Food and Beverage Solutions - Optimization Using Model Predictive Control Ebook

Food and beverage
solutions
Optimization using model predictive
control
BEGI
N
World’s largest 80+ 20+
company dedicated countries industries
to industrial
automation and
information
Rockwell Automation is a member of the Association for Packaging
and Processing Technologies (PMMI) – a trade association made up of
more than 700 member companies that manufacture packaging,
processing and packaging-related converting machinery,
at a
commercially-available packaging machinery components, containers glance
and materials in the United States, Canada and Mexico.
PMMI members are the industry-leading solutions providers on your

processing and packaging supply chain, and PMMI resources help
22,00 $6.8
you connect with them.
SERVING CUSTOMERS FOR MORE THAN 115

YEARS
• Technology innovation
0 B
employees fiscal 2020
• Domain expertise
sales
• Culture of integrity and corporate responsibility
Rockwell Automation Model Predictive MPC optimization Value First

at a glance Control (MPC) PG 9 PG 26 2
PG PG 4
2
Food and beverage
solutions
Helping you exceed your business goals
INNOVATION
• The most powerful predictive modeling software in the
industry
• Enhancing the profitability of global manufacturers
• Delivering the highest ROI in the industry
DOMAIN EXPERTISE
• Combining technology and application knowledge
• Targeted industries
• Best practices from multiple industries
PRODUCTS
• Certified project managers
• Repeatable, measurable, auditable
Single
• Risk management point of
2500+ contact
employees
80
Bringing you countries
a world of 20
Average 13+
years‘
experience languages experience
ENERGY OPTIMIZATION • ENVIRONMENTAL • FOOD AND BEVERAGE •

POLYMERS MINERALS/MINING/CEMENT • NATURAL GAS LIQUIDS •
WATER/WASTEWATER

PG PG 4
2
Expertise in food and
beverage
Model Predictive Control (MPC) solutions
Rockwell Automation industrial process solutions have provided more than
115+ years of innovation, experience and global support. Industry best
practices are incorporated into every application, which is built upon a
composite application framework that leverages a Service Oriented
Architecture (SOA).
This scalable solution, when combined with library-based content, offers

personalization options that promote faster user adoption and time-to-value
for Food and Beverage producers. Our customers enjoy industry specific
results to increase revenue and increase saving that directly effect the
bottom-line.
Our capabilities help improve your operations

• Increase throughput up to 9%
• Reduce energy per unit of product up to 9%
• Reduce off-spec product up to 75%
• Improve yield up to 1%
• Reduce variability in quality up to 60%
Milk-based ingredient Milk processor improves Improving product

supplier benefits from operational efficiencies and quality and operating
tighter controls on increases yields efficiency on a global
caseinate production by reducing moisture scale (EMEA)
(EMEA) variability on its dryers
(Asia/Pacific)

PG 2 PG 4
How MPC
10
generates
4 benefits
Model predictive control effectively reduces the
variation of key product variables; optimization
KEY PRODUCT PROPERTY PERCENT OF
10 drives the average toward a more profitable

2
operating point.
PROCESS/ SPECIFICATION
10 LIMIT Maximum profitability is realized by operating
0 a process as close to its constraints as possible
while maintaining an appropriate margin of
9 safety. The margin of safety is largely dependent
8 upon the control system for the process.
9 Model Predictive Control Solutions are able

6 to reduce process variability and enhance
stability over and above what is currently
94 possible with more traditional control schemes.
This is accomplished through our multi-variable,
Push process
nonlinear, model predictive control capabilities.
toward limits
9
2 • REDUCES variability
LIMIT
Variability under Variability reduction Maintain quality • ACHIEVES “plant obedience”

9 operator control with Pavilion just
0 • MANAGES the process within
MPC within specification
contraints
BEFORE DURIN AFTER • ACHIEVES UPLIFT - operate closer
G to specifications and performance
limits while maintaining safety
TIME margins

PG 2 PG 4
The facility consists
of: CVs, CCVs, MVs, D
DVs
CONTROLLED VARIABLES (CVs)
V
WIND,
Process variables that need to be ROAD,
maintained at a target or within a set range OTHER
CARS
CONTROLLED CONSTRAINT
VARIABLES (CCVs) C
Process variables that should not be V
allowed to violate limits (upper, SPEED
lower or within a range) (MAXIMIZE)
CC
V
MANIPULATED VARIABLES
OIL
(MVs) PRESSURE
Process variables you can adjust that
affect the CVs (typically PID set-
DISTURBANCE
points) VARIABLES
(DVs)
Measured process variables that affect
the CVs that are not MVs
MPC is the replacement of the

