded to the Westinghouse Research Laboratories for use in the tur-
bine blade material testing. These tests were conducted by the Westing-
house Research Laboratories, Combustion Systems Research Group under EPA
Contract 68-03-0049. The tests were conducted with a turbine simulator
operating under the pressure and temperature conditions expected for the
candidate turbines being considered for use with the CPU-400 systems.
The blade material investigated included samples of the types of metals
used in the CPU-400 pilot plant turbine, (the Ruston TA-1500), and sam-
ples of the types of metals used in the blades of the CPU-400 prototype
candidate turbines - the newer Westinghouse and General Electric turbines.
The test results showed no attack other than oxidation, i.e., no sulfi-
dation or "hot corrosion". Oxide scales were left intact even though
the deposit spalled, and no localized a-tack was found other than the
normal penetration of oxide at grain boundaries in *nit» rf th ;,<::»eno
of chlorine, sulfur, and alkali ecdls in the test environment. Conser-
vative extrapolation of the data indicated that a machine life in excess
of 10,000 hours should be obtainable.
3-121
-------
Table 3-18 FlYASH ANALYSIS SUMMARY
Compo-
nent
s,o2
M2°3
CaO
Fe203
Na-,11
K20
*lqO
Ti02
PbO?
5n02
ZnO
CoO
Mn203
MoOj
Ag20
V2°5
SrO
Zr02
B203
BaO
CuO
NiO
Cr2°3
P2°5
SO,
C'
F
Trapelo-West
P r i ma ry
flyash
col lection
point
48.0
19.0
14.0
2.4
6.7
2.4
3.4
1.3
0.1
0.1
0.5
0.001
0.4
0.01
0.001
0.008
0.02
0.01
0.02
0.4
0.05
0.06
0.20
2.2
Labs, Inr.
Secondary
flyash
col lection
point
35.0
19.0
14.0
2.0
13.3
4.8
3.4
1.3
2.0
0.35
0.70
0.001
0.25
0.01
0.001
0.008
0.02
0.01
0.02
0.4
0,05
0.04
0.20
2.2
Metallurgical
(secondary flyash
Wet chem. determ
42.96
26.95
10.64
5.08
5.08
2.96
2.82
2.31
0.24
0.11
n.ooi6
Labs, Inc.
collection point)
Emission spect.
Major
Major
10.0
5.0
4.0
2.4
3.0
1.3
0.25
0.1
O.1-.
0.002
0.3
0.01
0.001
0.009
0.02
0.01
0.02
0.3
0.07
0.09
0.09
0.6
All values ar«" ejor^ssfd in itprcent
3-122
-------
SECTION IV
CONTROLS AND INSTRUMENTATION
Controls and instrumentation were designed and developed to support the
24- and 48-hour test runs defined in tasks LP-8 and LP-11. The controls
and instrumentation used to support the 2-hour test run defined in task
LP-1 were "designed under a previous contract.
Task LP-8 specified 6 hours of continuous system operation in an exhaust
temperature within +_ 20° F of a selected setpoint and 18 hours of
continuous system operation in an exhaust temperature band within +30° F
of the selected setpoint. To achieve this requirement, an analog
(semi-automatic) type control system was developed.
Task LP-11 specified 48 hours of continuous pilot plant operation in
a fully automatic control environment. During this period the exhaust
temperature variance was to be maintained within +_ 10° F of a selected
setpoint for 36 hours and +_ 30° F for 12 hours, or the exhaust tempera-
ture variance could be maintained within ^ 20° F of the selected set-
point for the entire 48 hours. A computerized (automatic) control
system was designed and developed to support this task and the future
task requirements of high pressure turbine testing. The design of this
control system was authorized under task LP-9. Its procurement, fabri-
cation, installation, and checkout were accomplished under task LP-10.
In addition to the above, various gas analysis equipment and techniques
were developed to sample and analyze the combustor off-gas constituents.
4-1
-------
DESIGN AND DEVELOPMENT OF GAS ANALYSIS EQUIPMENT
Off-gas constituents selected for continuous analysis were limited to
those whose presence could cause adverse health effects, or whose de-
termination were useful for the study of combustion efficiency as well
as corrosion problems within the pilot plant system. The exhaust gas
compounds selected for gas composition analyses were: oxygen (02),
carbon monoxide (CO), carbon dioxide (C0?), hydrocarbons (HC), oxides
of nitrogen (NO ) sulphur dioxide (SO ), and hydrogen chloride (HCL).
" £
Gas Composition System
The on-line gas analysis systan developed for this program consisted
of two infrared analyzers for CO and CO,,, a flame ionization detector
for hydrocarbons, two electrochemical reaciton cells for NO and S02,
and one polarographic analyzer for oxygen. These instruments are in-
stalled into a mobile cabinet rack as rhown in Figure 4-1. A schematic
of the on-line gas analyzer system is shown in Figure 4-2. A descrip-
tion of the gas analysis equipment is presented in Appendix C.
An off-line titration technique was developed for determination of the
HC1 level in the exhaust gas. A volume of the exhaust gas was first
bubbled through a solution of sodium hydroxide. The liquid was then
analyzed in the laboratory for the concentration of sodium chloride.
An attempt was made to procure an on-line instrument for HC1 measure-
ment using a gas chromatograph. However, after trying a number of dif-
ferent columns and packing materials, it was determined by the instru-
ment subcontractors that it was beyond the present state of the art.
The main problem in this endeavor was the high moisture level in the
4-2
-------
Figure 4-1 Gas Analyzer Equipment
4-3
-------
FROM
STACK LXHAUST
1' ^--s \XJ
REGULATOR FILTER
GAS
DUMP
={Q^-£<^
T.AST
IUST
PUMP
PRESSURE
Figure 4-2 On-line Gas Analysis System Schematic
-------
exhaust gas along with the relatively low levels of HC1 that had to be
measured (150 ppm). The water tended to interfere with and mask the
HC1 reading.
Particle Analysis
The off-line particle analysis technique used for this test program is
shown schematically in Figure 4-3. Particles were taken at 2 to 5 hour
intervals, and were collected in a special electrolytic solution using
an impinger. Following the collection, the wet sample was analyzed in
the laboratory for the particle count versus particle size, using a
Coulter Model Z analyzer (Figure 4-4). The analyzer determined particle
sizes by passing a current through a calibrated orifice in the particle-
laden electrolyte, forcing a measured quantity of fluid through the ori-
fice, and measuring the change in resistance caused by liquid borne par-
ticles flowing through the orifice and correlating the change in resis-
tance to particle size. The particle sampling and analysis technique is
discussed in Appendix C.
Testing
One significant problem encountered during testing with the gas analysis
equipment has been the attenuation of SOo in the sampling system due to
water condensation. Wet chemistry backup samples were used to determine
attenuation levels and their changes. However, no consistent data were
obtained to allow correction lor SOp loss in tne system.
The results of the gas composition and particle analyses performed dur-
ing the LP-8 and LP-11 tests are summarized in Section V.
