The use of simulation software allows a more effective, less expensive approach to testing process automation system application software. 

The use of simulation software has been proven to save process industry users’ time and money on process automation projects.  There are a number of enabling technologies that support new approaches to application software acceptance testing, but the most significant is the use of process and IO dynamic simulation.

Traditional Factory Acceptance Tests

Traditional practice with Distributed Control Systems (DCS) was to test the performance and completeness of the control system by conducting a Factory Acceptance Test (FAT).  This test would be a significant portion of the project execution and delivery cycle.  In general, the tasks of a DCS FAT include the following:

• Assembly and staging of the complete control system in a large dedicated area including process controllers and IO, operator consoles, engineering consoles, cabinets and power systems.
• A burn-in period conducted to identify component “infant mortality” and other system failures.
• Testing of node-to-node communication, system availability, and diagnostics analysis.
• Testing of system application software through the IO subsystem using external switch panels or mirror control systems.

Problems with applying this approach to modern control systems include:

• Current global business environments demand faster project schedules.  The traditional practice takes considerable time, and more importantly, does not allow schedule compression by using parallel system development and testing activities.
• Modern automation systems are more modular and use Commercial Off-The-Shelf (COTS) computer hardware and network equipment. This allows system development in a more productive, less expensive, office environment.
• Testing application software using external switch panels or mirror control systems is insufficient to adequately test control strategies.
• Testing of the system configuration through the IO subsystem using external switch panels or mirror control systems is expensive and prone to human failure. This approach does not provide adequate infrastructure for operator training. Operator effectiveness and acceptance of the control system is possibly the second greatest area of risk in process automation projects.

Technology changes in the process automation industry have allowed significant improvements in the installed cost and performance of control systems installations. Significant improvements in the reliability and robustness of electronic components, due to integrated circuits and board design improvements have made system burn-in tests unnecessary. COTS network, PC technology, and integrated databases using object technology have improved the available bandwidth and performance of system workstations and networks so that node-to-node communication, system availability, and system diagnostics testing can be more easily and cost-effectively accomplished on site during system commissioning.

A Better Approach to System Testing and Acceptance

Modern control system technology allows a more cost-effective approach to system hardware testing, control system configuration SAT, and system commissioning.  This approach allows reductions in project schedule and more effective control system testing through the following means:

• Concurrent System Acceptance– System hardware testing, acceptance, and system configuration SAT can be accomplished as parallel independent tasks instead of serial tasks.  This is accomplished by the use of Process Simulation Software for the SAT portion of the task and conducting the hardware testing in stages. (See the project schedule in Figure 1 below.)

Automation Project Schedule Using Concurrent Tasks

• Concurrent System Commissioning– System commissioning, Installation Qualification (IQ) and system configuration SAT can be accomplished as parallel, independent tasks instead of as serial tasks.  (see project schedule figure 1)
• Comprehensive Software Acceptance Testing– Control system configuration can be thoroughly tested and analyzed using Simulation Software resulting in no process downtime due to control system configuration performance. This testing can include advanced control, batch and sequence control, interlock actions,
• Manufacturing Execution System (MES) and Process Information Management System (PIMS) integration, as well as simple verification of the operator graphics and system configuration database.
• Comprehensive Operator Training– Operators can be thoroughly trained using the same Simulation Software used in the SAT, resulting in no process upsets due to operator error.  This training can include operations during steady-state as well as upset conditions, process startup and shutdowns, and batch and sequence operations.

Simulation Software for Testing and Training

The use of Simulation Software for Automation System Software Acceptance Testing (SAT) allows a more comprehensive approach to Automation System Software Acceptance Testing.  The use of dynamic, easily developed process models allows the user to uncover more errors and eliminate risk from the application software. By using a dynamic process model the application software can be tested through its complete range of use. Tests that are too dangerous or risky to test on a live process can be accomplished in the safe off-line environment. Studies have shown that the cost of identifying and correcting system errors or validating application software in the simulation off-line environment is 10 to 100 times less than in the on-line plant environment.

