Previous Page

Chapter 10 - Virtual Plant

395

Application Detail
Virtual versions of scalable distributed control systems (DCS) are now
available that support the development of control-system configuration,
logic checkout, and operator training using a single PC or multi-node PC
environment. These packages are designed to be used on a laptop or office
PC. Using this capability, it is possible to configure all of the features that
the control system supports (for example, continuous control, batch control, advanced control, and their associated workstation displays, alarms,
and historian data collection) without purchasing system hardware.
Using this capability, the execution is simulated for the operator interface
and selected control modules defined for the plant control system. This
capability is used for control logic checkout. And, using the control and
I/O block and Fieldbus block simulation capability, field measurement
values and status are manually supplied to the simulationor provided
by blocks used to simulate the process.
The full range of OPC features in the control system are available and may
be used for the development and testing of OPC interfaces. Plus, before
your plant is even constructed, the OPC-compliant process simulation
package is used for process and control design checkout Thus, most of the
features of a complete control system are made available on the PC.
Using the virtual plant, it is possible to accomplish the following:
Engineer systems anywhere - Engineering work is done without
the constraints of working at a dedicated configuration station. The
virtual plant is installed on a home or office computer or on a
laptop. The work done on this computer is easily exported and
quickly imported to the actual control system.
Verify control logic and display implementation - Software
modules configured for your plant control are executed in your
computer for control and display checkout. Using the simulate
capability of the controller blocks and Fieldbus I/O blocks, the
measurement value and status are examined during the checkout of
control logic, operator displays, alarms, and trending.
Test OPC interfaces- The simulation environment fully supports
OPC. Executing modules in the computer allows an OPC interface
to be checked out completely. This feature is used to develop
applications that can execute in the control system.
Support control system training - Dynamic process simulations
are created to support operator and engineering training one
controller at a time. Function blocks can be used to create a process
model or a process simulation package can be incorporated in the

396

Advanced Control Unleashed

virtual plant to create a training tool.

The control system provides an integrated development environment in
which the continuous, sequential, and batch control for a plant are
engineered. Using drag-and-drop techniques, the configuration of control
is relatively easy, as shown in Figure 10-5.

Figure 10-5. Integrated Development Environment

The Fieldbus Foundation architecture has been adopted by the control system. The control functionality that is assigned to a controller or Fieldbus
device in the control system is executed in the PC and the results displayed in the configured operator interface, without any change in the
control or operator interface configuration. Using the simulate capability
of the control system and Fieldbus input-output function blocks, field
measurement values are available for control-system and operator-interface checkout as shown in Figure 10-6.
Using the SIMULATE parameter of the controller and Fieldbus I / O
blocks, it is possible to manually inject field measurement values and status to check out control logic, alarming, displays, etc. By changing the status or value associated with a measurement, the action taken by control
logic execution of the control system in the virtual plant environment may
be examined. Also, this capability is used to test whether block parameter
values and status are displayed correctly in the screens designed for the
operator interface. By changing the measurement value through the SIMULATE parameter, it is possible to force an alarm condition to verify
assigned alarm priority and the manner in which alarms become visible to
the operator; e.g., alarm banner in the operator interface.

Chapter 10 - Virtual Plant

397

Figure 10-6. Standard Support for I/O Simulation

Dynamic process simulations are created to support process design and

operator and engineering training. Using the standard function block
available in the control system, it is easy to create a simple dynamic model
based on the process design.
The standard calculation and dynamic blocks provided in the Analog
Control and Math palette in control-system configuration tools may be
used to construct a simple dynamic process model for testing control strategies and operator training. As shown in Figure 10-7, the process is simulated using function blocks that are wired to simulate inputs and outputs
of I / O function blocks.
Application packages that support an OPC interface are compatible with
the virtual scalable DCS. For example, as shown in Figure 10-8, this OPC
capability allows the high-fidelity dynamic process simulation software to
be used with the control-system configuration for plant and control design
as well as for control checkout and operator training. Many process-simulation packages support checkout and operator training by the use of
enhanced tieback simulations. These process-simulation applications utilize the simulate capability of the control system and Fieldbus I / O blocks,
as illustrated in Figure 10-8.
An application and interface are provided in the simulation environment
to allow the simulated control-system execution to be coordinated with
the process-simulation execution. Using this application, it is possible to
examine the state of the control simulation and whether I / O and control
blocks are correctly set for simulation. Also, from this interface, it is possible to change the simulated control execution to be faster or slower than

398

Advanced Control Unleashed

Figure 10-7. Process Models Constructed from Function Blocks

Figure 10-8. High-Fidelity Process Simulation

real time and to pause the execution. Common requirements such as the
initialization of the simulation, block mode, and block dynamics are
addressed. An OPC interface is provided to allow process-simulation
packages to also invoke the same requests that are made through this user
interface to the control simulation.

