Digital InlineDigital Inline
Blending with DeltaVBlending with DeltaV
by
Emerson Process Management
@Copyrights
Emerson Process Management Confidential
AgendaAgenda
What is Blending?
How do we do it?
 Blend Ratio Control
 Blend Quality Control & Optimization
Conclusions
Emerson Process Management Confidential
What is BlendingWhat is Blending
Definition – Uniform mixing of 3 or more different streams to produce a
product with specific properties
Problems:
 Flow Constraints
 Inventory Constraints
 Accuracy & Precision
 Poor Mixing (Tank stratification)
Approaches
 Sequential Blending
 Continuous Inline Blending
Emerson Process Management Confidential
History of BlendingHistory of Blending
Prior to 1990
 Most blending is sequential
1990 – 2000
 Most refinery (gasoline & diesel) blenders upgraded to inline blending
 Some chemical blending converted to inline
2000+
 Gasoline blenders forced to consider optimization
 Most remaining sequential blenders upgrade to inline
Emerson Process Management Confidential
Who Uses BlendersWho Uses Blenders
Refineries
 Gasoline
 Diesel
 Distillates, Solvents, Fuel Oils
 Lube Oils
 Asphalt
Chemical Plants
Other
 Baby Formula
The blender is the cash register of the refinery
Emerson Process Management Confidential
Blender’s Operating ProblemsBlender’s Operating Problems
Make more grades with more specifications
using many more components
Make the grade on-spec with minimum
giveaway, but don’t use too much of any one
component
Don’t contaminate any finished product tanks
because of bad line-ups, line fill, leaks, etc.
This is a manually intensive operation but
doing it right is key to the refinery’s
profitability
Increase capacity
Blend Operator
Emerson Process Management Confidential
Improved Blending BenefitsImproved Blending Benefits
Reduced “Giveaway” - Produce closer to specification, reduced cost of
producing average blend
Reduced Off-spec and co-mingled material
Reduced Reblending - increased capacity on blenders
Reduced Component and Product Inventory - reduced working capital
Reduced Blend period – Blend is on-spec from start to finish
Estimated benefits $0.05 to $0.10 per barrel of
gasoline blended
Emerson Process Management Confidential
Sequential Vs. Ratio BlendingSequential Vs. Ratio Blending
One Component at a time
Economical Method (Lower
Installation Cost)
Line Flush required for proper
product quality
Blend Accuracy assured only for
completed batches
Large volume blends will require
long periods of mixing
In-line or all at same time
Requires meter and valve for each
stream
Precise control of blending
Instantaneous alarms on blend
available
Blend quality assured on aborted
blends
Emerson Process Management Confidential
An Integrated Application SolutionAn Integrated Application Solution
Consulting
High Performance Field Instrumentation
 Control Valves
 Flowmeters
On-Line Analyzers
 Conventional Analyzers (GC’s, NIR, etc)
 Neural Network based inferential sensors
DeltaV Controller w/ Proven Configuration & Algorithms
Blend Quality Control, Optimization & Scheduling
Turnkey Skid
Installation, Start-up & Training
Financing
Emerson Process Management Confidential
FC
Component
Ratio Control
FI AI
Blend Quality
Controller
On-Line Blend
Optimization
Ratio Targets
Component Ratio
& Quality Targets
Component Qualities
& Limits
Desired
Rundown Rate
& Flow Limits
Blend
Planning LP
Prices
Batches
Target Recipes
Target Inventories
Heel Information
Component Quality &
Availability
Blend
Start/Stop
FC
FC
FC
DeltaV Blend
DeltaV Blend Control StrategyDeltaV Blend Control Strategy
Emerson Process Management Confidential
Modular DesignModular Design
Continuous or Batch (Tank) In-Line Digital Blending
Recipe Control
 30 recipes with 80 configurable parameters each
 Start/Hold/Resume/Complete logic
 Blend Reports
Component & Additive Ratio Control
 Pacing (Instantaneous ratio control)
 Memory (Total ratio control)
 Ramp up/down
Dual Analyzer Trim Control
 Decoupling
Emerson Process Management Confidential
OptionsOptions
Basic Customization
 Add/Delete streams
 Custom graphics
 Customer Tags
 Define I/O
 Add/Delete recipes/parameters
Optimization
 Low level optimization in Analyzer control strategies
 High level optimization integration (OPC)
Master Blend Control
 Coordination of multiple blend headers
 Integration of header and pipeline control (OM&S)
Emerson Process Management Confidential
OptionsOptions
Simulation
 Training
 Testing
Custom Interfaces (analyzers, optimizers, etc)
Special Calculations & Functions
 Special equipment scheduling and blend line-ups based on production rates,
equipment availability, etc
 Derived Quality Calculations
 Blend grade transition and Tank heel accounting
 Tank Farm Management
 Component, additive & product compatibility checking
 Scheduling and Time-to-completion calculations
Emerson Process Management Confidential
Standard DeltaV Blender AlgorithmsStandard DeltaV Blender Algorithms
Blend Ratio Control
 Streams
12 main component flows
4 additive flows
 2 Analyzer trim controllers with decoupling (Optional)
Start/Stop Control
 Sequential blender pump and valve line-ups
1 pump, 1 control valve, 1 flow transmitter per stream
Delay Sequences, ramping up/down
 Analog & Discrete controls and monitoring of auxiliary equipment
Feed and Product tank level Monitoring
Header pressure & temperature control
Emerson Process Management Confidential
Blend Start/Stop ControlBlend Start/Stop Control
Startup Sequencing
 Load Blend Order to on-line system