M
driver with automated control
V
that takes all variables into
ACCELERATO
account for defined performance
R

PG 2 PG 4
Controller consists of a matrix of process model
pairs that explain important interactions in the
Controller process...
matrix
Advanced Process Control (APC) software solutions deliver
unparalleled modeling, prediction, control, and optimization capabilities CV1 CV2 CV3
in one unified solution.
MV1
Through superior data pre-processing, parameter formulation, modeling, and
prediction techniques, APC provides extremely accurate representations of
even the most complex processes. A multivariable predictive control solution MV2 NO MODEL NO MODEL
is the only one in the market capable of synthesizing all available information
— process knowledge, first-principle equations, or empirical models — into
a comprehensive model of the production process.
MV3
Flexible and efficient, the control solution continuously drives the most out of DV1
your production assets to achieve your unique business objectives. Predictive
analysis capability provides insight into key process parameters, reducing
variability and eliminating the guesswork between infrequent lab sampling. DV2 NO MODEL
Individual input output models are combined together in a matrix to give an
overall picture of how the process behaves. This model matrix forms the
heart of the MPC Controller. Predict the Proactive
values control
of the CVs by to achieve CV setpoints,
movement of all the prevent them from
MVs and DVs being violated, hence
reducing variability

PG 2 PG 4
PID MPC
SINGLE VARIABLE IN & SINGLE VARIABLE OUT MULTIVARIABLE IN & MULTIVARIABLE OUT
Set up a target and control process variable to the target. Control strategy based on a holistic comprehension
No awareness of how control changes impact other PID loops. of key process variables and their interdependencies.
FEEDBACK CONTROL PREDICTIVE CONTROL

The controller will take no action unless PV deviates from target Dynamic models developed through process step tests.
Controller action based on current and future PV deviations from
target.
INDIRECT CONTROL OF LAB MEASUREMENTS DIRECT CONTROL OF LAB MEASUREMENTS

Control property variables through proxy Controller predictions of lab measurements used
(temperature, pressure, etc.) for control updated as available.
POOR ABILITY TO HANDLE PROCESS DELAYS EXPLICIT DYNAMIC MODELS

During complex dynamic interactions Full understanding of process dynamics and interactions
CONSTRAINTS CONSTRAINTS
Only internal awareness of loop’s limits on set point and output Predict and monitor future values of constraints

PG 2 PG 4
MPC: PROCESS OPTIMIZATION CUSTOMER CHALLENGES
Dryer and Inconsistent dryer feed milk solids
Inefficient energy use
evaporator
Manufacturers of milk and other food ingredient Reduced throughput
powders recognize the constant need to operate High moisture and content
their processes more efficiently. The goal is to variability
increase throughput while simultaneously Balancing the rates between the
reducing costs in order to be competitive globally. evaporator(s) and the dryer
Many manufacturers are also shifting more

MPC BENEFITS
production towards higher value specialty
INCREASE throughput up to 15%
products that require higher quality standards
(such as INCREASE outlet air relative humidity up to 16%
low-fat, no-fat, protein-enhanced, etc.) In order to
produce these different products, plant REDUCE energy use up to 12%
operations must perform within narrower, more
consistent, product specifications. REDUCE off-spec product up to 75%
In the dairy industry, a variety of dryers and INCREASE moisture yield up to 1%
evaporators are common in the production of
powdered products to achieve cost-effective REDUCE moisture variability up to
moisture removal from concentrated liquids.
Dryers and evaporators are the most energy 60%
intensive units used in the dairy processing
industries. Optimal operation of this process IMPROVE the consistency of dryer
is essential to reduce costs and help meet feed milk solids
final product quality specifications.
Process optimization for results: Model Predictive Control has proven to be an

effective technology to help our customers regardless of market conditions. Whether it
is getting more out of existing equipment, or improving operations for tighter margins,
MPC from Rockwell Automation is the answer to your challenges.