Chapter 10 - Virtual Plant

399

Rules of Thumb
Rule 10.1. The actual control system configuration and displays must be
imported and exportedfromthe virtual plant. The use of a reconstructed or
emulated control system and operator interface is fundamentally
unsound. The actual configuration and displays and all the features of the
actual system must be used.
Rule 10.2. The use of actual hardware instead of a virtual plant is costly, delays
and restricts the access, and decreases the utility of a simulation for training and
testing. If an engineer needs something he can touch, give him last year's
obsolete controller.
Rule 10.3. The process simulation must use an OPC interface to the controlsystem configuration and software tools. This will facilitate the connection of
the system to other advanced diagnostic, data-analysis, and control tools.
Rule 10.4. The definition of the informationflowbetween the control system
and the process simulation should be configurable and easily done without requiring programming knowledge. The connection should be as simple as browsing and dragging and dropping tags in an interface table.
Rule 10.5. The process simulation and control system should support execution fast or slower than real time. Also, the ability should be provided to
freeze both.
Rule 10.6. The virtual plant should support single commands to initiate setup
of the process simulation and control system to a known state. There should be
an ability to easily capture and restore the state of the process and control
system.
Rule 10.7. It should be possible to run the virtual plant in a single PC or be
distributed between multiple PC's. No change in control-system configuration should be required to relocate where the process simulation and control system is executed.
Rule 10.8. The process-simulation software must be graphical and have highlevel and a wide variety of physical property packages, state equations, and configurable functional blocks and unit operations. Custom simulations that require
the setup and numerical integration of differential equations are too complex to construct and maintain and do not have all the physical properties
and their dynamic interaction with process conditions. If you spent the
millions of dollars to create a custom simulation of an entire plant, only
the programmers who created the custom model could debug and modify
it.

400

Advanced Control Unleashed

Rule 10.9. To verify that the configuration matches the control definition and
to teach operators how to use the interface, an enhanced tieback simulation is suf-

ficient. Configuration engineers prefer to use a tieback simulation to test

their work so they don't have to get involved in the process or learn how
to start and stop a process simulation. Also, these tieback models have an
input and output file card channel (FCC) screen that lets the configuration
engineer conveniently set and monitor input and output signals.
Rule 10.10. To test whether the control definition is correct, to learn how the
process and control system behave for upsets, startups, and failures, and to prototype advanced control solutions, a high-fidelity dynamic process simulation is

needed. While the configuration engineer is too busy and will move on to
the next project too quickly to take advantage of a high-fidelity virtual
plant, it is a valuable ongoing resource for operators and engineers at the
plant and can lead to real-time optimization. While this method does not
check the I/O file card channels, these assignments are verified anyway as
a part of the normal course of the checkout of the field wiring.

Guided Tour
This tour illustrates the potential ease of use and convenience of an integrated interface that is possible for control simulation embedded in an
industrial control system. The following areas are addressed:
Control Simulation Setup
Coordination of Process and Control Simulation
Viewing Simulation State
Modern control systems allow modules for control, monitoring and
calculate applications to be engineered independent of where they are
executed. Thus, a module is assigned to execute on controller hardware or
a PC platform with no change in its configuration. Such capability makes
it possible to take a control system configuration and quickly use this to
construct engineering and operator training system based on PC
platforms i.e. no controller hardware.
Modules that have been engineered for the plant control and normally
execute in hardware controllers are re-assigned to run in the PCs that have
the control system software installed, as shown below. When these modules are loading in the module folder on the PC as shown in Figure 10-9,
the assigned modules automatically begin to execute at their configured
rate of execution. Late binding of module information in the communications between the PC stations allows the operator and engineering interface to access the inputs, alarms, and calculate values associated with
these modules without any change in system configuration.