4-5
-------
EXHAUST
PROBE
IMPINGER
ELECTROLYTE
SOLUTION
VACUUM
PUMP
FLOW METER
DESICCANT
SAMPLE COLLECTION
VACUUM PUMP
PULSE HLIGHT
ANALYZER
COUNTER
ELECTRODES
TEST SOLUTION
(PARTICLES IN ELECTROLYTE)
MERCURY COLUMN AND ELECTRODES
FOR DRAWING MEASURED QUANTITY OF
FLUID THROUGH ORIFICE
SAMPLE ANALYSIS
221-151
Figure 4-3 Particle Analysis, Off-line Technique
4-6
-------
I
-J
Figure 4-4 Coulter Z Analyzer
-------
DESIGN AND DEVELOPMENT OF SEMIAUTOMATIC CONTROLS AND INSTRUMENTATION
The semiautomatic control system was designed to maintain the hot gas
subsystem operating at a steady state level. Operator intervention is
required only to perform various system checks prior to operation and to
change levels or operating configurations. This control configuration
was used as the primary pilot plant control medium during the LP-8 test-
ing series and served as a backup control source during the LP-11 test-
ing series.
Location of Controls
The semiautomatic control system controls were located in a centrally
situated control room (Figure 4-5) adjacent to the pilot plant. The
following paragraphs describe the primary control loops used in the
semi-automatic control configuration. Simplified block diagrams of the
pilot plant control loops are presented in Section V.
Gas Temperature Control
Originally, during LP-1 testing it was planned to utilize two primary
control loops to control the gas temperature. The first primary control
loop was to have been a high rpsponse loop controlling gas temperature
through control of a mixing air flow (a bypass flow around the combustor
which would be equivalent to a bypass air loop in a turbine system).
The second primary control loop was to have been a low response loop
controlling the combustor temperature through control of the solid waste
feed rate. Short term variations in the gas outlet temperature were to
4-8
-------
i
vo
MOTOR CONTROL PANEL
A
SEMI-AUTOMATIC
CONTROL PANEL
INSTRUMENTATION PANELS
-------
have been damped by the thermal inertia of the first and second stage
separators.
However, preliminary testing of this control scheme uncovered a basic
stability problem in that the bypass air control tended to couple with
the combustion process as a result of the change in the fluidizing air
flow. It has been assumed that the bed would tend to absorb or attenuate
the effects of any unbalance in energy between solid waste input and hot
gas output, but, instead of the bed acting as a thermal inertia, the in-
crease in bypass air (reduced fl ndizing air flow) resulted in increased
afterburning, thereby offsetting the temperature reduction which the
increased bypass air flow was to provide.
The gas temperature control scheme for the pilot plant eventually evolved
as a single cascaded loop, shown in Figure 4-6. The inner loop measured
the gas temperature in the top of the first stage separator, compared
this temperature to a setpoint or commanded temperature, and then
adjusted the commanded solid waste flow rate to correct any errors. The
gas temperature at the top of the first stage separator was used because
it proved to be the highest- temperature in the system, thereby being the
most representative of the combustion process. In the outer loop the
second stage separator (LP-8). or third stage separator (LP-11) outlet
temperature was used as the controlled parameter. Any deviation in this
temperature from an operator selected setpoint would cause a change in
the commanded first stage separator top temperature, which, in turn,
would cause a change in the commanded solid waste flow rate. The inner
looo orovided the means to respond rapidly to variations in the solid
4-10
-------
LEVEL
f t\
SEP. niiT.)
\TEMP /
.h-
Ix'
SI AGE
SETPOINT
TRANSFER J
rnwv/rvfitj s-»
LEVEL J I
EED jf[MPTY
rvnn %-
/1ST \
[ STAGE \
\ SEP. IN./
\sUMP /
^--\
\ COMMANDED *
SOLID WASTE
X^ 1ST STAGE IX' VOLUMETRIC
SEP. TEMP FLOW
CONTACT OPENED WHEN
TRANSFER CONVEYOR
V" SUPPLY FULL
\ (PHOTO ELECTRIC CIRCUIT)
1 -f
J 1
EMPTY
FULL
^
+ ^S ~~
OUTFEED
DRIVE PUMP
DISPLACEMENT
CONTROL
SWEEP
DRIVE PUMP
DISPLACEMENT
-CONTROL
VARIAL.LL
r KLljUt. Ml 1
CONTROLLER
- -
OUTTEED
DRIVE
PUMP
SWI I P
DRIVE
P! Mf
uUTFELD
nravL
HYDRAULIC
MOT OK
SWEEP
DRIVE
HYuRAUl 1C
MOTOR
SECOND STAGE SEPARATOR FOR LP-8; THIRD STAGE SEPARATOR FOR LP-11
Figure 4-6 Gas Temperature Control Loop (LP-8 and LP-11)
-------
waste makeup, or feed system. The outer loop was a relatively slow
responding loop which provided the fine tuning of the system to maintain
the outlet temoerature at the desired level.
Feedline Air Flow Control
Secondary control loops (Figure 4-7) were used to control feed line air
flow rate in the two solid waste feed lines to the combustor. These
loops utilized a pitot tube differential pressure measurement (corres-
ponding to an air flow rate) as the controlled parameter and variable
position throttling valves in the feed Mnes as the controlled units.
Dump Air and Combustor Air Control
Additional control loops (Figure 4-8) were initially incorporated to
control the dump air flow rate (via valve 005) and the plenum inlet flow
(via valve 012) based on pressure measurements. However, problems were
encountered in obtaining a meaningful flow measurement, and it was also
established that the position of these valves did not have to be adjusted
once a stable condition had been reached. Therefore, these add-'fional
control looos were eliminated and the valve positions were subsequently
controlled directly by the operator.
Fisher Process Controllers
Fisher process controllers were incorporated into all these control
loops. These controllers contain readouts for the controlled parameter,
commanded setpoint and controller output signal, and adjustments for
proportional loop gain, lag time constants and lead time constants.
These Controllers can be operated in either a manual mode or an automatic
l-l ?
-------
AIR ' "
PI nu
f
OPERATOR
SETPOINT
1
PROCESS 4-2() MA SIGNAL ,!^^,.TO
CONTROLLER * ^NVERTER
<"t-c.O MA !
r!r,r;AL
4-20 PS1 AIR j
DIFFERENTIAL
PRESSURE N
TRANSDUCER ^ T
0-5" H20 ft ~=~~^ i
^O- V^^ r-^/
/
1 y^\
/
AIR
221-226
Figure 4-7 Feedline Air Flow Control
-------
DISTRIBUTION
PLATE
DIFFERENTIAL
PRESSURE
PROCESS
CONTROLLER
AIR
DUMP CONTROL
VALVE
(005)
SETPOINT
COMBUSTOR
INLET FLOW
PITOT TUBE
(DIFFERENTIAL
PRESSURE)
PLENUM INLET
CONTROL VALVE
(012)
SETPCINT
221-227
Figure 4-8 Dump Air and Combustor Air Control Loops
-------
mode. In the manual mode, the test conductor actuates a thumb wheel
which causes an increase or decrease in the iignal being sent to the
associated unit controlled. A meter on the front of each controller
provides an indication of the signal being sent to the controlled unit.