Better, more comprehensive, testing can be done in the simulation off-line environment for less time and money. The operations manager can use the off-line simulation system to reduce the three greatest areas of risk in any process automation project. These areas of risk are:

• Application Software that has hidden errors due to implementation mistakes of design flaws.
• Operator actions and reactions to the process that are either inappropriate or insufficient.
• Operating procedures that have errors or are incomplete to address the process situations.

The automation system application software performance and effectiveness is the most critical part of the control system implementation.  The best engineered control networks, IO systems, and control room layouts will have no effect if the control system application software is poorly designed or implemented.  Under performing application software is one of the leading causes of control system startup delays, unscheduled process downtime, and off-spec or re-worked product.  Conducting a thorough SAT of the application software may be the most critical task to ensure the automation project is a success.  Simulation technology must be used in order to conduct a thorough testing of the application software, however this simulation technology must meet several requirements in order to be used successfully for automation system SAT.

SimulationTechnology for Automation System SAT

Non-intrusive Process Simulation

Non-intrusive simulation technology allows the user to test the control system configuration without making any modifications to the configuration database including conversion of function blocks or re-assigning control functions to run in other control system nodes.  A non-intrusive approach allows testing and verification of the control system. This non-intrusive approach also permits plants to move the tested configuration to the live system without retesting and re-verification.  A non-intrusive approach implies that the simulation is foreign to the control system configuration and does not require modifications to the application software.This eliminates the requirement to re-test and re-verify the control system configuration before downloads.

Simulation through the IO Subsystem Including Digital Busses

Simulation through the IO subsystem allows the user to test the control system configuration through the IO assignments and configuration.  This requirement becomes important when using a distributed processing digital protocol like Foundation Fieldbus, which supports distribution of process control function blocks to field devices.  If the process simulation is to be used to eliminate errors in the control system configuration, a comprehensive IO Subsystem must be simulated with the process models.

Cost Effective Development of Process Simulation Models

Effective Control System SAT require process models that accurately model mass-balance and heat-balance of the process unit.  In addition field devices such as motors, valves, heat exchangers, etc., must be accurately and cost-effectively modeled.  The development of an adequate process model must be possible even with the most compressed project schedule.

Flexible Simulation Data Visualization Tools

Many tasks that are performed thorough SAT are repeated for each element of configuration.  Significant testing can be saved by the use of flexible data visualization tools that can be customized to the needs of the project and the task.  Depending upon the testing function and requirement, the user may need dynamic tabular data views, process trends, or dynamic process flow diagrams.  Configurable, easy-to-use, data visualization tools can speed up the SAT process.

Using MiMiC Simulation Software for Comprehensive Control System SAT

Users of modern control system architecture have a process simulation technology solution that supports all the requirements listed above for comprehensive control system SAT.  MiMiC Simulation Software is a 3rd generation simulation software platform based upon the latest in open software development standards and computer technology.

Non-Intrusive Process Simulation

MiMiC Simulation Software provides a completely non-intrusive process modeling approach to control system SAT.  The DeltaV controllers are connected through their IO bus to a Virtual IO Module that emulates the IO cards and devices of the controller. The Virtual IO Module provides all the IO and diagnostic messaging required by the controllers to run the control algorithms without modifying the control system configuration. This virtual IO subsystem supports controller/IO module communication including auto-sensing of IO cards and devices and diagnostic functions. DeltaV Diagnostics for a controller with “virtual” IO shows all cards are in GOOD status. Good diagnostics ensures that control modules run in their TARGET modes and do not behave in process upset conditions during normal operation.

DeltaV Software Acceptance Testing Architecture using MiMiC Simulation Software

This approach to process simulation allows the user to download his control system configuration to the controller before SAT, enable MiMiC Simulation Software, and begin testing without any modifications to the control system configuration.  At the end of SAT, when all configuration corrections have been made and tested, the user can download the same control system configuration to the live system without any modifications and re-testing. The controller now begins to communicate to the live IO and process in the same manner as the virtual IO and process.