PG 2 PG 4
Upper & lower Upper Lower
Maximize Target
constraint constraint constraint
Manipulated Controlled Disturbance

M CV Variable Constraint D
CCV
MPC: PROCESS OPTIMIZATION V
Variable Variable
V
Variable
Dryer and evaporator EVAPORATOR

DV FEED %
SOLIDS
FEED
EVAPORATOR EVAPORATOR EVAPORATOR
SILO MV FEED MV TVR STEAM CV STAGE TEMPS
Dairy dryer/evaporator balance FLOW PRESSURE
SURGE TK
application SURGE TK
TIC
TIC
SIC
STEAM
FIC
The Rockwell Automation Model Predictive Control FIC
(MPC) and Optimization Application is tailored to

the individual configuration of each unit and
specific site objectives. It combines the Rockwell EVAPORATOR
CV CONDENSER
VALVE TI
Automation Dryer and Evaporator Balance Control DRYER
MV EXHAUST TEMP
STICKY STEAM PI
CV POINT
Applications for a complete solution which FEED
incorporates full thermal control and optimization. DRYER TIC DRYER FEED
MV INLET CV RATE
TEMP
Energy balancing AIR
Implementing this dual application solution DRYER INLET

MV AIR
offers a unique thermal energy balancing

FLOW DRYER
CONCENTRATE
opportunity between the evaporator and the CV %SOLIDS
CV
DRYER FEED
PUMP
SPEED
dryer. Optimal energy use and cost control can STEAM
TIC PRODUCT
be achieved. STEAM
TO
STORAGE
TIC
AIR DRYER FEED

DV %
SOLIDS
MOISTURE
AIR
PRODUCT
CV MOISTURE
Milk Energy Increase Moisture

CHALLENGES variability inefficiency throughput variability
Stabilize Push Operate Dry to

OBJECTIVES process evaporators at specification
limits
Increased Reduced Increased Decreased
BENEFITS milk solids energy throughput variability
12% 16% 60%

PG 2 PG 4
0
MPC: PROCESS OPTIMIZATION CUSTOMER CHALLENGES
Dryer Need to increase capacity to meet
demand Too much off-spec product
s
Manufacturers of milk and other food ingredient Inefficient energy utilization
powders recognize the constant need to operate Moisture variability: under and over drying
their processes more efficiently. The goal is to product
increase throughput while simultaneously
reducing costs in order to be competitive globally. MPC BENEFITS
Many manufacturers are also shifting more
production towards higher value specialty
products that require higher quality standards
REDUCE energy use up to 9%
(such as
low-fat, no-fat, protein-enhanced, etc.). In order to REDUCE off-spec product up to
produce these different products, plant
operations must perform within narrower, more 75% INCREASE moisture yield up to
consistent product specifications.
1%
In the production of powdered products to
achieve cost-effective moisture removal from REDUCE moisture variability up to
concentrated liquids, dryers require maximum 60%
efficiency and optimization to improve yield and
save energy costs. Dryers are one of the most
energy intensive units used in the dairy
processing industries. Optimal operation of this
process is essential to reduce costs and help
meet final product quality specifications.

PG 2 PG 4
1
Maximize Target
MPC: PROCESS OPTIMIZATION

Dryer
Variable CCV Variable Variable
V V
Application scope
The MPC application works with single and multi-stage DRYER
MV EXHAUST
spray drying processes as well as flash dryers and STEAM P TEMP
processes with fluidized beds. It controls product I FEED
average moisture to a desired target while reducing TIC
DRYER
DRYER CV FEED
overall moisture variability. Relative humidity and capacity MV INLET
RATE
are also managed to help achieve optimal performance. TEMP
Moisture management AIR