Chapter 10 - Virtual Plant

401

Figure 10-9. Module Assignment for Execution

Within the PC nodes that make up the virtual plant environment, an application may be added to provide process simulation. The process-simulation application may use OPC to read the actuator positions from the I / O
blocks within the modules. Based on these values, the process-simulation
package may calculate the process inputs and writing these values to the
simulate parameter of the input blocks. Thus, within the accuracy of the
process simulation, the measurement and control response will duplicate
what is seen in the actual installation.
To provide an effective engineering or training system, it is necessary to
coordinate the process-simulation package execution with the controlmodule execution. A simulate interface is provided in the control system
for this coordination as illustrated in Figure 10-10.
Through the selection provided in the Setup, it is possible to:
Enable or Disable the Simulate function of I / O blocks.
Change Control and Output blocks to their setup mode so the
process-simulation application may change these outputs to a
known starting point in the simulation.
Initialize the dynamic blocks in all modules assigned to the node so
that they are at steady-state solution.
Allow the operating state of all modules to be saved or restored at
any point in time. This allows the simulation to be restarted at
different conditions.

402

Advanced Control Unleashed

Figure 10-10. Interface for Simulation Coordination

Allow all modules assigned to the node to be run faster or slower

than real time. This capability allows the control simulation to
match the speed of the process simulation.
The simulate state of individual modules is examined by selecting the
Summary tab as shown in Figure 10-11.

Figure 10-11. Examining Module Simulation State

Chapter 10 - Virtual Plant

403

The status of module I/O blocks to support simulation is examined by

selecting the Summary tab. From this view, it is possible to view all modules assigned to a selected area. A frown face is used to identify conditions
that would prevent a process simulation from interacting with the control.
These can be of the following types:
Simulate - One or more I/O blocks do not have simulation enabled.
Status - The input status of an I/O block is Uncertain or Bad. This
might indicate that the process simulation is is calculating an
incorrect value or no value.
Mode - The I/O block is not in its normal mode of operation.
When a module is selected from this view, details concern the I/O and
control blocks in the module are automatically displayed. Also, the user
may enable or disable simulation of an individual block from this
interface. Using this capability, the user can simulate disturbance in the
process during control-system checkout and operator training

Theory
The software and the associated configuration of some modern control
systems are designed to be independent of the hardware, and thus may
execute in multiple operating environments without change. When this
approach is taken, it is possible to distribute control-system functionality
in a variety of ways without needing to reengineer the software or reconfigure the associated applications. All features of the control system are
combined on a single platform or distributed between multiple PCs to
support configuration and dynamic checkout of the controls and interface
associated with a process control system.
This approach makes it easier to update the hardware platform since
migration issues are minimized. This capability is used to allow portions
of the controller software to be executed in the workstation. Internally, this
software capability was utilized by control-system manufacturers to facilitate checkout of new controller code. By installing the controller software
on a PC, the superior debugging tools provided on the PC can be used in
troubleshooting the software. Also, the capability of installing the controller software on a PC has been used to create internal demonstration systems where it was not practical to have controller hardware. Also, some
control-system manufacturers have realized that this capability to run all
the control software on a single PC would be valuable to customers in configuration development, simulation and operator training. Thus, as illustrated in Figure 10-12, some manufactures offer this capability as a
standard feature that is purchased independent of the control system
hardware. This design approach allows the controller software, Fieldbus

404

Advanced Control Unleashed

software, and workstation software to run on a single PC or in a multinode environment as shown in Figure 10-12.

Figure 10-12. A Complete Training and Testing Center

When a control system utilizes the Fieldbus Foundation function block

architecture, the function blocks used to define control, calculation and
measurements may be distributed in a variety of ways. For example, the
function blocks assigned to multiple controllers may be reassigned to a
single PC or to multiple PCs without modification or the need to reconfigure block parameters. Also, as defined in the Fieldbus standard, a SIMULATE parameter is available in each I / O function block. A person may
supply a value and status to the I / O function block for the field measurement when simulation is enabled, as illustrated in Figure 10-13.
This capability is used to check out control logic and operator displays.
Also, using simple tieback from output blocks to the simulation input values of input blocks, it is possible to provide simple process simulation for
operator training.
An OPC interface is provided in modern control systems to allow applications to access function-block information. Utilizing this capability, an
application may enable simulation in I / O blocks and write the simulated
value and status to be used in place of the field measurement. This capability, combined with the ability to redistribute control functionality without reconfiguration, provides a strong foundation for the integration of
high-fidelity simulation with the control system configuration in a single
or multi-PC environment.