All process and IO system models run foreign to the control system configuration as simulation objects. Process device tiebacks are modeled in the simulation software and tied into the control system configuration through the Virtual IO Module.The simple tieback models generated by MiMiC Utilities can be enhanced quickly and easily in the Simulation Studio application.

Simple Tieback Model in MiMiC Simulation Software

Simulation through the IO Subsystem Including Digital Busses

MiMiC Simulation Software provides a method to test the control system configuration through the IO subsystem of the controller. In the case of traditional IO, this ensures the IO assignments are correct in the control system configuration.  IO assignments can be verified using a Dynamic Tabular view of the Simulated IO. The MiMiC Data View application is flexible and can be modified to meet the user’s project and testing requirement. 

If the process control system installation uses digital busses, such as Foundation Fieldbus, simulation through the IO subsystem is mandatory. Foundation Fieldbus provides the capability of distributed function block processing including PID control in the field devices. No other effective means to test the control system configuration exists.  Process simulation provides a simulation interface for digital bus IO cards and devices that emulates all the necessary messaging for the controller.

With MiMiC the emulation of Foundation Fieldbus devices is complete and supports all device states.  The simulation also supports the assignment of function blocks to field devices and dynamic simulation of their processing.

Simulated IO Definition Including Foundation Fieldbus Devices and Signals

The control system configuration for this loop and the rest of the process controller can be downloaded in the same manner as real IO and field devices.  The loop configuration runs in the simulated field devices. The same process modeling objects, used for traditional IO simulation, are used for the digital busses.  This method provides a comprehensive approach to testing control systems that use digital busses.

Cost Effective Development of Process Simulation Models

MiMiC Simulation Software provides an automated utility for generating base-level simulation of the control system.   Users of PlantWeb architecture can run the utility on the DeltaV database export file (.fhx).  The utility generates the Simulated IO (including digital buss device definition), and base level IO, discrete and analog tieback models.  The simulation that results from running this utility is more comprehensive than a panel board or a controller configuration tieback solution.  In addition, it does not require IO and device hardware and is completely non-intrusive to the control system configuration.  If the user or project requires a higher resolution process simulation, this base level simulation is easily enhanced using the Simulation Studio application.

Simulation Studio provides a graphical user interface for the development of process models.  Simulation is built using the extensive library of IO, modeling, and logical function blocks. The modeling blocks follow the IEC1131for ease-of-use and intuitive model building. There are no restrictions on how the user can combine blocks into comprehensive process models. Models can be viewed in on-line execution or off-line construction mode.  Moving between model construction and model execution viewing is as easy as clicking an icon on the Simulation Studio tool bar.

Complex, dynamic process models are easily developed using first-principles modeling blocks like the Heat Exchanger function, Steam Properties function, Thermodynamic Properties function, and IEC1131 structured text Calc function.The Simulation Studio graphical user interface and simulation functions provide a user environment that allows a typical process control engineer to develop a mass-balance, heat-balance simulation in less than one week using for each process controller.

Flexible Simulation Data Visualization Tools

In the course of testing the system configuration, flexible, graphical user interfaces can greatly reduce testing time and increase testing accuracy. MiMiC Simulation Software provides a comprehensive suite of configurable user views that support system testing and increase the efficiency of the testing engineer.

MiMiC Data MView application provides user defined views of model and Simulate IO data in a dynamic table format. MView is launched from the MiMiC Explorer. One or more views can be created in the MiMiC Explorer, opened and configured in the MView application, and placed online to show the runtime model data in tabular format. Each MView can have multiple sub-views, which appear as child windows inside an open MView application window. Each sub-view can display a different aspect of the running simulation. The user can toggle individual discrete channels, set values of analog channels, and ramp analog channels over a specific time and range. This application supports digital busses such as classic discrete and analog IO.

MiMiC’s Data Monitor application allows the user to build trends of real-time data from MiMiC models and any other OPC Server source.  This is a helpful tool for troubleshooting and tuning simulations as well as predicting what will happen in the process under certain conditions.  Because process models in MiMiC have many variables that never are reported to the control system, process information that is unavailable to the operator can be monitored and trended.