The MPC software controller compensates for
variations in the milk solids concentrate as well as inlet DRYER INLET
air relative humidity and other disturbances that affect MV AIR
FLOW DRYER
drying efficiency. By reducing moisture variability, the
average moisture target can be increased without CONCENTRATE
CV %SOLIDS DRYER FEED
compromising product quality. This increase in CV PUMP
moisture target can also produce higher yields from SPEED
the same milk solids, increase dryer capacity and STEAM
reduce product re-work TIC PRODUCT
and energy usage per ton of finished product. This STEAM
TO
application includes early engagement control, which STORAGE
TIC
sees the dryer under MPC and optimization. This helps DRYER FEED
AIR DV %
in achieving optimal quality and thermal efficiency SOLIDS
within minutes of the dryer commencing powder MOISTURE
AIR
production. Capacity management is achieved by PRODUCT
improving the thermal efficiency of the dryer and CV MOISTURE
controlling chamber temperatures and airflows to
achieve the capacity target. At the same time, the
absolute chamber relative humidity constraint and
Increase Increase Excessive
product quality are optimized. CHALLENGES capacity
Off-spec
yield energy
use
Performance metrics and visualization
Manage Reduce moisture Control Reduce
Real-time visualization in a browser-based interface OBJECTIVES process variability variables energy
presents metrics that allow operators and management
to monitor the performance measures of production, Increased Reduced Increased Reduced
quality and energy. These dashboards allow current and BENEFITS throughput off-spec yield 1% energy
predicted plant performances to be viewed and managed 9% 75% 9%
to realize and sustain the lifetime business value of MPC
investment.

PG 2 PG 4
2
MPC: PROCESS OPTIMIZATION CUSTOMER
Evaporators CHALLENGES
Inconsistent concentrate solids
Prescriptive and predictive analytic process
Inefficient energy use
solutions: Evaporators are critical in the
Need to increase capacity to
production of powdered products to achieve
meet demand
cost effective moisture removal from MPC BENEFITS
concentrated liquids. Evaporators are one of the
most energy intensive units used in the dairy INCREASE throughput up to 9%
processing industries. Optimal operation of this
process is essential to reduce costs and help REDUCE energy use up to 9%
meet final product quality specifications.
REDUCE concentrate solids variability up to 60%
Rockwell Automation recognizes these
challenges and offers an application focused on IMRPOVE the consistency of dryer feed milk
delivering valuable evaporator control for solids
optimization and performance.
Dairy evaporator control application:

Application assists in controlling and optimizing
the evaporation process during the production
of
powdered dairy products. Powered by
FactoryTalk Prescriptive and Predictive Analytics
patented technology, this application is tailored
to the individual configuration of each unit and
specific site objectives. It may be a stand-alone
option or combined with the Dryer Control
Application for
a complete drying solution.

PG 2 PG 4
3
Maximize Target

CCV
Variable Variable
V
Variable
Evaporator
EVAPORATOR
Application scope D FEED %
V SOLIDS
The Evaporator application works with single and
multiple effect evaporators and can optimize FEED EVAPORATOR EVAPORATOR EVAPORATOR
SILO
M FEED FLOW
M TVR STEAM CV STAGE
energy usage by including mechanical and thermal V V PRESSURE TEMPS
vapor recompression loops if they exist. It has also SURGE TK
been successfully applied in the most complex TI
SURGE SI
evaporator configurations, including rising-film, TI C
TK C
STEAM C
falling-film, wiped film and units with pre- FI
C
evaporation or finisher. FI
C
Milk solids management
This application reduces the variability of the milk
solids concentrate, allowing the average
production target to be shifted upward without EVAPORATOR
CV CONDENSER
compromising the downstream drying process. VALVE T
Variations in I
STICKY
milk feed temperature and concentrate feed to CV
POINT
the dryers, helping to improve the consistency
in product quality.
Increase Energy Inconsistant

CHALLENGES demand costs concentrate solids
Increase Reduce Reduce concentrate

OBJECTIVES throughput energy solids variability
Increased Reduced Reduced

BENEFITS throughput energy variability
9% 9% 60%

PG 2 PG 4
4
Dryer MPC
results Dryer/Evaporator
or MPC
commissioned
ENERGY THROUGHPUT
SAVINGS INCREASE
14 .5% less +10.6% gal/year
BTU/gal
The Rockwell Automation Dairy Model Predictive Control (MPC) applications assist in controlling and optimizing the drying and
evaporating process during the production of powdered dairy products. Powered by Rockwell Automation patented technology,
this application is tailored to the individual configuration of each unit and specific site objectives. It may be a stand-alone option
with drying, evaporating, or combined for a complete drying solution.