In the automatic mode the process controller compares a signal from
measurement of a preselected system parameter to a selected level or set-
point. A_switch on each controller allows this setpoint to be set in
locally by the operator or remotely by the computer. The controller then
increases or decreases the signal to the controlled unit based upon the
difference between the measurement and the setpoint.
Mode Switches
Starting and stopping the system and switching operating modes were
accomplished through use of mode switches (Figure 4-9) designed to mini-
raize the load on the operator. These switches were vr,red to preclude
entrance or switching to any operating mode unless certain prerequisites
were met. These prerequisites were established to insure that the system
was ready for safe operation in the selected mode. When the prerequi-
sites were met, the mode switch would turn on appropriate motors, valves,
and sensors for the selected operating mode. If the prerequisites were
not acceptable, the system operating state would remain unchanged. Also,
if for some reason one of the prerequisites changed state during ipera-
tion in a specified mode (indicating an unsafe condition), the system
would automatically shut down.
4-15
-------
LOW PRESSURE OPERATING HOPES
+24 VDC
PREREQUISITE CONTACTS
(CLOSED WHEN SYSTEM
ELEMENTS ARE IN SAFE
OrZRATING STATES FOR
THE APPROPRIATE MOOE.]
H HHH
01
NORMALLY CLOSED
CONTACTS OPENED <
BY OTHER MODE
RELAYS
1 UACKHEAT
2 FLUIDIZED OPERATION ON DIESEL FULL
3 FLUIDIZED OPERATION ON DIFSEL FUEL AND SOLID WASH
4 FLU'DIZED OPERATION ON SOLID WASTE
5 FLUIDIZED OPERATION WITHOUT FUEL
MODE SWITCH
(TOGGLE SWITCH
WITH MAGNETIC
LATCHING.)
VDC
NORMALLY OPEN CONTACTS TO
TURN ON APPROPRIATE SYSTEMS
ON THE FLOOW
NORMALLY CLOSED CONTACTS
IN LATCHING CIRCUITS OF
OTHER
Figure 4-9 Basic Mode Control Circuit Logic (Illustrated for Three Modes)
-------
Temperature and Pressure Measurements
Instrumentation was primarily associated with temperature and pressure
measurements. Thirty chomel-alumel thermocouples were distributed
throughout the system. Two of these were used in the control system as
previously noted; the remaining 28 were used for operational analysis.
Six pressure (or pressure difference) measurements were required to de-
fine system operation and four pressure head (flow) measurements were
used to monitor air flow. The location of the instrumentation is de-
picted in the tesc setup schematic aoolicable to each major test series
discussed in Section V.
Data Acquisition System
Pressure and temperature measurements were recorded in a data acquisition
system (Figure 4-10 and 4-11). A scanner, digital clock, data coupler,
teletypewriter, digital temperature indicator, and a digital multimeter
comprise this system.
The scanner transfers multiple analog input signals sequentially, under
internal control, to a common set of output lines. Numbered front panel
pushbutton switches permit skipping of unwanted channels. The push-
buttons are illuminated during the scan cyclo to notify the operator of
the scanner's position. A digital clock controls the scan process
through 63 different selectable time intervals and provides a time head-
ing for the data prior to each scan.
A data coupler resembles a small, hard-wired processor. It consists of
a mainframe, data control buses, device input cards, devicr output cards,
4-17
-------
00
MEASUREMENTS
MINI
SCANNER
DIGITAL
CLOCK
DIGITAL DIGITAL
TEMPERATURE MULTIMETER
INDICATOR
t '
1 - 4
DEVICE
INPUT
CARD
DEVICE
INPUT
CARD
DEVICE OLV1CI
INPUT INPUT
CARD CARD
ir.TERr.Ai. 1 11
fO'JROL DATA AND CONTROL P. H,fS J
CIRCUITS 1 '
DATA CO'JfLtR
CijDF
CARD
1
DtVKE
OUTPUT
CARD
TELETYPE
Figure 4-10 Data Acquisition System Block Diagram
-------
Figure 4-11 Data Acquisition System
4-19
-------
code cards, and formatting matrices. This unit couples the digital
output devices (multimeter, thermometer, or clock) to the teletypewriter;
and it also allows for formatting the data (i.e., provides the capability
to include engineering units in the record).
Thermocouple measurements are converted from millivolts to temperature
via the digital temperature indicator. This unit provides linearization
ior the thermocouples (type K) along with a reference junction compensa-
tion. Voltage and resistance measurements are made via the digital mul-
timeter. This unit is capable of autoranging and of having the mode
externally selected.
DESIGN AND DEVELOPMENT OF AN AUTOMATIC CONTROL SYSTEM
The automatic control system was designed to provide full control of
the pilot plant with a minimum of operator intervention. This includes
automatic verification of system readiness, automatic verification of
valve operability, automatic transition from one operating mode to ano-
ther, and automatic monitoring of the system for alarm or out of toler-
ance conditions.
Location of Controls
Most of the automatic control system controls are located in a centrally
situated overhead control room adjacent to the pilot plant. Some of the
control operators, dedicated to specific elements of the system, are
Incated in the proximity of the controlle unit in order to reduce cable
run lengths and to reduce the number of signals being sent to the control
room.
4-20
-------
The control area is a two-story arrangement with signal conditioning,
patch panels, transmitters and termir.al strips located on the lower
floor, while the control consoles and a process control computer are
located above in the control room (Figure 4-12). Where practical,
control and instrumentation (data gathering) functions are separated in
order to minimize interference during system operation. Air condition-
ing and conditioning of the electrical power for the control system are
provided in order to insure reliable operation of the electronics.
Process Control Computer
The primary function of the computer is to control the pilot plant pro-
cess through an entire operating cycle. This includes monitoring of
various operations within the process for out-of-tolerance or abnormal
conditions. In this capacity, the system operates as a real time com-
puter, the performance of each operation being based on a specific time
for execution. A basic block diagram of the computer with the process
is shown in Figure 4-13.
A background function of the computer during this process control func-
tion is data acquisition; reading various voltages, converting these
voltages to engineering units, processing the information, storing the
results and/or printing, punching, or recording the information on one
of the digital system output devices.
The process control computer selected for this application is the Texas
Instruments Model J60A Digital Computer. It is shown schematically in
Figure 4-14. The following paragraphs provide a brief Description of
each major element in this system.