 Log starting conditions to Blend Order DB
 Flush lines
 Zero totalizers
 Select, Line-up, & Start pumps
 Ramp valves open and put FC’s into control
 Start ratio control, ramp to target rate
Shutdown Sequencing
 Ramp down blend rate to minimum
 Place FC’s in manual and close valves
 Shutdown pumps
 Log final blend results to Blend Order DB
Emerson Process Management Confidential
Operator InterfaceOperator Interface
Primary Operating Displays
 Blend Setup
 Blend Overview Graphic
 Component & Additive loops
 Analyzer Trim Control
Interlock/Permissive Status
Displays
 Active/Bypassed
Recipe management
Displays
–Load recipes
–Modify recipes
–Save recipes
Support Displays
–Trends
–Alarm management
–Reports
Emerson Process Management Confidential
Primary User ViewsPrimary User Views
Blend Control
Recipe Management Blend Setup
Blend Components
Emerson Process Management Confidential
FC
Component
Ratio Control
FI AI
Blend Quality
Controller
On-Line Blend
Optimization
Ratio Targets
Component Ratio
& Quality Targets
Component Qualities
& Limits
Desired
Rundown Rate
& Flow Limits
Blend
Planning LP
Prices
Batches
Target Recipes
Target Inventories
Heel Information
Component Quality &
Availability
Blend
Start/Stop
FC
FC
FC
DeltaV Blend
Blend Quality Control & OptimizationBlend Quality Control & Optimization
Emerson Process Management Confidential
US Gasoline Regulatory Time LineUS Gasoline Regulatory Time Line
20052004200320022001200019991998199719961995
CARB
CBG
EPA RFG
Phase ll
EPA Low
Sulfur Gasoline
EPA RFG
Complex Model
CARB MTBE
Phase Out
EPA RFG
Simple Model
Each New Regulation Means More Complicated Blending
199419931992
Mandated
Oxygenates
Emerson Process Management Confidential
Complicated Gasoline SpecificationsComplicated Gasoline Specifications
Parameter Current California
Specification
Octane Per Grade
RVP 7.0 psi max
Sulfur 80 ppm max
30 ppm av
T50 220 F max
170 F min
200 F av
T90 330 F max
290 F av
Olefins 10 vol % max
4 vol % av
Aromatics 30 vol % max
22 vol % av
Oxygen 2.7 wt % max
1.8 wt % min
To meet these
specifications is a
complicated
multivariable control
problem
Emerson Process Management Confidential
Longer term
coordination
Production
Planning
Medium term
definition of activities
Production
Scheduling
Blend
Control
Real-time regulatory
control & optimization
Initial ConditionsPlanned production and/or
material and resource constraints
Current Status and PerformanceTargets and recipes
Interactions between Planning,Interactions between Planning,
Scheduling, and ControlScheduling, and Control
Emerson Process Management Confidential
So why isn’t recipe control enough for
blending?
Production PlanningProduction Planning
Linear Program (LP)
Run monthly (weekly?) with average product demand
Selects crudes, intermediates, and product purchases for future period
(one to three months) to meet demand at maximum profit
Sets average blend recipes for period (assumed component quantities and
qualities)
Emerson Process Management Confidential
ProblemsProblems
Simplified LP models
Crude assay database
Component properties are different than that assumed by Planning LP
Crudes are run individually or in blends different from average resulting from
the Planning LP
Specific batches of gasoline have to be produced to a specific grade with the
stocks on hand
How can we help solve these
problems?
Emerson Process Management Confidential
Multi-Period Blend Planning LPMulti-Period Blend Planning LP
Multi-period, multi-blend, executed daily or weekly
Linearized blend models (recursive LP)
Determines optimum blends required to meet specific batches of product demands
Integrated with refinery planning LP
Inputs:
 Product shipments, prices
 Starting inventories
 Projected component rates & qualities
 Target ending inventories
Outputs:
 Finished product batches
 Target recipes for each batch
 Target ending inventories for each period
(products and components)
Emerson Process Management Confidential
Traditional Blend Control ArchitectureTraditional Blend Control Architecture
Blend Models
AI
FC FC FC FC FC
AI
MPC Dynamic
Models
Ratio Controls
(Ramping & Pacing)
Flow Targets
Blend Property
Control
Ratio Targets
Blend Scheduling
Target Recipe & Availability
Blend Models
Blend Optimization
Product Quality Targets
Component Flows
Emerson Process Management Confidential
Blend Order Management System
Blend Order ManagementBlend Order Management
Blend Order Database
Blend
Start/Stop
Blend Order
Execution
MMI
Blend
Reports
Blend Reporting
Blend
Optimizer
Emerson Process Management Confidential
Blend Order ManagementBlend Order Management
User Interface to Blend Order Database
Security functions
 User authority changes with order status
Order Entry
 Load from past orders
Order Management
 Tracks orders through phases
 Confirms required data input before changing phases
 Approvals
 Data validation
Blend Reporting
Emerson Process Management Confidential
Blend Order Data ObjectBlend Order Data Object
Blend Order
• Blend Order ID
• Blended Product Name
• Blender ID
• Order Status
• Planned start/end time
• Target Quantity
• Product Quality Targets & Limits (up to 25 qualities)
• Components and target quantities (up to 15 components)
• Additives and target quantities (up to 10 additives)
• Special instructions
• Actual time for each status change
• Final blend results
• Operator comments
Emerson Process Management Confidential
Blend Order PhasesBlend Order Phases
Planned
Approved
Selected
Starting
Active
Stopping
Complete
Hold
Blend Planner in Control