Chapter 10 - Virtual Plant

405

Figure 10-13. Input/Output Function Blocks Support the Simulate Parameter

The major benefits realized from structuring the interface to the control
system around OPC and Fieldbus function blocks are:
Standard Component Object Model (COM) /OLE Automation
models, tools and languages (VB, C++, Delphi) are used to access
the server.
OPC interfaces provide a common protocol that reduces driver
effort by an estimated 50% to 80%.
OPC is Distributed COM (DCOM) based, so distributed capability
is built-in.
The OPC server of this control system abstracts the deployment of the
controller and Fieldbus-based control items from the OPC client. For
example, consider a system that employs a large number of Fieldbus
devices. In order to simulate such a system easily, the measurement and
control function blocks need to be reassigned to run on a workstation. If
the OPC server is set u p correctly, the OPC client does not need to know
that control execution has been reassigned. The same client configuration
connects to the actual scalable DCS or the control assigned to a PC with no
change in configuration.
The flexibility to extend a dynamic process-simulation tool is important
where proprietary process elements exist, including Unit Operations,
Reaction Kinetics, Property Packages, and Pressure-Drop Correlations.
With the OLE Automation Extensibility Interfaces supported by the simulation tool, extensions are easily written using standard OLE automation

406

Advanced Control Unleashed

tools and plugged seamlessly into the simulation engine process. Customers do not have to rely on the simulator manufacturer to provide the
required functionality.
The OLE automation interface, OPC browser interfaces, and tools based
on these technologies help customers flexibly configure large, integrated
simulation systems. Maps must be created to associate elements in the
control system with elements in the process-simulation system. This mapping is generated during the configuration phase of a project. Both visual
drag-and-drop and more automated programmatic generation features
are desirable for flexible map configuration.
While computing power continues to increase at phenomenal rates, the
computational problem of performing rigorous dynamic simulations for a
large plant will remain more than one machine can handle for some time.
The process models and control-system sections in a final application are
generally distributed, to some degree. Adopting hardware and software
interface standards such as OPC helps reduce the cost and development
times for operator training projects. Using OPC, it is possible to distribute
the virtual plants between multiple PC nodes as illustrated in Figure
10-14.

Figure 10-14. Distribution of Virtual Plant Among PCs

To allow control-system execution to be coordinated with the process simulation, command parameters are added to the PC nodes that execute the
control-system modules. Typically, a user interface is provided that allows
a process designer or operator training instructor to use these parameters

Chapter 10 - Virtual Plant

407

in setting up the simulation. Some common types of request that are supported by the virtual plant are:
Enable or disable I/O simulation for all blocks assigned to a PC
node.
All time-dependent blocks (dynamic blocks) contained in modules
assigned to the node are initialized to their steady-state condition
(based on input value).
Set target mode of output blocks to Manual or Auto so it is possible
to write initial values.
Set target mode of output blocks to Normal mode so normal
operation is achieved.
In addition, node parameters will typically be provided that allow the
control system to execute faster or slower than real time. This capability
reduces the time needed for control checkout and operator training.
Support is also normally provided for stopping the execution of all blocks
assigned to the PC with a single request.
The instructor interface to the virtual plant may allow all modules
assigned to a node to be examined. Modules that are not correctly set up
for simulation are automatically identified. The blockmodes and simulate
parameters are independently changed. The module summary indicates if
modules are correctly set up for simulation. The module detail view indicates abnormal conditions for simulation. Block Mode, Simulate, and OUT
values are changed and Simulation is enabled or disabled with a single
request.
To allow the control simulation to be coordinated with process simulation,
these node parameters are accessed by the process-simulation application,
as illustrated in Figure 10-15.
A PC node parameter may also be used to specify an execution factor. The
modules in the node run faster or slower than their configured execution
rate based on the factor. However, time-dependent calculations in the
block still use the configured execution rate in their calculation. Thus, true
faster or slower than real time response is exactly provided.
Typically, the virtual plant may support speed factors from 1/30 to 30
times real time. How fast the virtual plant can run depends upon the distribution of the models among multiple PCs, the speed of the PC processors, and the complexity of the process models.

408

Advanced Control Unleashed

Figure 10-15. Coordination of Simulation Using PC Node Parameters

References
1.

Mansy, Michael, M., McMillan, Gregory, K., and Sowell, Mark, S., "Step Into
the Virtual Plant," Chemical Engineering Progress, February, 2002, pp 56-61.

2.

Butler, Douglas, L., Cameron, Robert, A., Eckelman, Larry, D., and McMillan,
Gregory K., "Virtual Plants for Hands On Advanced Control," ISA, Houston,
September 10-13, 2001, Paper 1053.

3.

Lo, P., et. al., "Distillation Tower Pump Failure," Control, October, 2000, pp.
71-82.

4.

Stanley, Peter, "Dynamic Simulation for Insight," Chemical Processing,

December, 1999, pp. 47-50.

5.

Chin, K., "Learning in a Virtual World," Chemical Engineering, December,

2000, pp. 107-110.