4-21
-------
CnVTVH. ASSY /
(12-IJNIT)
Figure 4-12 Pilot Plant Control Room
-------
A.'IALOG SENSORS
THERMOCOUPLES
PRESSURE TRANSDUCERS
POSITION TRANSDUCERS
VELOCITY TRANSDUCERS
TRANSMITTERS
MV 10 CURRENT
PS I TO CURRENT
VOLTS TO CURRENT
INPl'T1
PPOCFSS
CONTROLLERS
POWER CORTROLLERS
VALVES
SERVO CONTROLS
COMPUTER
MGITAL SiJi
PELAY1,
LWITCHES
ALTERNATE OUTPUTS
POWE_R CONTROLLERS
PELAVS
ALARM
LELL
IIGHTS
Figure 4-13 Generalized Control System Schematic
-------
t*
no
DIRECT MEMORY
ACCESS CHANNEL
(DMAC)
CPU
DIABLO DISC
ANALOG OUTPUT
SIGNALS
DIGITAL
INPUT/OUTPUT
CHANNELS
MONITOR LAFT
CPU
SEMICONDUCTOR
MEMORY
730 TERMINAL
HAZELTINE CRT
Figure 4-14 Computer System Block Diagram
-------
Central Process Unit-
The heart of the system is a 16-bit word length mini-computer (CPU) with
20,480 words of semiconductor main memory, a battery pack for continuous
memory refreshing, memory parity, variable length memory protect, power
failure interrupt, extended arithmetic option, expanded internal communi-
cations register unit (CRU) which provides for computer/process inter-
facing, and a control panel with a keylock.
Mass Data Storage Unit-
The mass storage unit is a 1.14 million, 16-bit word moving head Diablo
disc pack and controller. It interfaces directly with the CPU through
a direct memory access channel (DMAC). The disc pack has an average
access time of 100 ms, and is used for process program compilation,
process task storage, proces-s data storage, task overlay operations,
systems generation, and for storage of overall systems process control
programs configuration.
Telewriter-
The silent 730 teletypewriter is a 30-character per secord teletype that
is used as the system's main data logging device, and the primary pro-
grammer's interface peripheral..
Alpha-Number Display Terminal (CRT)-
The CRT is an alpha-numeric display terminal with a maximum datr
transfer rate of 2400 baud. This device is used as th? nrimary process
operator's interface peripheral.
4-25
-------
Paper Tape Punch Unit (PTP)-
The eight column paper tape punch unit is a 60-character per' second de-
vice that is used for permanent storage of process source programs and
process data.
Paper Tape Reader Unit (PTR)-
The paper tape reader unit (eight columns) is: a 300 character per second
device t^iat is used to load memory from source taoes, and from stored
process data tapes.
Interval Timer-
The interval timer is a 1 ms resolution device that provides controller/
orocess synchronization, and a system's real time clock capability for
time of day acquisition and recording.
Low Level, Differential, Analog Input Subsystem-
The low level A/D subsystem provides 30 channels of 0.05 percent accu-
rate, low level analog inputs fur measurement nf process temperatures
utilizing the thermocouple method.
High Level, Single-Ended, Analog Input Subsystem-
The high level A/D subsystem provides 80 channels of 0.05 percent accu-
rate, high level analog inputs for measurement of process transducer
outputs. Some examples of process outputs are pressure, level, speed,
position, setooints, voltage, cur-^nt, power, flow rate, and weight.
4-26
-------
High Level, Single-Ended. Analog Output Subsystem-
The high evel analog output subsystem provides 18 channels of 0.05 per-
cent, accurate, high level analog outputs for process setpoint generation.
Discrete Contact Closures (11)-
The 160 sets of isolated, buffered, contact closures are used to control
the energizing and de-energizing of numerous process control relays and
contactors for programmed sequential plant operation.
Discrete Contact Sense (12)-
The 160 channels of the discrete contact sense subsystem are used to
monitor numerous process contact closures, to ensure cotrect sequential
plant operation, and to indicate urgent and s^mi-urgent process alarm
conditions.
Digital Input Subsystem (13)-
Thirty-two bits of the 96-bit digital input subsystem are used to monitor
the modes of operation of the various process controllers. The remaining
64 bits are used in conjunction with the 64-bit digital output subsystem
to interface the computer system with the process data acquisition system.
External Interrupt Subsystem-
The 16-channel external interrupt subsystem is used to generate immediate
mode process alarm conditions, to get supervisor control for certain
peripheral devices, and to get immediate operator control of the process
under predefined conditions. Each external interrupt will cause the
computer to trap the predefined memory locations for programmed servic-
ing of that particular interrupt.
4-27
-------
Software
The remainder of tMs section describes the software approach to the
process control, monitoring, and data acquis4tion functions.
Vendor Supplied Software-
To accomplish the required operations in a real time environment, a
multl-programming operating system denoted PAM/D* was utilized on the
Texas Instruments (TI) model 960A computer. The PAM/D monitor system
uses an executive worker method of program control and incorporates a
multi-level priority scheme for program execution.
Several high priority software elements ("tasks") are supplied by TI
as part of the PAM/D system and are available throughout all process
operations. Two of thase, denoted DEBUG and JOB CONTROL, provide a
basic means of directing and controlling tht execution of application-
oriented tasks. These software packages comprise the principal Inter-
face between the programmer and the computer during program development
phases. Following Initial development, however, many of the Important
functions performed by these packages can more efficiently be handled
by a high priority, pilot plant process oriented task. In this approach
test conductor personnel can communicate with the process and the coi.ipu-
ter In familiar terms, thus avoiding the need to become proficient in
DEBUG and JOB CONTROL coding. Accordingly, little additional attention
is addressed to these software elements 1n the remainder of this dis-
cussion.
An acronym for Process Automation Monitor/Disc
4-28
-------
A basic decision was made to employ the PCL^ language wherever possible
in implementing pilot plant software. The relatively high level PCI
language offers numerous programming advantages such as ease of expres-
sing arithmetic operations on real numbers, convenience in certjin types
of control and input/output (I/O) operations, and reduced actial pro-
gramming labor costs. Nevertheless, there is a significant Mst of limi-
tations and constraints in the present version of PCL; this list clearly
indicates that PCL should not be expected to compete with contemporary
large-machine FORTRAN compilers.
The next highest level language supplied with the 960A is the Symbolic
Assembly Language (SAL) . This language, which produces machine language
code in a two-pass assembly, includes a relatively powerful set of 78
basic instructions. Since PCL uses a one pass compi ,er to produce a
corresponding re-entrant SAL program, it is possible to include in-line
SAL coding commands in the original PCL program. This technique, to
be avoided wherever possible, unless particularly advantageous, can be
used to remove some of the PCL limitations.
An acronym for Process Control Language, an extension of ANSI baiic
FORTRAN II.
I
A SAL command in a PCL program, distinguished by a period in the first
column, is passed through compilation intact to join compilation gener-
ated commands. One example of necessary SAL coding is to interrogate
the real time clock in order to generate the time of day.
4-29
-------
Another important element of vendor-supplied software, known as LIB/
LNK960**, is utilized to link the object modules produced by the assem-
bler for main programs ,md the required subroutines. Output from this
program constitutes the overall object tape which can be loaded for
execution by the computer.