Operator in Control
Emerson Process Management Confidential
Offline Blend OptimizationOffline Blend Optimization
File based
Load inputs from stored blend orders
Alternative objective functions
Correct for heel
User can modify all inputs
Inputs:
 Target recipe
 Desired finished batch size
 Finished product quality specs
 Component qualities and availability
 Heel quantity and qualities
Outputs:
 Optimal blend qualities
 Optimal recipe
 Final batch size
Emerson Process Management Confidential
Blend OptimizationBlend Optimization
• Optimum recipe
• Component usage
• Predicted qualities
Results
RTO+
Real-Time
Optimization Engine
Inputs
• Prices
• Target Recipe
• Product Specs
• Component
Availability
• Current Heel
Quantity & Quality
Objective Function
• Maximize Profit
• Minimize Deviation
from Target Recipe
• Min/Max use of
Specific Component
Blend Models
Emerson Process Management Confidential
• Total Blended
Amount
• Predicted Density
• Predicted Qualities
Outputs
Blend ModelsBlend Models
• Component Name
• Amount to be
blended
• Density
• Qualities
• Component Name
• Amount to be
blended
• Density
• Qualities
Components
• Component Name
• Amount to be
blended
• Density
• Qualities
Heel
• Product Name
• Amount
• Density
• Qualities
Blend Models
Blend Quality Library
• Density
• Sulfur
• Octane
• ASTM Distillation
• Aromatics
• Olefins
• Benzene
• RVP...
Emerson Process Management Confidential
Inputs
• Prices
• Target Recipe
• Product Specs
• Component
Availability
• Current Heel
Quantity & Quality
RTO+
Real-Time Optimization Engine
Objective Function
• Maximize Profit
• Minimize Deviation
from Target Recipe
• Min/Max use of
Specific Component
Blend Models
• Optimum recipe
• Component usage
• Predicted qualities
Results
Off-Line
Optimizer
Multiple Optimizer ApplicationsMultiple Optimizer Applications
Real-Time Database
Inputs
• Prices
• Target Recipe
• Product Specs
• Component
Availability
• Current Heel
Quantity & Quality
RTO+
Real-Time Optimization Engine
Objective Function
• Maximize Profit
• Minimize Deviation
from Target Recipe
• Min/Max use of
Specific Component
Blend Models
• Optimum recipe
• Component usage
• Predicted qualities
Results
On-Line
Optimizer
Blend Order Database
Inputs
• Prices
• Target Recipe
• Product Specs
• Component
Availability
• Current Heel
Quantity & Quality
RTO+
Real-Time Optimization Engine
Objective Function
• Maximize Profit
• Minimize Deviation
from Target Recipe
• Min/Max use of
Specific Component
Blend Models
• Optimum recipe
• Component usage
• Predicted qualities
Results
Blend Quality
Control
Emerson Process Management Confidential
Online Blend OptimizationOnline Blend Optimization
Executes every 5-15 minutes
Runs for blend orders that are “Active”
Operator selects objective function
Outputs recommended results even in open-loop
Handles reduced degrees of freedom (e.g. 1 or more controllers in Auto)
Accommodates actual component qualities and blend model update from
analyzers or lab
Inputs:
 Automatically loaded from RT database and blend order system
Outputs:
 Optimal instantaneous quality targets
 Optimal recipe
 Target batch size
Emerson Process Management Confidential
Components
Product
Tankage
Total
Blend
On-Spec
Heel
Components
Product
In-Line
Each
Part of
Blend
On-Spec
Online Analyzers
Blend Property Control
Online Blend Optimization
Product Certification
Two Basic Types of BlendingTwo Basic Types of Blending
Emerson Process Management Confidential
Accelerated Heel CorrectionAccelerated Heel Correction
Final
Target
Percent of Blend
Octane
0% 100%
Heel
Current
Tank Quality
Blend header
target adjusted
to meet
final target
Current Blend
Header Target
Emerson Process Management Confidential
Blend ExecutionBlend Execution
FC
Component
Ratio Control
FI AI
Blend Quality
Controller
Ratio Targets
Component Qualities
& Limits
Desired Rundown Rate,
Header Pressure, Flow and
Valve Limits
FC
FC
FC
PI
Product Quality
Targets
Emerson Process Management Confidential
Blend Quality ControlBlend Quality Control
RTO+
Real-Time
Optimization Engine
Blend Models
• Executes 1-2 minutes
• Specifications can be:
- Target
- Range
- Max or min
Real-TimeDatabase
• Current Blend Quality Targets & Specs
• Current Component Ratios
• Blend Quality Analyzers
• Component Limits & Qualities
• Component ratio targets
• Predicted qualities
Objective Function
• Primary: Meet Quality Specs
• Secondary: Minimize Deviation from
Current Recipe
Emerson Process Management Confidential
SummarySummary
Pre-engineered proven application
 Reduced risk
Leading Technology hardware platform
 Non-proprietary
 Longest Life Expectancy
Easily customized and extended
 Integrate virtually any feature or function as required
Complete Engineering support from start to finish
 Consultation
 Main Automation Contractor
The Global Industry Leader
Emerson Process Management Confidential
Technical AdvantagesTechnical Advantages
Integrated suite of applications on DeltaV
Scaleable, Modular
Automated Startup/Shutdown
No need for multivariable control layer
 Dynamics are trivial
 No step tests required
 Incorporates non-linearity
 Simplifies architecture
Integrated set of blend models used by all applications
 Non-linear
 Biased by lab or analyzer data
 Incorporate RFG simple and complex models
Integrated blend order management & reports
Emerson Process Management Confidential
Moving ForwardMoving Forward
Complete Blend Optimization Modules
Parallel Effort to Develop Movements & Inventory Solution
 Leverage Terminal Automation Solution