Simulation Data Trends in MiMiC Data Monitor

Finally, Component Studio in MiMiC is a dynamic process flow diagram environment for building graphical representations of the process for instructor stations or troubleshooting process models.  Graphical elements can be linked to real-time data from MiMiC models or any other OPC server source.

Graphical Views of Process Models in MiMiC Component Studio

The configuration and design of all three types of views are saved in the MiMiC database.  They can be imported and exported from one MiMiC system to another.  User views allow complete flexibility with how the user can work with his simulation system.

Comprehensive Operator Training

One of the benefits of using a process simulation for the Software Acceptance Testing of a process automation project is that the same simulation now can be used to conduct comprehensive operator training and certification.  The operators can be trained and certified on the entire scope of the process automation project and get runtime experience running the plant and operating the control system.  This training can be done without the potential loss of product or process incident due to operator action.  Also, the operators can use their experience to critique and test the process automation solution.

Operator Training has several additional needs in addition to the process simulation requirements. First, Operator Training sessions should be conducted in a structured, repeatable manner.  Training Scenarios are defined sets of process events and malfunctions that provide consistent operator training sessions. Scenarios can be simple failures or complex sets of process events, time delays, and equipment failures, depending upon the process and training requirements. Scenarios can be initiated by the trainer or automatically by a process event.

Training systems should allow the user to develop expected results or scoring conditions for each malfunction or process event. The training system should have the ability to set scoring conditions based upon expected results in the simulation or the off-line control system. The system should dynamically calculate the results of the training session with an expected score and actual score.

In order to meet regulatory agency requirements, the results of the operator training sessions must be documented. Training Session report should be automatically in open document formats such as HTML, RTF, and PDF formats. This report should contain any operator actions as well as process alarms and events.

Operator Training Scenario in MiMiC Operator Training Manager

Advanced Scenario training may require the trainer to start the operator at specific process and controller conditions.  In order to meet this objective, the training system must have the ability to take snapshots of the process and controller, save the snapshots, and restore the process and controller to that snapshot condition.  Snapshots should be saved in such a manner that the trainer can select them in the same manner that any other file or document is selected.

The result of conducting comprehensive operator training is safer and more efficient plant operations. Operator training programs increase the effectiveness of the operations staff, resulting in better finished product, lower operating costs, and less process downtime.

Conclusion

In summary, the use of MiMiC Simulation Software allows users of modern control technology, a new and more cost-effective approach to system hardware testing, control system configuration SAT, and system commissioning.  The use of MiMiC Simulation Software allows the user of these systems to move from a DCS model of Factory Acceptance Testing to a new approach, with significant reductions in project schedule and improvements in control system performance.  This new approach combined with Process Simulation Software allows the following methods:

• Concurrent System Acceptance.
• Concurrent System Commissioning
• Comprehensive Software Acceptance Testing
• Comprehensive Operator Training

The use of this new approach to process automation system testing using MiMiC Simulation Software will save process industry users’ time and money.

Financial Justification of Process Simulation Systems

To follow is a summary of costs and project returns on the use of MiMiC Simulation Software for process automation solution SAT.

Control System Size and Scope:

Batch Process
2 Process Controllers
Medium Fidelity Simulation with mass-balance, heat-balance
Sale value of One Batch, Grade A Product - $120,000
Sale value of One Batch, Grade B Product - $80,000
Batch Cycle – 24 hours
Cost of unscheduled downtime - $5,000 / hour

Process Simulation Solution Costs

Software License - $18,000
Interface Hardware, PC - $5,000
Simulation Development Costs - $10,000
Total Cost for Simulation System Development - $33,000

Benefits of Using Process Simulation for Software Acceptance Testing

One-Time Savings, First Batch in Startup Grade A instead of Grade B - $40,000
One-Time Savings, Reduction in Startup Schedule by One Day - $120,000
Annual Savings, Reduction of Off-Spec Product (Grade B) by One Batch per Month – $480,000
Annual Savings, Reduction of Unscheduled downtime from 1 day per month to 1 day per quarter due to operator training - $960,000