Pilot PI ant-So f t ware^ Structure-
The multiple task capability of the TI960A was fully utilized in the
applications software written for the pilot plant. A set of 23 tasks
comprise the process control software configuration used for low pressure
pilot plant testing operations.
In the PAM/D operating system it is possible to have ono task bid another
It is also possible for a task to "suspend" itse f. In the suspended
state, a task remains in memory with its instruction counter fixed at
a known statement in the program. This state can be terminated by one
of two methods. First, another executing task can "unsuspend" the task
whenever specific conditions are satisfied. Secondly, a specific pro-
grammed external discrete signal from outside the computer can be used.
This external signal can be either a process oriented contact sense
line or an operation oriented switch line.
The foregoing capabilities have been developed into concepts of '"execu-
tive" and "worker" tasks. In the pilot plant software, for example.
** An acronym for Libraiy Link
4-30
-------
The overall executive -sk 10 is first bid bv the operator using a
DEBUG command. Most of the subsequent task bidding is then performed
automatically by the tasks themselves. Typically, for the greater per-
centage of the time, executive tasks are in the suspended state, wait-
ing to be unsuspend^d by either an interrupt or a worker task that has
completed some orerequisite activity.
A basic feature ->f the PAM/D system is that many tasks may be in the
"bid" state simultaneously. These tasks are loaded into various sepa-
rate portions of the computer memory (management of this function being
a prime attribute of PAM/D) and receive attention according to their
various assigned priority integers. Each 0.1 second a new scan of bid
tasks is begun by the monitor and those tasks that are not suspended
are serviced.
Common data base - The principal means of inter-task communication is
a system-wide common data base (CDB) that is accessible to ill tasks.
The CDB itself occupies a special portion of computer memory beginning
with hexadecimal location BO anr^ extending for 426 words.
A discussion of the CDB reveals many of the software structural design
concepts. The structure is prpcisely defined by fhe group of PTL
COMMON and DIMENSION declarations shown in Table 4-1. Thp meaning and
use of the various variables cited in the table are summarized below:
INDXI A 72 row by 2 column array of integers that form an index and
scan control scheme for the analog input channels sampled by
the computer. Information pertaining to the 3-digit CPf mea-
4-31
-------
Table 4-1. COMMON DATA BASE DEFINITION FOR LP-11
Decla ation
C0MM0N
C0MM0N
C0MM0N
C0KM0N
C0MM0N
C0MM0N
DIMENSION
DIMENSION
DIMENSION
DIMENSION
DIMENSION
Variable
(ZBO) INDXI, MDATA, NM[ AS. I MEA3
INDX0, H0UTP, N0LITP, I0UTP
PARAM
M0DE. IQU'T, IREAD, III
IDLES, LTIME
L0G, MSTAT. STAT
INDXI (72,2), MDATA (7?)
INDX0 (18), M0UTP (18)
PARAM
LTIME (4)
MSTAT (20). STAT (10)
4-32
-------
surement number, the high (or low) level of the signal, and the
sampling frequency is packed into the first column for each
channel. The analog converter channel number is stored in the
second column.
MOATA A 72 member array of integers that contain the most recent re-
sults of sampled, digitized conversions on each of the analog
input channels. The measurement associated with the i_ row of
INDXI appears in the i element of MDATA.
NMEAS, A pair of integers that implement an asynchronous operator or
IMEAS
programmatic command for a read of a specific measurement. To-
gether with the automatic periodic sampling, this mechanism is
presumed to maintain sufficiently current information in the
MOATA array about the process state.
INDXO This integer array associates a 3-digit^CPC number with each
of the 18 available analog input channels. It thus provides an
Index function that enables program reference to the CPC num-
bers.
M0UTP This 10 member integer array is a register for the last output
values sent from the computer to each of the analog output chan-
nels.
N0UTP, Two Integers used in a way similar to NMEAS and IMEAS for oper-
I0UTP
ator Introduced changes to analog output channels.
PARAM A 60 element real number array containing adjustable parameters
for process control and monitoring. Certain key measurements
1n engineering units are also stored here. A computer readout
-------
of this array is presented in Figure 4-15.
MODE An integer that reflects the commanded mode of pilot plant op-
eration. For example, MODE^l pertains to bed preheat opera-
tions.
IQUIT Task termination integer. Most looping tasks look to this num-
ber; if their ID number occurs, they branch to terminate.
IREAO An integer argument of a WAIT statement that determines the ba-
sic analog input sampling ,'requency. frequency - 0.l/(IREAD+1)
reads/second.
ILI An integer for general task control communication purposes.
IDLES An integer argument of a WAIT statement for timing control be-
tween the executive logging tasks and their worker tasks.
LTIME An integer array where the start time (year, day of year, hour
minute, and second) associated with the task 50 (i.e., flui-
dized operations) start point is stored.
LOG An integer argument of a WAIT statement that determines the ba-
sic logging frequency. Same relationship as for IREAD.
MSTAT An array of integer stdte variables involved in digital filter-
ing of analog input measurements.
STAT An array of real «tate variables involved in digital filtering
of computed performance parameters (e.g., flow rates).
Many of the common data base elements are filled with initial values by
reading from a disc data file at the beginning of software operations.
Those initialized in this fashion include INDXI, INDXO, PARAM, and
IREAD. The display in Figure 4-15 is a. listing of that portion of the
-------
r -,.- f r. I^;>-.T dtp n:'i. 'FT fCIMT- IF HIM
' , f>"K.i ( p.'f . '"-HIM
Figure 4-10 Parameter Section of Coitmon L)ata Base
-------
disc file that is used as a source for the initial PARAM array.
A summary identification of the 23 tasks used is contained in Table
4-2. In Table 4-3, a normal chronological sequence of events is out-
lined to provide additional insight. For reference, a PCL listing for
the principal task, task 50, is included as Figure 4-16.
Executive task 10 conmands - A major function of executive task 10 is
to provide an opei Jtor/con-.u-jter/process interface at the CRT console.
As previously noted, task 10 is normally suspended and waiting for an
interrupt. The external devide providing the interrupt is the ACCEPT
button near the CRT: depression by the operator constitutes the inter-
rupt that unsuspends the task.
After interruption, task 10 normally calls a communication subroutine
that receives one of several comnands from the CRT keyboard. These
commands begin with a two-character code that is normally followed by
a three-digit integer, and may be followed by an additional value. The
definition of the commands used in LP-11 testing are described in Table
4-4.
It should be noted that all. DEBUG and JOB CONTROL commar.Js are avail-
able during process control operations. This provides a straightfor-
ward method for real-time display and modification of all integers in
the COB, including the index array elements.
4-36
-------
Table 4-2. TASK IDENTIFICATION SUMMARY
I.D.