More Related Content

PPT
Blending with deltav
PPT
Gasoline blending
PDF
Advanced Process Control for ARC
PPT
In Apc Training Presentation
PPTX
Advanced Process Control
PPT
Practical Advanced Process Control for Engineers and Technicians
PPT
Recovery from a process saturation condition benefits of using delta v pid_plus
PPTX
Control Process Design parameter
Blending with deltav
Gasoline blending
Advanced Process Control for ARC
In Apc Training Presentation
Advanced Process Control
Practical Advanced Process Control for Engineers and Technicians
Recovery from a process saturation condition benefits of using delta v pid_plus
Control Process Design parameter

What's hot (14)

PPTX
Solution Manager Technical Monitoring - SAP NW BW on HANA Setup Part 2 of 3 (...
PPTX
Process Design and control
PDF
(ATS6-PLAT08) AEP in a Validated Environment
PDF
IPC_project final
PDF
AFT Fathom 11 - Ficha de dados
PPT
Basics of Automation in Solid Dosage Form Production (Formulation & Developem...
PDF
Solution Manager Technical Monitoring - BOBJ (Part 2 of 3 - Auto-Configure)
PPTX
Process Control
PPT
07 Line Balancing
PPT
Watkins q1 2nded_outlook_on_quality_final
PPTX
Pharmaceutical automation
PPSX
Control valve sizing, selection & quotation software
PPTX
Administration settings
PPTX
An Integrated Supply Chains Solution for Cold Chains
Solution Manager Technical Monitoring - SAP NW BW on HANA Setup Part 2 of 3 (...
Process Design and control
(ATS6-PLAT08) AEP in a Validated Environment
IPC_project final
AFT Fathom 11 - Ficha de dados
Basics of Automation in Solid Dosage Form Production (Formulation & Developem...
Solution Manager Technical Monitoring - BOBJ (Part 2 of 3 - Auto-Configure)
Process Control
07 Line Balancing
Watkins q1 2nded_outlook_on_quality_final
Pharmaceutical automation
Control valve sizing, selection & quotation software
Administration settings
An Integrated Supply Chains Solution for Cold Chains
Ad

Viewers also liked (20)

PPT
Control System Cyber Security - A Different Approach
PDF
01 sable
PDF
MHM Smart Machinery Technologies
PPT
Safety system
PPT
Safety Instrumentation
PPT
Advances In Digital Automation Within Refining
PPT
Introduction to Functional Safety and SIL Certification
PDF
80 instrumentation interview questions with answers
PDF
Prezentacja B2B - SuperPrezenty
PPT
DR Planning - Improving Recovery Time
PPT
Session 5 financing strategies
PDF
Ympäristösitoumus 2018, Green Tourism of Finland GTF®
PPT
Elevaktiv formativ bedömning ppt
DOC
Prova doc 3-1
PPS
Blessed to Bless
PPTX
DAKTARI Newsletter Q1 2015
PDF
impact of viewer perception on effectiveness of couple image sex appeal adver...
DOCX
Gifts of the_spirit
PPT
Interactive RFID Credit Card
PPTX
Perilaku Nonverbal by Sahid
Control System Cyber Security - A Different Approach
01 sable
MHM Smart Machinery Technologies
Safety system
Safety Instrumentation
Advances In Digital Automation Within Refining
Introduction to Functional Safety and SIL Certification
80 instrumentation interview questions with answers
Prezentacja B2B - SuperPrezenty
DR Planning - Improving Recovery Time
Session 5 financing strategies
Ympäristösitoumus 2018, Green Tourism of Finland GTF®
Elevaktiv formativ bedömning ppt
Prova doc 3-1
Blessed to Bless
DAKTARI Newsletter Q1 2015
impact of viewer perception on effectiveness of couple image sex appeal adver...
Gifts of the_spirit
Interactive RFID Credit Card
Perilaku Nonverbal by Sahid
Ad

Similar to Blendingwith DeltaV (20)

PDF
Process house automation
PPTX
Bunker Blending Presentation 1
PPTX
Process Control in Chemical Engineering by MATLAB
PPT
Smart Process Distillation Application Improves Recovery And Saves Energy
PPT
Manufacturing Execution & Information Management System
PPTX
Engineering
PPTX
Architectural styles 3
PPT
PID Tuning for Self Regulating Processes - Greg McMillan Deminar
PPTX
Adding new elements in old Control Loops
PPTX
Use Cases for Big Data and the Connected Enterprise
PPT
Gasoline blending
PDF
Risk management in complex changeovers through cfmea
PPTX
process control system
PPT
How to Setup and Adjust the Dynamic Compensation of Feedforward Signals
PPT
FMEA Presentation V1.1
PPT
Proportional Integrator Der Training.ppt
PDF
Sistema de control online de calidad de pellets de madera
PPTX
BatchMaster ERP for Food & Beverages
PPTX
Condition Based Asset Management R K Gupta
PPTX
Process Control Fundamentals and How to read P&IDs
Process house automation
Bunker Blending Presentation 1
Process Control in Chemical Engineering by MATLAB
Smart Process Distillation Application Improves Recovery And Saves Energy
Manufacturing Execution & Information Management System
Engineering
Architectural styles 3
PID Tuning for Self Regulating Processes - Greg McMillan Deminar
Adding new elements in old Control Loops
Use Cases for Big Data and the Connected Enterprise
Gasoline blending
Risk management in complex changeovers through cfmea
process control system
How to Setup and Adjust the Dynamic Compensation of Feedforward Signals
FMEA Presentation V1.1
Proportional Integrator Der Training.ppt
Sistema de control online de calidad de pellets de madera
BatchMaster ERP for Food & Beverages
Condition Based Asset Management R K Gupta
Process Control Fundamentals and How to read P&IDs