PG 2 PG 4
5
Reduction in
variability Example: dryer
results (moisture)
49%
BEFORE AFTER
Model Predictive Control Model Predictive Control
Moisture Management
The MPC software controller compensates for variations in the milk solids concentrate as well as inlet air relative humidity and
other disturbances that affect drying efficiency. By reducing moisture variability, the average moisture target can be increased
without compromising product quality. This increase in moisture target can also produce higher yields from the same milk solids,
increase dryer capacity and reduce product re-work and energy usage per ton of finished product.

PG 2 PG 4
6
Exhaust relative humidity
control
MPC Early engagement and capacity management
OFF This application includes early engagement control, which sees the dryer under MPC and
optimization. This helps in achieving optimal quality and thermal efficiency within minutes of the
dryer commencing powder production. Capacity management is achieved by improving the thermal
efficiency of the dryer and controlling chamber temperatures and airflows to achieve the capacity
target. At the same time, the absolute chamber relative humidity constraint and product quality are
optimized.
MPC Performance metrics and visualization

ON Real-time visualization in a browser-based interface presents metrics that allow operators and
management to monitor the performance measures of production, quality and energy. These
dashboards allow current and predicted plant performances to be viewed and managed to
realize and sustain the lifetime business value of MPC investment.
MPC MPC
OFF
• Humidity frequently exceeds limit •ONKeeps process Relative Humidity under limit even with changes in Ambient
• Humidity
Wide distribution = poor manual
control • Narrow distribution = tighter control with MPC
• Over drying = wasted energy • Less over drying = energy savings

PG 2 PG 4
7
MPC: PROCESS OPTIMIZATION CUSTOMER
Dairy CHALLENGES
standardization
Variability in final product composition
In milk processing, powder compositional
Give away of fat & protein in final
control is one of the most difficult process
product Out-of-spec quality
control challenges. Typically fat composition is
regulated through ratio control of cream to MPC BENEFITS
skim milk feed.
Protein is controlled through the addition of REDUCE final product component variability up to 60%
lactose or permeate. Traditional regulatory control
of liquid fat and protein composition is REDUCE fat and protein in the final product by up to
problematic due to the interactions of the 0.35%
Adjusting
ingredientratios and variables may affect both
streams.
(to specification limits)
the fat and protein composition. Furthermore, DECREASE product give-away
other complexities arise when relating the final
powder composition back to the wet-side LIMITING out-of-spec product
composition. Managing the effects of
evaporation, process dynamics and
compositional instrumentation measurement
accuracy is nearly impossible without an
automated system.
A process optimization solution using MPC
allows control of dairy standardization for
definitive results.

PG 2 PG 4
8
Maximize Target

CCV
Variable Variable
V
Variable
Dairy
standardization
The MPC application from Rockwell Automation
Standardization application
provides a closed loop standardizing solution with MV
RAW MILK
FROM
SILOS
superior targeting of final powder composition
POWDER
and constraint management through real-time MV CV % FAT
control of wet-side standardizing and process

accurate inferential powder composition models. TIC
FIC
Monitor and control

Powder composition inferential models are
developed to provide a continuous virtual CREAM
MV FEED
measurement of the final powder composition.
POWDER
The models account for mixing in silos and CV %
balance tanks, loss of water through evaporation, PROTEIN
post standardizing additives and process

dynamics through the evaporator and drying
systems.
The inferential models are biased with
validated in-process testing (IPT) results to
account for instrument drift. Appropriate Product Out-of-spec Final
CHALLENGES give- quality product
Statistical Process Control (SPC) rules are away variability
applied in the biasing algorithm. Decrease Limit Reduce product
OBJECTIVES product out-of-spec composition
give-away variability
Reduced fat .35% and Reduced Reduced
BENEFITS protein .25% in final out-of-spec variability
product 60%