Description
10
A2
20
30
31
32
33
50
15
BO
Bl
17
bl
B4
B5
Bf
87
B8
B9
BB
BC
BO
BF
Overall execution
Reads initial COB data, zeroes all outputs
Analog input scan
Controller readiness checks, transducer zeroes
Valve exercise
Preheat mode initialization
Preheat mode ignition, control, and monitoring
Fluidized operation control and monitoring
Preheat logging executive
Preheat control parameter logging
Preheat pressure and temperature logging
Fluidized logging executive
Data filtering - high speed
Data filtering - low speed
Solid waste flow logging
Bed temperature logging
System pressure logging
System pressure drop logging
Miscellaneous temperature logging
Proportional valve position logging
Control system parameter logging
Air flow calculation and logging
Time origin logging
4-37
-------
Table 4-3. COMPUTER SOFTWARE TIME SEQUENCE OF EVENTS
Step Event
1. The DEBUG command actuated by the operator bids task 10, the primary
executive task.
a. Tas.k 10 bids task A2 to initialise the common data base (CDB)
with values read from disc file 84.
(1) IREAD, an integer that (roughly) says how many seconds
between each basic analog input scan cycle.
(2) 72 sets of MEAS, LEVEL, IFREQ, NCHAN integers to form the
INDXI array for analog input channel scanning.
(3) 18 integers of the INDX0 array for analog output.
(4) 60 real values to initialize the parameter array.
b. Task 10 bids task 20, the analog input scan task.
c. Task 10 bids task 30, the pre-test health check task (see stop 3).
d. Task 10 then suspends itself to wait for the ACCEPT button inter-
rupt. If interrupted, it writes a message ("waiting *or task 30
to terminate") and checks to see if task 30 is done. If not,
it loops back to wait for another ACCEPT interrupt. If done,
see step 4.
2. Task 20 immediately starts scanning 72 analog input channels, storing
the resultant U'liHje'l rnnvpr'=, i< r $ in rUi;"V!'i i itv flcrpnt1, (it the
MDATA array The MDATA array is also part of the CDB so its contents
are available to all other tasks.
The periodic scan algorithm features 4 different input scan frequen-
cies (unless a shared transducer is used) in addition to a "never-
read" choice if IFREQ^O for the channel. An index, J, is cycled
from 1 to 0, making a sten ever/ IREAD seconds.
a. J Index 1 -
(1) C0730 and C0737 are closed so that pressures 101 and 123 re
switched to transducers 101 and 123 for later reading when
J 2.
4-38
-------
Table 4-3 (Continued). COMPUTER SOFTWARE TIME SEQUENCE OF EVENTS
Step Event
(2) All channels with IFREQ=1 are read and stored in oroper
MDATA elements.
b. J Index 2 -
(1) All channels with IFRE°=1 or 2 are read and stored.
(2) Pressures 101 and 123 are read and stored in MDATA(30) and
MDATA(37). C0730 and C0737 are opened.
c. J Index 3 -
(1) C0731 and C0738 are closed so that pressures 102 and 124 are
switched to transducers 101 and 123 for later reading when
J 4.
(2) All channels with IFREQ=1 are read and stored.
d. J Index 4 -
(1) Ai, channels with IFREO=1,2,4 are read and stored.
(2) Pressures 102 and 124 are read and stored in MDATA(31) and
MDATA(38). C0731 and C0738 are opened.
e. J Index 5 -
(1) C0734 nd C0739 are closed so that pressures 107 and 12C are
switch.-d to transducers 101 and 123 or later readinq when
J = 6.
(2) All channels with IFREOrl are read and stored.
f. J Index 6 -
(1) All channels with IFREQ=1 c,r 2 are read and -.tored.
(2) Pressures 107 and 125 are read and stored in MHATA(34) and
MDATA(39). C0734 and C073'J ore opened.
4-39
-------
Table 4-3 (Continued). COMPUTER SOFTWARE TIME SEQUENCE OF EVENTS
Step Event
g. J Index 7 -
(1) All channels with IFREQ=1 are read and stored.
h. J Index 8 -
(1) All channels with IFREQ=1,2,4 or 8 are read and stored.
3. Task 20 returns to reside J=l repeatedly until it sees IQUIT=20, the
latter havii.j been set Ly the executive task 10 when it decides to
stop 20. In that case, task 20 opens all contact closures, sets
IQUIT to 0, and terminates.
4. Task 30 is a pre-test health check task that accomplishes the follow-
ing in a one-pass sequence:
a. Writes an initial message (CRT and 700).
b. Verifies that protectal system contact CI733 is closed. If not,
it writes a message, and continues checking every second. When
corrected, the task writes an OK message before continuing. On
the other hand, if there never was a fault, no message is written
at all; the task simply moves to the next check at once. This
same pattern is followed in succeeding checks.
c. Verifies that CI615 is closed indicating sufficient facility
air pressure.
d. Checks each of eight pressure transducers to see whether they
are putting out 1 vdc t 0.1 (i.e., 0 psig). The sequence is:
121, 120, 146, 140, 141, 101, 123, and 142.
e. Verifies that CI706 is closed indicating a satisfactory diesel
oil level.
Two other contact closures are controlled by task 20. At. every step
in J, the task looks at M0DE (also in CDB). If M0DE <_ 1 (as in the
present case), C0735 and C0744 are closed to switch polarity of pres-
sure transducers 120 and 121 for the preheat mode. Otherwise, they
are opened.
4-4D
-------
Table 4-3 (Continued). COMPUTER SOFTWARE TIME SEQUENCE OF EVENTS
Step Event
f. Verifies that CI730 Is closed Indicating that the transfer con-
veyor belt has not drifted.
g. Verifies that CI731 is closed indicating that the weighing con-
veyo- belt has not drifted.
h. Verifies that five temperature transducers are operable by
checking for an onen condition on the following:
CI506 (Bed)
CI507 (Freeboard)
CI504 (I/S2 outlet)
CI503 (A/S top)
CI505 (I/SI outlet)
i. Displays recommended settings for controllers 1 through 8, in
turn, pausing after each message with an ATTENTI0N note. Pres-
sing CONTINUE causes the task to move on.
j. Verifies that each of 7 "remote setpoint" switch controls are
closed.
Contact Controller
CI905 1
CI986 3
CI980 4
CI982 5
CI983 fi
CI981 /
CI971 8
k. Writes some termination messages, sets IQUIT=30 (for executive
task 10 in step 4), and terminates.
5. Now that task 30 has concluded, task 10 bids the valve exercise
task 31 and again suspends itself waiting for an interrupt at the
ACCEPT button. If interrupted in this loop it writes a different
message and checks to see if task 31 is done (i.e., IQUIT=31), If
not, it loops back for another ACCEPT interrupt. If 31 is done,
see step 7.
6. Task 31 moves various system valves and verifies appropriate response
from associated position transducers and/or switches. It also is a
one-pass task.
-------
Table 4-3 (Continued). COMPUTER SOFTWARE TIME SEQUENCE OF EVENTS
Step Event
a. Task 31 first writes two opening messages, one of which is a
suggestion to advise personnel to stand clear of the valves.