More from Luis Atencio (8)

PPTX
Safety life cycle seminar IEC61511
PDF
HUman centred design applied to Graphics (Emerson, DeltaV)
PDF
Unified industrial wireless networks (cisco)
PPT
Emerson Wireless Presentacion a Clientes
PPT
Wireless hart conceptos basicos
PPT
Delta v advanced control overview_en
PPTX
Electronic marshalling vs remote io
PPT
DeltaV overview v11
Safety life cycle seminar IEC61511
HUman centred design applied to Graphics (Emerson, DeltaV)
Unified industrial wireless networks (cisco)
Emerson Wireless Presentacion a Clientes
Wireless hart conceptos basicos
Delta v advanced control overview_en
Electronic marshalling vs remote io
DeltaV overview v11

Recently uploaded (20)

PDF
IT-ITes Industry bjjbnkmkhkhknbmhkhmjhjkhj
PDF
NewMind AI Weekly Chronicles – August ’25 Week IV
PDF
MENA-ECEONOMIC-CONTEXT-VC MENA-ECEONOMIC
PPTX
Configure Apache Mutual Authentication
PDF
Accessing-Finance-in-Jordan-MENA 2024 2025.pdf
PDF
The-Future-of-Automotive-Quality-is-Here-AI-Driven-Engineering.pdf
PPTX
SGT Report The Beast Plan and Cyberphysical Systems of Control
PDF
Aug23rd - Mulesoft Community Workshop - Hyd, India.pdf
PDF
Co-training pseudo-labeling for text classification with support vector machi...
PDF
Comparative analysis of machine learning models for fake news detection in so...
PDF
CXOs-Are-you-still-doing-manual-DevOps-in-the-age-of-AI.pdf
PDF
4 layer Arch & Reference Arch of IoT.pdf
PPT
Galois Field Theory of Risk: A Perspective, Protocol, and Mathematical Backgr...
PDF
SaaS reusability assessment using machine learning techniques
PDF
INTERSPEECH 2025 「Recent Advances and Future Directions in Voice Conversion」
PDF
Rapid Prototyping: A lecture on prototyping techniques for interface design
PDF
“A New Era of 3D Sensing: Transforming Industries and Creating Opportunities,...
PDF
Dell Pro Micro: Speed customer interactions, patient processing, and learning...
PDF
sbt 2.0: go big (Scala Days 2025 edition)
PDF
Transform-Your-Factory-with-AI-Driven-Quality-Engineering.pdf
IT-ITes Industry bjjbnkmkhkhknbmhkhmjhjkhj
NewMind AI Weekly Chronicles – August ’25 Week IV
MENA-ECEONOMIC-CONTEXT-VC MENA-ECEONOMIC
Configure Apache Mutual Authentication
Accessing-Finance-in-Jordan-MENA 2024 2025.pdf
The-Future-of-Automotive-Quality-is-Here-AI-Driven-Engineering.pdf
SGT Report The Beast Plan and Cyberphysical Systems of Control
Aug23rd - Mulesoft Community Workshop - Hyd, India.pdf
Co-training pseudo-labeling for text classification with support vector machi...
Comparative analysis of machine learning models for fake news detection in so...
CXOs-Are-you-still-doing-manual-DevOps-in-the-age-of-AI.pdf
4 layer Arch & Reference Arch of IoT.pdf
Galois Field Theory of Risk: A Perspective, Protocol, and Mathematical Backgr...
SaaS reusability assessment using machine learning techniques
INTERSPEECH 2025 「Recent Advances and Future Directions in Voice Conversion」
Rapid Prototyping: A lecture on prototyping techniques for interface design
“A New Era of 3D Sensing: Transforming Industries and Creating Opportunities,...
Dell Pro Micro: Speed customer interactions, patient processing, and learning...
sbt 2.0: go big (Scala Days 2025 edition)
Transform-Your-Factory-with-AI-Driven-Quality-Engineering.pdf