PG 2 PG 4
9
Maximize Target
Liquid composition MPC controller
By utilizing in-line analytical measurement of wet-side M Variable CV Variable CCV
Constraint
Variable D Variable
composition with instrumentation, standardization is run V V
as a continuous process. The FactoryTalk® Analytics ™
MPC Liquid Composition Controller adjusts the cream and
lactose or permeate ratios typically at a 30-second
LIQUID
interval to control to liquid composition targets. The
CONTROLS
multi-variable predictive control models handle the
interactions between the ingredient streams.
CV CV CV
Powder composition MPC controller
The Powder Composition Controller, which accurately
PROTEIN/LACTOSE
encompasses process dynamics, is used to make POWDER
STANDARDIZED
adjustments to the targets of the Liquid Composition CONTROLS
MILK FLOW
Controller based on feedback from the statistically
FAT/SNF
biased inferential powder composition models.
TOTAL
CV CV
MILK
MILK
Performance and visibility
Historical and predicted milk compositions along with M RAW MILK FROM
SILOS
X X
past and planned controller actions are displayed V
X
POWDER %
POWDER %
through the Pavilion8® Console. The web browser M LACTOSE X X
PROTEIN
interface allows operators, engineers, and managers to FEED
V
clearly see all pertinent composition and ingredient flow X
FAT
information M CREAM
FEED
X X
in both comprehensive and custom trends. V
Optimization Multivariable models handle interactions D RAW MILK
X X M MILK X X
that cannot be solved with regulatory PID control thus COMPOSITION FAT/SNF
V V
reducing liquid compositional variability.
Variable dynamic models and SPC logic allow for superior D CREAM
X
X M MILK X X
COMPOSITION V PROTEIN/LACTOSE
powder compositional targeting further reducing the V
variability in the final product. Detection of instrument X
drift provides valuable calibration assistance. Tighter
compositional control allows fat and protein targets to be BENEFITS • Optimized fat by 0.35%
driven to the specification minimums leading to •
• 50% Reduced fat variability by 50% Reduced product give–away
significant
financial savings. Web-based visualization platforms facilitate • Reduced protein variability by 60% • Reduced off-spec
easy access to real-time controller and performance data • Optimized protein in final product by • Improved fat and protein
enabling operations and management to make proactive 0.25% consistency
decisions further maximizing benefit.

PG 2 PG 4
MPC: LINE OPTIMIZATION CUSTOMER CHALLENGES
Food Inconsistent production rates
Excessive energy consumption
lines
Food Line Optimization provides control and Off-spec product increasing scrap or rework
optimization of food line processes, from raw Loss of production from not balancing equipment
material introduction to final pre-packaged
product, using patented, industry-leading
technology. MPC BENEFITS
The Food Line Application provides control
performance by reducing the disturbance
effects from raw material variations and feed
REDUCE energy use up to 12%
interruptions. The application responds to raw
material disturbances such as color, density and REDUCE quality variability up to
material size, as well as changes in steam
supply, cooling capacity and ambient 60% INCREASE yield up to 1%
conditions. With
real-time optimization, increased throughput and IMPROVE operational stability
yield is achieved by operating the plant closer to
process constraints. An MPC solution drives
more production, optimizes energy, and lowers
off-spec product with existing equipment that
does not require additional capital investment to
achieve greater profits.

PG 2 PG 4
1
Maximize Target

CCV
MPC: LINE OPTIMIZATION V
Variable Variable
V
Variable
Food lines
Monitor and control
Monitoring and controlling production can
LINE COLOR
significantly reduce process inconsistency and MV FEED CV VARIATION
variability, limiting off-spec product and RATE
rework of materials. Through predictive DEFECTS SORTER OVEN COMPRESSOR

CV (CRIT, MAJ, MV SETTINGS MV TIMES/TEMP CV LOAD
control, our solution delivers better process TOT)
stability and higher consistency in product
quality. The MPC
software provides real-time visibility of
production metrics compared to a baseline and
promotes a performance-driven culture.
Process constraints are monitored and future MIXER EXTRUDER CORE SHAPER HOISTER OVEN FLAVORING LINE
FILLER
violations predicted. Corrective action is taken as
PRODUCT
necessary to ensure that the process runs close D
FEED
MV
NH3
TEMPERATURE
QUALITY SUCTION CV
to constraints without violating acceptable limits. V PRESSURE
(EXIT
FREEZER)
The application stabilizes the production process

by coordinating multiple loops with a
multivariable predictive control algorithm (Model
Predictive
Control). In addition, the Food Line Optimization Inconsistent Excessive energy Off-spec Imbalanced
CHALLENGES production consumption product equiptment
MPC enables an innovative approach to optimize
sorters, fryers, freezers and other food line Balance Reduce Reduce concentrate Balance
equipment, OBJECTIVES production rates energy solids variability equiptment
by calculating and controlling temperatures, raw
Increased Energy Decreased Reduce product
material rates, additive rates, dwell times and BENEFITS throughput reduction variability quality variability
other key process parameters for balanced 15% 12% 60%
equipment efficiency.