Then it pauses; press CONTINUE to move on.
b. Dump, valve 005 should initially be closed because of the shorting
contact (C0820=0). This is verified (CI736=1 and CI223<0.1 vdc).
Then C0820 and C0814 are closed so that controller 1 can receive
a setpo-'nt of -3.0 volts from C0005. Within 8.0 seconds the
computer looks to see if CI736=0 and CI223>9.5 vdc as should
happen if the valve goes full open. I each case above, 'ailure
produces a message and a pause. Then pressing C0NTINUE starts
a new try. When successful &n OK message occurs and the task
moves on. Valve 005 is left open for preheat. This same general
pattern is repeated in the following steps:
c. Air plenum inlet valve 007 is commanded closed (it may already
be closed) by 1.0 volts on C0007 through controller 2. Contact
C0013 is first closed to permit this. The verifications are
CI724=open, CI725=closed, and CI222<0.1 volts.
Then 5.0 volts is put on C0007 and the task later looks to see
if CI724=closed, CI725=open, and CI222>9.5 volts.
Valve 007 is left open for preheat.
d. Five valves are then cycled open and closed in turn. In each
case, first 5.0 volts and then -3.0 volts are applied to a
specific analog output channel and a potentiometer signal is
nonltored.
Prior Analog
digital output Analog
Valve outputs channel Xducer
Diesel oil 002 C0822 = 1 050 236
C0808 = 1
C0809 = 0
C0810 - 0
Feed line 000 C0811 - 1 000 220
Feed line 001 C0812 = 1 001 221
I/SI Residue 009 C0825 = 1 009 225
A/S Residue 010 C0816 = 1 010 226
4-42
-------
Table 4-3 (Continued). COMPUTER SOFTWARE TIME SEQUENCE OF EVENTS
Step Event
-e. Feed line air shutoff valve 729 is opened. The valve is left
open for preheat.
f. Preheat air shutoff valve 704 is opened using C0704=l. It is
left open for preheat.
g. The task writes some termination messages, sets IQUIT'Ol (for
executive task 10 in step 5). and terminates.
7. After task 31 concludes, task 10 bids the preheat initialization
task 32 and again suspend:, itself waiting for an interrupt at the
ACCEPT button. It remains in this loop, able to service operator
CRT commands, until IQUIT=32 is observed.
8. Task 32 conducts the following operations preceding ignition of the
preheat burner:
a. Positions system valves for the prtheat operating mode,
b. Turns on aopropriate residue removal system components to imple-
ment a purge exercise,
c. Turns the main blower on,
d. Steps valve OU/ partially closed to set up an initial air flow
condition through the preheat burner,
e. Implements a 30 second preheat air flow control loop so that
measured flow approaches the value stored in parameter 01,
f. Turns off the residue removal system components,
g. Sets IQUIT=32 to inform task 10 of completion.
9. Afte- task 32 concludes and ACCEPT is depressed, Usk 10 bids preheat
ignition and control task 33. It then enters a loop containing the
usual suspend-on-ACCEPT interrupt feature.
10. Task 33 takes over the preheat sequence, performing the following
operations:
a. Initiates preheat logging executive task 15 (see item 10).
4-43
-------
Table 4-3 (Continued). COMPUTER SOFTWARE TIME SEOUKNCE OF EVENTS
Step Event
b. Generates the ignition spark pulse, opening the propane valve
and oil valves, checking for established flame,
c. Continues the burner air flow control begun in step 8e,
d. Refreshes a setpoint output to controller . for analoq compari-
son with measured treeboard temperature (2nd subsequent oil flow
control).
e. Monitors bed, freeboard, and inertial separator stage 2 outlet
gas temperature,
f. Initiates a termination shutdown sequence when bed temperature
exceeds the readiness value stored in parameter 24. Included
is setting IOUIT=33 to communicate with task in in the normal ,
way. Also included is termination of preheat logging executive
task 15.
11. Preheat logning executive task 15 is active in parallel with task 33.
It first bids task BF which stores a time origin on disc file B4,
and then periodically bids two worker tasks during preheat operations:
a. Task BO logs 11 control parameters of interest,
b. Task Bl logs 11 selected pressures and temperatures.Logging
frequencies for thp two tasks are individually adjustable.
12. After task 33 concludes and ACCEPT is depressed, task 10 bids flui-
dized control and monitoring task 50. It then enters a loop con-
taining the usual suspend-on-ACCEPT interrupt feature (see step 15).
13. Task 5i supervises and monitors operations in the fluidized mode.
Actions include the following:
a. Appropriate pre-positions system valves,
b. Activates residue removal system components (transport blower,
baghouse motors, valves, hopper vibrators).
c. Activates main blower,
d. Bids fluidized logging executive task 17,
e. Refreshes valve 007 position command based on parameters 36 and 37,
4-44
-------
Table 4-3 (Continued). COMPUTER SOFTWARE TIME SLQUENCE OF EVENTS
f. Supervises coni-pi Of the setpomt to controller 4 such that
T031 approachr , the value stored in parameter 48,
q. Refreshes the setpoints for feedline air flow controllers,
h. Monitors bed, freeboard, and first stage separator top tempera-
tures,
i. Performs bang-bang control of bed temperature by opening/closing
bypass valve 704,
j. Generates a periodic CRT and 700 log disolay.
k. Conducts an orderly shutdown sequence including terminating
executive logging task 17.
14. Fluidized logging executive task I/ is active in parallel with task
50. It first bids fsk BF to ^tore a time origin on disc file 64.
Then it periodically bids a string of worker tasks, most of which
also generate disc file records.
a. Task 83 digitally filters da'a to be logged by tasks B5 and B7.
b. Task B4 digitally filters data to be logged by task B8.
c. Tas. 85 logs solid waste flow rate and cumulative values.
d. Task 86 logs bed temperatures.
e. Ta^k B7 logs system pressures
f. Task B8 logs system pressure drops.
g. Task 89 logs miscellaneous system temperatures
h. Task BB logs proportional valve positions.
i. Task BC logs control system performance parameters.
j. Task BD computes and "ogs system air flow rates.
15. After bidding task 50, pvecutive task 10 is normally suspended in
an operator service '.cop. The only exit from this loop is through
setting IQUIT=10. When task 50 and other tasks have terminated,
this is the normal final step in clearing the memory of all tasks.
4-45
-------
r
c
c
rpiitipf-i . _-f-,, ]f,;, j.
CL:.'"".'1< ,ri'.J. "I TF . hD.'TF1- .'Off
C'Jf""''"i k >-«H" '.(,'*' !. T or ' . r '! . ' i i.: . ; f i .
1 T''i.\ -F ! >! .t ' . - .F !,: ' .F U 1
f.FfiT1'! i-i'TF i> ''I T . :tt»! . II. i
I'P ' . HIM !;; 1 -: .,' . 'TIM,". r_- .