Blendingwith DeltaV

  • 1. Digital InlineDigital Inline Blending with DeltaVBlending with DeltaV by Emerson Process Management @Copyrights
  • 2. Emerson Process Management Confidential AgendaAgenda What is Blending? How do we do it?  Blend Ratio Control  Blend Quality Control & Optimization Conclusions
  • 3. Emerson Process Management Confidential What is BlendingWhat is Blending Definition – Uniform mixing of 3 or more different streams to produce a product with specific properties Problems:  Flow Constraints  Inventory Constraints  Accuracy & Precision  Poor Mixing (Tank stratification) Approaches  Sequential Blending  Continuous Inline Blending
  • 4. Emerson Process Management Confidential History of BlendingHistory of Blending Prior to 1990  Most blending is sequential 1990 – 2000  Most refinery (gasoline & diesel) blenders upgraded to inline blending  Some chemical blending converted to inline 2000+  Gasoline blenders forced to consider optimization  Most remaining sequential blenders upgrade to inline
  • 5. Emerson Process Management Confidential Who Uses BlendersWho Uses Blenders Refineries  Gasoline  Diesel  Distillates, Solvents, Fuel Oils  Lube Oils  Asphalt Chemical Plants Other  Baby Formula The blender is the cash register of the refinery
  • 6. Emerson Process Management Confidential Blender’s Operating ProblemsBlender’s Operating Problems Make more grades with more specifications using many more components Make the grade on-spec with minimum giveaway, but don’t use too much of any one component Don’t contaminate any finished product tanks because of bad line-ups, line fill, leaks, etc. This is a manually intensive operation but doing it right is key to the refinery’s profitability Increase capacity Blend Operator
  • 7. Emerson Process Management Confidential Improved Blending BenefitsImproved Blending Benefits Reduced “Giveaway” - Produce closer to specification, reduced cost of producing average blend Reduced Off-spec and co-mingled material Reduced Reblending - increased capacity on blenders Reduced Component and Product Inventory - reduced working capital Reduced Blend period – Blend is on-spec from start to finish Estimated benefits $0.05 to $0.10 per barrel of gasoline blended
  • 8. Emerson Process Management Confidential Sequential Vs. Ratio BlendingSequential Vs. Ratio Blending One Component at a time Economical Method (Lower Installation Cost) Line Flush required for proper product quality Blend Accuracy assured only for completed batches Large volume blends will require long periods of mixing In-line or all at same time Requires meter and valve for each stream Precise control of blending Instantaneous alarms on blend available Blend quality assured on aborted blends
  • 9. Emerson Process Management Confidential An Integrated Application SolutionAn Integrated Application Solution Consulting High Performance Field Instrumentation  Control Valves  Flowmeters On-Line Analyzers  Conventional Analyzers (GC’s, NIR, etc)  Neural Network based inferential sensors DeltaV Controller w/ Proven Configuration & Algorithms Blend Quality Control, Optimization & Scheduling Turnkey Skid Installation, Start-up & Training Financing
  • 10. Emerson Process Management Confidential FC Component Ratio Control FI AI Blend Quality Controller On-Line Blend Optimization Ratio Targets Component Ratio & Quality Targets Component Qualities & Limits Desired Rundown Rate & Flow Limits Blend Planning LP Prices Batches Target Recipes Target Inventories Heel Information Component Quality & Availability Blend Start/Stop FC FC FC DeltaV Blend DeltaV Blend Control StrategyDeltaV Blend Control Strategy
  • 11. Emerson Process Management Confidential Modular DesignModular Design Continuous or Batch (Tank) In-Line Digital Blending Recipe Control  30 recipes with 80 configurable parameters each  Start/Hold/Resume/Complete logic  Blend Reports Component & Additive Ratio Control  Pacing (Instantaneous ratio control)  Memory (Total ratio control)  Ramp up/down Dual Analyzer Trim Control  Decoupling
  • 12. Emerson Process Management Confidential OptionsOptions Basic Customization  Add/Delete streams  Custom graphics  Customer Tags  Define I/O  Add/Delete recipes/parameters Optimization  Low level optimization in Analyzer control strategies  High level optimization integration (OPC) Master Blend Control  Coordination of multiple blend headers  Integration of header and pipeline control (OM&S)
  • 13. Emerson Process Management Confidential OptionsOptions Simulation  Training  Testing Custom Interfaces (analyzers, optimizers, etc) Special Calculations & Functions  Special equipment scheduling and blend line-ups based on production rates, equipment availability, etc  Derived Quality Calculations  Blend grade transition and Tank heel accounting  Tank Farm Management  Component, additive & product compatibility checking  Scheduling and Time-to-completion calculations
  • 14. Emerson Process Management Confidential Standard DeltaV Blender AlgorithmsStandard DeltaV Blender Algorithms Blend Ratio Control  Streams 12 main component flows 4 additive flows  2 Analyzer trim controllers with decoupling (Optional) Start/Stop Control  Sequential blender pump and valve line-ups 1 pump, 1 control valve, 1 flow transmitter per stream Delay Sequences, ramping up/down  Analog & Discrete controls and monitoring of auxiliary equipment Feed and Product tank level Monitoring Header pressure & temperature control
  • 15. Emerson Process Management Confidential Blend Start/Stop ControlBlend Start/Stop Control Startup Sequencing  Load Blend Order to on-line system  Log starting conditions to Blend Order DB  Flush lines  Zero totalizers  Select, Line-up, & Start pumps  Ramp valves open and put FC’s into control  Start ratio control, ramp to target rate Shutdown Sequencing  Ramp down blend rate to minimum  Place FC’s in manual and close valves  Shutdown pumps  Log final blend results to Blend Order DB
  • 16. Emerson Process Management Confidential Operator InterfaceOperator Interface Primary Operating Displays  Blend Setup  Blend Overview Graphic  Component & Additive loops  Analyzer Trim Control Interlock/Permissive Status Displays  Active/Bypassed Recipe management Displays –Load recipes –Modify recipes –Save recipes Support Displays –Trends –Alarm management –Reports
  • 17. Emerson Process Management Confidential Primary User ViewsPrimary User Views Blend Control Recipe Management Blend Setup Blend Components
  • 18. Emerson Process Management Confidential FC Component Ratio Control FI AI Blend Quality Controller On-Line Blend Optimization Ratio Targets Component Ratio & Quality Targets Component Qualities & Limits Desired Rundown Rate & Flow Limits Blend Planning LP Prices Batches Target Recipes Target Inventories Heel Information Component Quality & Availability Blend Start/Stop FC FC FC DeltaV Blend Blend Quality Control & OptimizationBlend Quality Control & Optimization
  • 19. Emerson Process Management Confidential US Gasoline Regulatory Time LineUS Gasoline Regulatory Time Line 20052004200320022001200019991998199719961995 CARB CBG EPA RFG Phase ll EPA Low Sulfur Gasoline EPA RFG Complex Model CARB MTBE Phase Out EPA RFG Simple Model Each New Regulation Means More Complicated Blending 199419931992 Mandated Oxygenates
  • 20. Emerson Process Management Confidential Complicated Gasoline SpecificationsComplicated Gasoline Specifications Parameter Current California Specification Octane Per Grade RVP 7.0 psi max Sulfur 80 ppm max 30 ppm av T50 220 F max 170 F min 200 F av T90 330 F max 290 F av Olefins 10 vol % max 4 vol % av Aromatics 30 vol % max 22 vol % av Oxygen 2.7 wt % max 1.8 wt % min To meet these specifications is a complicated multivariable control problem
  • 21. Emerson Process Management Confidential Longer term coordination Production Planning Medium term definition of activities Production Scheduling Blend Control Real-time regulatory control & optimization Initial ConditionsPlanned production and/or material and resource constraints Current Status and PerformanceTargets and recipes Interactions between Planning,Interactions between Planning, Scheduling, and ControlScheduling, and Control
  • 22. Emerson Process Management Confidential So why isn’t recipe control enough for blending? Production PlanningProduction Planning Linear Program (LP) Run monthly (weekly?) with average product demand Selects crudes, intermediates, and product purchases for future period (one to three months) to meet demand at maximum profit Sets average blend recipes for period (assumed component quantities and qualities)