PG 2 PG 4
Control
structure before
??!!??!!
Step B, C, A,
MPC
A PID loop system is limited to a single
D? area of your entire process. This does
Limits? not allow for proactive actions should any
How much?
disruption in your system occur, such as
unexpected changes in humidity, pump
speeds, or feed variability.
PID PID PID PLC/DCS
Without a holistic view of any process,
an operator will always be in reactive
mode, and not in a predictive or
prescriptive mindset.
Lab Process
Model Predictive Control sees the
whole picture throughout the process
and then is able to make adjustments
BASIC REGULATORY prior to any slow-down or stoppage.
CONTROL

PG 2 PG 4
MPC/optimization PRODUCTION RATE MAX
Higher yields, lower energy
architecture
overview
With MPC we can create virtual
Virtual MPC
online analysers. These are inferential Online
measurements that use easy to Analyzers
measure information in your process

to provide continuous feedback of
quality parameters that typically PID PID PID DCS
require lab sampling to measure. layer
These VOA’s are typically used to

provide feedback to our Pavilion8 MPC
controller enabling continuous closed Lab Process
loop control of important quality
parameters.

PG 2 PG 4
Predictive
Analytic Machine
Learning
Predictive Analytic Machine Learning can be leveraged throughout a cement
facility to provide additional oversight. Many Soft Sensors are in use to drive
quality control within the previously described MPC applications where no
online analyzer is available. Predictive Maintenance targets any equipment
that rotates or has concerns on failure or high maintenance costs: mills or
mill components, kiln drives, combustion air fans, conveyors, critical pumps
or even trucks with connected sensors. Improve maintenance dispatch
timing to improve the return for every maintenance dollar spent.
Anomaly Detection is very useful for any equipment section that periodically
goes wrong and providing operators early insight into developing problems
with kiln wall ring formation or gas flow blockages, ball mill losses, or
equipment failure concerns where there is not an existing history of
identified failures. With Anomaly Detection you can frequently gain
advanced warning on detectable issues (measurement deviations) and
resolve them to minimize periods of upset or poor performance.
If MPC is continuously driving good periods of plant performance to the

highest levels, many of these predictive analytics are reducing time
spent performing abnormally. Two collaborative drives to increase
cement plant
profitability.

PG 2 PG 4
ValueFirst® project
THE WAY WE DO methodology
BUSINESS
ValueFirst is the Rockwell Automation unique
customer engagement process that guarantees Assess Deliver Audit Sustain
measurable value and lasting results. ValueFirst
aligns every aspect of the Rockwell Automation
people, process and product with customers’ • Propose & plan • Design, develop, deploy • Commissioning • Support
business needs to deliver and demonstrate the

• Confirm business value • Data gathering & • KPI configuration • Knowledge transfer
incremental and cumulative value of our validation
• Set expectations • Training • Performance metrics
solutions. • Model
The ValueFirst methodology encompasses every • Determine benchmark development • Performance validation • Ongoing support
step in our customer engagement process. metrics
• Application
From a simplified contracts process to a uniform deployment
set of performance metrics on every Model VALUE-BASED VALUE-BASED MEASURE ON-

Predictive Control (MPC) and Environmental PROPOSAL SOLUTION VALUE GOING
VALUE
Compliance solution, we are is dedicated to
delivering the highest possible value to
customers, faster and better than alternative
solutions.

PG 2 PG 4
Sustained value
services
Support programs to help protect your investment LEVEL 1 SUPPORT
The Sustained Value Team is committed to keeping your investment at its
optimal performance level, and our program of value-added services is • Email and live telephone support
designed to help deliver results to your bottom-line year after year. To
achieve this goal, we off er proactive support to maximize your operations. • Service-Pack Releases provide updates and system
This includes new software features and industry-specific applications, changes
solutions and services to help customers increase production, reduce costs,
improve quality and increase profitability. • MPC System recovery assistance from server failures
Our Sustained Value Program provides you with the foundation to continue
to leverage our industry leading solutions, ongoing insight into plant, LEVEL 1 AND LEVEL 2 SUPPORT
application and solution performance and a proven path to achieve the
greatest possible annual return on your investment in solutions from • Email and live telephone support
Rockwell Automation.
• Service-pack releases provide updates and system changes
Key Benefits
• Proactive maintenance for optimal performance • MPC System recovery assistance from server failures
• Improved application availability
• Refresher training for maintaining operator expertise • Quarterly Pavilion8 MPC status reporting and
troubleshooting APC application issues
• Greater software availability through product enhancements, updates
and quarterly application backups • Web-based Pavilion8 Support Knowledgebase and
• Enhanced visibility of application performance through annual onsite visits available for APC
quarterly status reports
• Disaster restoration of MPC application

