See how Industrial AI-powered hybrid modeling dramatically improves the Design, Operation and Maintenance of plants to accelerate your profit, safety and sustainability goals.
Why Hybrid Models?
Today’s process industry companies
face unprecedented challenges and
must respond faster to:
1
Aspen Hybrid Models™
More Information +
Aspen Hybrid Models™
Less Information +
address these challenges by
integrating AI and machine learning
into existing models to enable more actionable insights. This empowers teams across your organization to adapt more quickly and efficiently in solving increasingly complex equipment and process demands.
More Information +
More Information +
2
What are Hybrid Models?
First Principles
More Information +
More Information +
More Information +
These are the modeling systems—typically built with
Aspen HYSYS® and Aspen Plus®
—that your experienced engineers
have leveraged to get your plant
to where it is today.
First Principles
Less Information +
Aspen Hybrid Models combine AI data
with first principles models and 40 years
of industry expertise to create a more predictive model that accurately simulates
a wide range of plant conditions faster and easier than ever before.
What Types of Hybrid Models are Available?
3
Leverage machine learning, first principles and domain knowledge to quickly create an empirical model based on lab or plant data to model novel process conversion units.
AI-Driven Hybrid Models
Use AI to create an empirical model based on running hundreds of simulations to improve accuracy and speed to resolution and deploy to multiple applications.
Reduced Order Hybrid Models
Capture hard to measure properties like color and stiffness of plastics or model special refinery and petrochemical processes.
AI-Driven Hybrid Models
Example +
Example +
Run gas separation unit models continuously online while extending the scale of modeling from units to the entire site like in a large gas-to-liquids plant.
Reduced Order Hybrid Models
Augment existing models with AI data from operations, a lab or pilot plant to calculate previously unknown variables and improve predictions.
First Principles Driven Hybrid Models
Example
Example
Example +
Heat exchangers, columns and reactors tuned to reality with AI-enabled calibration of efficiencies, reaction rates and more.
First Principles Driven Hybrid Models
What are the Key Benefits of Hybrid Models?
4
Expand Modeling’s Scope and Impact – Hybrid models enable process engineers to model several equipment types that are difficult using first principles alone, such as specialty chemical reactor models.
Democratize Modeling for Shifting Workforce - Hybrid models, with built-in operational data and data science, empower newer process engineers to develop models for equipment and assets without expert modeling skills.
Create Accurate, Fit-for-Purpose Models – Traditionally, different models have been used across functional areas with little transparency. Now with Reduced Order Hybrid Models, teams can leverage the same operating data within the model to create a closed loop that improves overall production optimization.
Accelerate Collaboration Between Disciplines – Reduced
order modeling also enables model alliance across departments.
For example, hybrid models simulating rigorous reactor data
can be used by planners to update their plan closer to targets.
Changing market demand
What is the Value Creation from Hybrid Models?
5
Over 80 companies have tested hybrid modeling and validated more than 100 use cases within the past year, reporting these benefits:
More accurate planning models will generate another $10+M USD yearly for a 200K barrel/day global refiner
A hybrid model used to monitor a gas separation plant will reduce energy demand 5 – 20% annually
Refining reactor hybrid models will create $5 – 10M USD per year by extending catalyst life and improving yields
Hybrid models of difficult polymer applications will save manufacturers at least $1M USD per line, per year
Some early use cases from companies testing hybrid modeling identified these measurable results:
Feedstock and crude price volatility
Sustainability pressures
New global competition
Bottom line
Hybrid models will fundamentally change how teams work together. Rather than spend hours building and rebuilding models and handling data streams in largely manual spreadsheets, teams will make forward-thinking, data-driven decisions about what to do – driving exponential gains in profitability, safety and sustainability, regardless of changing
market conditions.
What Customers
Are Saying
Learn More
Close X
What Customers Are Saying
Close X
“Aspen Hybrid Models are a major
advance in the field of chemical engineering. They bring together AspenTech’s process models and machine learning, and are a game changer in process engineering
and plant improvement.”
Dr. Karuna Potdar, Vice President, Technology Centre of Excellence, Reliance Industries Limited
Get the full story
Learn more about Aspen Hybrid Models in the paper
“Hybrid Modeling: AI and Domain
Expertise Combine to Optimize Assets”
DOWNLOAD NOW
Get the full story
Learn more about Aspen Hybrid Models in the paper
“Hybrid Modeling: AI and Domain Expertise Combine to Optimize Assets”
X Close
Download Now
Plant
Lab/Pilot
Process data
ML builds empirical model, guided by domain knowledge
Add 1st principles
& constraints
AI driven
Hybrid Model
Simulation
Simulation
data
ML builds fit for purpose simple and more robust model
Add
constraints
Reduced Order Hybrid Model
Plant
Simulation
Operations data
ML closes the gap between 1st principles model and data
Scroll Here
Scroll Here
X
Aspen Hybrid Models™ named 2021 Best Modeling Technology by Hydrocarbon Processing.
Deploy to process simulation
Simulation
FP model constraints
Simulator
Deploy to multiple applications
Example
Example