  • 23. Emerson Process Management Confidential ProblemsProblems Simplified LP models Crude assay database Component properties are different than that assumed by Planning LP Crudes are run individually or in blends different from average resulting from the Planning LP Specific batches of gasoline have to be produced to a specific grade with the stocks on hand How can we help solve these problems?
  • 24. Emerson Process Management Confidential Multi-Period Blend Planning LPMulti-Period Blend Planning LP Multi-period, multi-blend, executed daily or weekly Linearized blend models (recursive LP) Determines optimum blends required to meet specific batches of product demands Integrated with refinery planning LP Inputs:  Product shipments, prices  Starting inventories  Projected component rates & qualities  Target ending inventories Outputs:  Finished product batches  Target recipes for each batch  Target ending inventories for each period (products and components)
  • 25. Emerson Process Management Confidential Traditional Blend Control ArchitectureTraditional Blend Control Architecture Blend Models AI FC FC FC FC FC AI MPC Dynamic Models Ratio Controls (Ramping & Pacing) Flow Targets Blend Property Control Ratio Targets Blend Scheduling Target Recipe & Availability Blend Models Blend Optimization Product Quality Targets Component Flows
  • 26. Emerson Process Management Confidential Blend Order Management System Blend Order ManagementBlend Order Management Blend Order Database Blend Start/Stop Blend Order Execution MMI Blend Reports Blend Reporting Blend Optimizer
  • 27. Emerson Process Management Confidential Blend Order ManagementBlend Order Management User Interface to Blend Order Database Security functions  User authority changes with order status Order Entry  Load from past orders Order Management  Tracks orders through phases  Confirms required data input before changing phases  Approvals  Data validation Blend Reporting
  • 28. Emerson Process Management Confidential Blend Order Data ObjectBlend Order Data Object Blend Order • Blend Order ID • Blended Product Name • Blender ID • Order Status • Planned start/end time • Target Quantity • Product Quality Targets & Limits (up to 25 qualities) • Components and target quantities (up to 15 components) • Additives and target quantities (up to 10 additives) • Special instructions • Actual time for each status change • Final blend results • Operator comments
  • 29. Emerson Process Management Confidential Blend Order PhasesBlend Order Phases Planned Approved Selected Starting Active Stopping Complete Hold Blend Planner in Control Operator in Control
  • 30. Emerson Process Management Confidential Offline Blend OptimizationOffline Blend Optimization File based Load inputs from stored blend orders Alternative objective functions Correct for heel User can modify all inputs Inputs:  Target recipe  Desired finished batch size  Finished product quality specs  Component qualities and availability  Heel quantity and qualities Outputs:  Optimal blend qualities  Optimal recipe  Final batch size
  • 31. Emerson Process Management Confidential Blend OptimizationBlend Optimization • Optimum recipe • Component usage • Predicted qualities Results RTO+ Real-Time Optimization Engine Inputs • Prices • Target Recipe • Product Specs • Component Availability • Current Heel Quantity & Quality Objective Function • Maximize Profit • Minimize Deviation from Target Recipe • Min/Max use of Specific Component Blend Models
  • 32. Emerson Process Management Confidential • Total Blended Amount • Predicted Density • Predicted Qualities Outputs Blend ModelsBlend Models • Component Name • Amount to be blended • Density • Qualities • Component Name • Amount to be blended • Density • Qualities Components • Component Name • Amount to be blended • Density • Qualities Heel • Product Name • Amount • Density • Qualities Blend Models Blend Quality Library • Density • Sulfur • Octane • ASTM Distillation • Aromatics • Olefins • Benzene • RVP...
  • 33. Emerson Process Management Confidential Inputs • Prices • Target Recipe • Product Specs • Component Availability • Current Heel Quantity & Quality RTO+ Real-Time Optimization Engine Objective Function • Maximize Profit • Minimize Deviation from Target Recipe • Min/Max use of Specific Component Blend Models • Optimum recipe • Component usage • Predicted qualities Results Off-Line Optimizer Multiple Optimizer ApplicationsMultiple Optimizer Applications Real-Time Database Inputs • Prices • Target Recipe • Product Specs • Component Availability • Current Heel Quantity & Quality RTO+ Real-Time Optimization Engine Objective Function • Maximize Profit • Minimize Deviation from Target Recipe • Min/Max use of Specific Component Blend Models • Optimum recipe • Component usage • Predicted qualities Results On-Line Optimizer Blend Order Database Inputs • Prices • Target Recipe • Product Specs • Component Availability • Current Heel Quantity & Quality RTO+ Real-Time Optimization Engine Objective Function • Maximize Profit • Minimize Deviation from Target Recipe • Min/Max use of Specific Component Blend Models • Optimum recipe • Component usage • Predicted qualities Results Blend Quality Control
  • 34. Emerson Process Management Confidential Online Blend OptimizationOnline Blend Optimization Executes every 5-15 minutes Runs for blend orders that are “Active” Operator selects objective function Outputs recommended results even in open-loop Handles reduced degrees of freedom (e.g. 1 or more controllers in Auto) Accommodates actual component qualities and blend model update from analyzers or lab Inputs:  Automatically loaded from RT database and blend order system Outputs:  Optimal instantaneous quality targets  Optimal recipe  Target batch size
  • 35. Emerson Process Management Confidential Components Product Tankage Total Blend On-Spec Heel Components Product In-Line Each Part of Blend On-Spec Online Analyzers Blend Property Control Online Blend Optimization Product Certification Two Basic Types of BlendingTwo Basic Types of Blending
  • 36. Emerson Process Management Confidential Accelerated Heel CorrectionAccelerated Heel Correction Final Target Percent of Blend Octane 0% 100% Heel Current Tank Quality Blend header target adjusted to meet final target Current Blend Header Target
  • 37. Emerson Process Management Confidential Blend ExecutionBlend Execution FC Component Ratio Control FI AI Blend Quality Controller Ratio Targets Component Qualities & Limits Desired Rundown Rate, Header Pressure, Flow and Valve Limits FC FC FC PI Product Quality Targets
  • 38. Emerson Process Management Confidential Blend Quality ControlBlend Quality Control RTO+ Real-Time Optimization Engine Blend Models • Executes 1-2 minutes • Specifications can be: - Target - Range - Max or min Real-TimeDatabase • Current Blend Quality Targets & Specs • Current Component Ratios • Blend Quality Analyzers • Component Limits & Qualities • Component ratio targets • Predicted qualities Objective Function • Primary: Meet Quality Specs • Secondary: Minimize Deviation from Current Recipe
  • 39. Emerson Process Management Confidential SummarySummary Pre-engineered proven application  Reduced risk Leading Technology hardware platform  Non-proprietary  Longest Life Expectancy Easily customized and extended  Integrate virtually any feature or function as required Complete Engineering support from start to finish  Consultation  Main Automation Contractor The Global Industry Leader
  • 40. Emerson Process Management Confidential Technical AdvantagesTechnical Advantages Integrated suite of applications on DeltaV Scaleable, Modular Automated Startup/Shutdown No need for multivariable control layer  Dynamics are trivial  No step tests required  Incorporates non-linearity  Simplifies architecture Integrated set of blend models used by all applications  Non-linear  Biased by lab or analyzer data  Incorporate RFG simple and complex models Integrated blend order management & reports
  • 41. Emerson Process Management Confidential Moving ForwardMoving Forward Complete Blend Optimization Modules Parallel Effort to Develop Movements & Inventory Solution  Leverage Terminal Automation Solution