Mr1 i, \rt> ;;. " i - . nfi'TF r: .
liir^r- !':! t,-*^--! j'. ,
FL^-C- tFH. '.'^ Tb
C
C .CPU .*re IFICHTICN:
L
LFLI Cl'Tfi.'T ':0'''c' ' CC'O-- ' CC^OTT'
CPU CK'TFUT 'T'lI'-Mi'. . C'.;.- -7 Ur CL^ir-
>-i O'/ri ,'T 'T'.]' :'. i.T'..: : n«; LOI-!-
CPU CJijTFi.iT '-[.'; ii. JJ i,- . n.-i.i , -: i ;-: , , -_(]-,
CF-ii rUTFUT ':C7i>- ' jr^ . . rCT"-?-T
CPU UUTK'T ru~<-^-,'f ~ . CC' '.w1 ->
CFij CM.ITH.IT ''c;\ « '.": 'rcTc.'-r
CPU OUTPUT I'C'-J7^. ">'l ' -I'L'Oii : T!
CrU ir'f-'uT FMMIC Jr i.' i :i ;u- .> i
H TC ='CTTCM T3 .' C1! '> £ THF -TF FFOILEM
C-OTO /rnjii
TUFrt f-Hir FLDI CF ri--. "iM* i? :f i;'"i> FP
I : N CiM.L r; .fiC-C r". - '
i. MLt ' f-r|l.'I ' 'i 1
CM! : ! ii 1 I
LH " ' 1
Figure 4-16 Task 50 Listing
4-46
-------
c
c VMLVE ori.*
c
IFt vT-f t-'ff'i^ ?*. , i ; r ;/ . it 0 ?? 0
370 \--ef-n<.(,f\, >
38U CfiLL M^'CUT' 7 . v;*1"
r
c rite or ; :.? OUTLET itr.f HCPTI>T
c
rP- 71 '
' 4? '-Tn?i
VL= 1.0 » 0 . fmi'FHF'MM' 49^
VL- - \.'L « 1 1 . C
i:'ir.i?fi=-l-^ri, 44 ,
F,i£,iM. 4J ,rM''1H1 ' VL -l^
-)c". f'fltt r.' -l.:- .^F;~-
4?4 iTOHTIM'.'E
r
C OIL ;V:TFM HH-OFF LU6IC
0
ODro« ^fiii.^.n-i.r.ciO.f.oi 5ni > .rote
500 >.o-:-t£*i
'. o 1 ( O? ^£ = ci
cn; IM = H
r
IDLII MHlTK tvrTEM DM-OFF I HO- 1C
i.
'.-OTCJ. i- "I .t.nj .^,on .r on -" 01
n' n-i
. u-, i =1
I.IH : 1 1; = i
:-rru -..v."
nl t'.ii" .'".= 0
>; n-r r -- f«
i uTir"''
|i- . i.'M'TF l.'MlTi 49
I.W:TE*H
i. tr:-.n--u
i 0'.. ! - '".
Figure 4-16 Task 50 Listing (Continued)
4-47
-------
c
C UPDftTt FEETl IfE flip VIH.VE CET PniNT
C
CftLL ftt'Clt.iT< 0.1 . + f, .o»prt(;-MM> =71-1
CF-LL MMULiT' 1
C
C TEP:PE**TI>
C
TOOO=TFMP< i.V'
Tui-;- T£r,i- 14 .
IP. 1 014-FHfMr. jo . 70.: . 70i. -70:
".'01 CKLL M^MlE- 11.4.
70c IF i TCC.l-FHFMV. 4 1 > ,704 '^04 .70?
70? Ch" I M;MI?-P 1C. ' ' '
704 IF. .fiGT .EFH . C-OTC 7 'i: _>
IF' -f-H^hf.' 4~ -p..-.t^f»r j,;' ' ^70* «70«: .70^
70^ CHlL M'MOE' ! . .^ '
onm 70?
70^. IF' -f !-"'P«r.' 47 -rt-t«Vf'M. 4£ '7A7 .70:.' .708
7or CHLL M'^GE. :-».4.
70* COr.TINl.'E
c
C CHEC^ FO^ lEFriUHL OUTPUT LDC EhL H ri
C
L - ^ - ' f o »^
. L O.L LCnn ; ncji. -.
. PC ll.C^'i.in £.FMf)CH IF TIME FQP
C001 COmiM-F
c
c IHMG-!MM:- CUMTFDI. or IFII TEMPEPfiTur'E MITH
c
IF' T ri ii n- F ric Hh » ? 1 -. '. 7£' 0 7t 0 72 1
7?u CD7C'4 = 1
L;OTD 7iJ
7JM IF- ' Onn-f-'
7?^ C 137 04 = 0
C-DTO ?0i.i
c
C ."HUrDO'.'f rEi"".'FHfE?
C
r Qf, u- - J
CD"-" '-0
C C'., 1 1 - '.'
r D" 1 - ri
'! 1'iLl. 'i"C?'.'T ' 7 '.' . 0 >
i:nr;,_ -
c D 7 1. * - n
f. Of.-"'- '
Figure 4-16 Task 50 Listing (Continued)
4-48
-------
car !)==. i
core.- = 0
CHLL i-iHOUT< I 0. 1 .0
i:ori.4=ii
GHTO 1 ii£
*oci CHI. i M:HOE ' IT* i ^
VH IT' ;. n 0 .
IFHL1 1 O.E'. -'.'"'.I i'i
CHLL M: Hi:,tt it .4-'
IL1---0
GOTO >:' 9
C
C lEPrtlMF-L CUTPUT LOS ZELTIDM
C
£000 i'PiTE>c » >.*. Tnrii, .TOM yc^.To;. i
UPITF'l-4'> > r. . rnfin.Tni4.TO;'i .TO: I
4 FORMHTc I?. ?LT = .F7.1 FI
c
Ji-'fO CONTINUE
CHLL MI HOE' i -4 '
CHLL MTfii-t ' c »
CHLL rfn.'.'E
c
C RE71HFT POIMT H^Ii O'EO FCP TFttK TESH!MMT10H
r
I 00 IF' IL I 1 Oi .H'4. I (.2
i o; CMLL M :*>?£ 3 .4 .
IL1 = 0
CALL Fhi'-F
104 I F ' I ('M I T -: 0 1 Hi:. :- "' fi 1 06
C
C CONTF-tlLLE(=' :. THVEC H'F'I.IT FTDM B-t THP TEMP 0
C GF'EM VHLV't: ."14- '""'' 'ill'! C1.07E
c
CHI L nrn^E' 4.4
CHLl r-ifillUT' I {)
Figure 4-16 Task 50 Listing (Continued)
4-49
-------
HIS
. c in
EHP
F .'JIFPDi.-f f-M; kFc''i.i;cf
TCI'P
Figure 4-16 Task 50 Listing (Continued)
Tablt 4-4. TASK 10 EXECUTIVE COMMANDS
Coiimand
oi N t
------- |