PG 2 PG 4
BACK
Expanding Human Possibility, FactoryTalk, FactoryTalk Analytics, ValueFirst, Pavilion8, PlantPAx, and Rockwell Automation are trademarks of Rockwell Automation,
Inc.
Trademarks not belonging to Rockwell Automation are property of their respective companies.
Publication RSPPFB-PP001B-EN-P - January 2020

Copyright © 2020 Rockwell Automation, Inc. All Rights Reserved. Printed in
USA.

Food and Beverage Solutions - Optimization Using Model Predictive Control Ebook

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Food and Beverage Solutions - Optimization Using Model Predictive Control Ebook

Uploaded by

Copyright:

Available Formats

Food and beverage

PMMI members are the industry-leading solutions providers on your

SERVING CUSTOMERS FOR MORE THAN 115

Rockwell Automation Model Predictive MPC optimization Value First

Bringing you countries

experience languages experience

ENERGY OPTIMIZATION • ENVIRONMENTAL • FOOD AND BEVERAGE •

Rockwell Automation Model Predictive MPC optimization Value First

This scalable solution, when combined with library-based content, offers

Our capabilities help improve your operations

Milk-based ingredient Milk processor improves Improving product

Rockwell Automation Model Predictive MPC optimization Value First

10 drives the average toward a more profitable

9 Model Predictive Control Solutions are able

Variability under Variability reduction Maintain quality • ACHIEVES “plant obedience”

Rockwell Automation Model Predictive MPC optimization Value First

MPC is the replacement of the

Rockwell Automation Model Predictive MPC optimization Value First

Rockwell Automation Model Predictive MPC optimization Value First

FEEDBACK CONTROL PREDICTIVE CONTROL

INDIRECT CONTROL OF LAB MEASUREMENTS DIRECT CONTROL OF LAB MEASUREMENTS

POOR ABILITY TO HANDLE PROCESS DELAYS EXPLICIT DYNAMIC MODELS

Rockwell Automation Model Predictive MPC optimization Value First

Many manufacturers are also shifting more

Process optimization for results: Model Predictive Control has proven to be an

Rockwell Automation Model Predictive MPC optimization Value First

Manipulated Controlled Disturbance

Dryer and evaporator EVAPORATOR

Dairy dryer/evaporator balance FLOW PRESSURE

The Rockwell Automation Model Predictive Control FIC

(MPC) and Optimization Application is tailored to

Energy balancing AIR

Implementing this dual application solution DRYER INLET

offers a unique thermal energy balancing

AIR DRYER FEED

Milk Energy Increase Moisture

Stabilize Push Operate Dry to

Rockwell Automation Model Predictive MPC optimization Value First

Rockwell Automation Model Predictive MPC optimization Value First

MPC: PROCESS OPTIMIZATION

Moisture management AIR

Rockwell Automation Model Predictive MPC optimization Value First

Dairy evaporator control application:

Rockwell Automation Model Predictive MPC optimization Value First

Manipulated Controlled Disturbance

Increase Energy Inconsistant

Increase Reduce Reduce concentrate

Increased Reduced Reduced

Rockwell Automation Model Predictive MPC optimization Value First

Rockwell Automation Model Predictive MPC optimization Value First

Rockwell Automation Model Predictive MPC optimization Value First

MPC Performance metrics and visualization

Rockwell Automation Model Predictive MPC optimization Value First

Rockwell Automation Model Predictive MPC optimization Value First

Manipulated Controlled Disturbance

control of wet-side standardizing and process

Monitor and control

post standardizing additives and process

Rockwell Automation Model Predictive MPC optimization Value First

Rockwell Automation Model Predictive MPC optimization Value First

Rockwell Automation Model Predictive MPC optimization Value First

Manipulated Controlled Disturbance