Editor's Notes

  • #6: Virtually every liquid refinery product is blended to spec prior to sale. More attention has been given to the “ProcessUnits”, but poor blending can eat away at margins as fast as anything. We have a solution to help you reduce reblending and giveaways.
  • #9: Most blending is performed in-line today. It avoids the need for extra tankage and mixing and generally provides a more uniform product when dealing with large quantities characteristic of most refineries. In-line blending requires proper instrumentation and controls.
  • #10: Our Digital Blending Package is an application solution which incorporates many potential products and services from Emerson/Fisher-Rosemount. This list of elements is explained in more detail in later slides. Not every project will require every element, we have can provide any and all of these and more if required.
  • #12: Again, these are standard capabilities, not limits. Pacing is an important requirement for digital blending and keeps the blend components in ratio even if one of the component streams becomes flow limited or constrained. It allows the blend rate to be set by recipe or by component availability. We also keep track during the blend so if there is any discrepancy, it is blended out over the remainder of the blend. Naturally, we include the ability to ramp up and down for the beginning and end of a blend. It is also possible to stop and restart or resume blending if necessary.
  • #13: The blend algorithms can be modified in virtually any way necessary to achieve the customer’s objectives. Basic customization is actually built into the cost of the product. Other functions can be provided as required.
  • #14: This slide shows even more options. DeltaV allows essentially unlimited calculational and reporting capabilities. We can even add Batch software with S88 recipe management for complex applications like lube oils or high performance specialty/racing fuels.
  • #15: This product should handle all but the most complex blending applications, but adding even more streams is easy. An example of the decoupling routine would be octane and RVP control. Components used to control octane (eg MTBE) and RVP (Butane) will affect both octane and RVP. The basic recipe will provide a coarse control, but analyzers can be used to trim octane and RVP. The decoupler overcomes the problem of interaction with the trim control using those components. The other features should be fairly obvious and are covered in more detail in the functional description. Note: I have updated this slide 3/27/00 to show 16 components and 8 additives as standard. Some of the other pieces including the functional description may not yet be updated.
  • #18: This is the most visible part of the package from the oerator’s perspective. We have designed the OI for ease of use leveraging the features of DeltaV and our experience with blending.
  • #41: In summary…. We can help identify potential savings and deliver a system to maximize those savings over time with a solution that offers the fastest payback and lowest risk possible. And after the new blender controls are up and running, our tools and services will ensure sustained performance over the lifetime of the system. We can easily adapt our solution to meet your special and unique requirements. With our resources and position as a global leader in this business, you can count on us to deliver results from start to finish!