GUIDELINES USEFUL FOR INCORPORATING DCS IN EXISTING PROCESS PLANT

Craig E. Wentworth
T.H. Russell Co.
Tulsa

Proper design, installation, and start-up of a distributed control system (DCS) as part of the overall automation of a process facility require consideration of several factors.

Here are some fundamental guidelines to ensure the project's success.

JUSTIFICATION

The first step in justifying the installation of a system is to set obtainable and verifiable goals for it.

Some typical goals are:

• Improving plant production through improved process control.

  The plant's existing instrumentation may have deteriorated so much that simply replacing the instrumentation will improve performance and production.

  Also, advanced control applications in some facilities can greatly improve production. And, accuracy over existing process measurements may have improved.

  It must be emphasized that these types of improvements are difficult to estimate accurately. Caution should be exercised when revenues from these goals are used for justification.

• Reducing operating cost through the combination of operational tasks.

  A study of the present operational manpower must be made to determine if the DCS will eliminate any operational tasks in the facility.

  The DCS will provide a means to monitor and control the facility from a central location. If the plant under consideration currently has local instrumentation, it is possible that tasks such as taking local process readings and monitoring local controllers would be combined into the DCS terminal operators' duties.

  Another task that could be eliminated would be the retrieving and organizing of the plant's process data.

  The DCS is capable of producing reports that provide the needed process data and operational information. This includes shift, daily, or weekly operational reports.

  There should also be a reduction in maintenance tasks over pneumatic instrumentation. Electronic instrumentation will require less maintenance and, when required, maintenance will be simpler.

• Providing single-source data acquisition for engineering and operations.

  The advantages of having a single-source data base for process information are frequently overlooked.

  The DCS will provide engineering and operations with a single data base to retrieve valuable process data. This allows the plant's operational data to be quickly and more thoroughly viewed, resulting in more prompt operational decisions based on current process data.

  This capability can save a significant amount of time in operational decisions being made, thereby saving on operating costs. In troubleshooting of operational problems, having immediate access to current and historical process data can save facility down time and significantly reduce the problem-solving time.

  It also provides operations the ability to convey accurate plant process, alarm, and shutdown information to engineering and management.

  The DCS single-source data base can also assist in handling information for the new U.S. Occupational Safety and Health Administration (OSHA) process-safety management (PSM) requirements.

  When properly programmed and designed, the system can be a data source for operational personnel in dealing with the plant's PSM procedures.

INSTALLATION

Typically, DCS hardware costs are competitive and closely monitored. Sometimes overlooked is the actual installation design which will affect the hardware selected and can significantly affect the overall project cost.

The primary way to reduce field-installation cost is to reduce the amount of field-installation material, such as wire and conduit. This can be accomplished by locating the DCS input and output (I/O) modules near the field instrumentation.

Most of the major DCS manufacturers can remotely mount the 1/0 devices, but consideration must be given to what is required to install the devices in the plant's classified areas (Fig. 1).

A majority of the facilities in the gas processing and refining industry deal with Class 1, Division II, Group D areas, under the National Electric Code published by the National Fire Protection Association. This means that the remotely mounted I/O modules must be able to meet the area classification.

One way of doing so is by air purging or modifying the DCS 1/0 enclosures which will have the following effects on the system's installation and operating costs:

• Additional instruments associated with the enclosure for air purging

• Higher utility costs and use of instrument air

• Additional installation labor and material.

For these reasons, evaluation of the DCS should include an investigation of whether the manufacturer can provide remote 1/0 modules that can be installed in Class 1, Division II, Group D areas without requiring enclosure purging or conditioning.

Some major manufacturers currently have this capability, and others are working towards it for the near future.

Remotely mounting the DCS 1/0 modules reduces field-installation costs by reducing wire, conduit, and fittings that must be purchased and installed, the control-building space and cable entry requirements, and the termination material and labor costs compared with the use of marshalling cabinets (Fig. 2).

The advantages become even more significant if the plant's automation and upgrade additions call for skid-mounted process equipment.

With the use of remotely mounted 1/0 modules, the enclosures can be installed on the skid at the fabrication shop, and all on-skid instrumentation prewired. Upon arrival at the facility, the skids require only power and data-highway installation (Fig. 3).

This approach will result in significant savings.

CONFIGURATION

Engineering and configuring the DCS project is usually the most interesting part of the project but can be the most difficult. It is imperative that a plan be mapped out by the facility's engineering and operational personnel.

Following are some basic questions that must be answered by the plant's personnel before the DCS is configured:

FORMAT

Mat will be the basic operational format used in the DCS?

In other words, how does the operator monitor the plant: through mechanical flow sheet graphics? Group controller displays? General overviews?

The question should be answered with the help of the operational personnel. If the operators feel uncomfortable with the format, they will resist the DCS.

The format must be one in which DCS display-to-display movement through the plant can be easily performed and follows the basic operation of the facility.

The most commonly used format is the facility's mechanical flow sheets. These represent the flow through the plant and also give a reference for all controllers and other instrumentation.

If the mechanical flow sheet graphics are to be used, an evaluation should be performed to determine exactly what is to be taken from the flow sheets and configured into the graphics.

Usually, some information on the flow sheets is not required on the system graphic displays. Unnecessary lines, symbols, and information can make the graphics "busy" and cause confusion.

Modify the mechanical flow sheets to represent the information as it is to appear on the system screens. If certain process information is to be displayed on the graphic screens, show it on the mechanical flow sheet that the configuration engineer will use to build the screen graphics.

If different colors are desired for the process lines for different service, indicate appropriately. Performing these tasks before configuration will save time and money.

ACCESS

How will operators access the process controllers?

Several options are generally available with all distributed control systems concerning how process controllers can be accessed from the operational graphic.

Three of the most commonly used are:

• Overview displays: A symbol for the controller is picked from the operational graphic, and a controller faceplate is displayed over the graphic.

• Overview display with trends: Same as the previous but the controller will have a trend display included.

• Controller group displays: A single controller symbol is picked, and a group of control faceplates is displayed.

Other options and methods are as effective and can be easily configured into the DCS. The method chosen should emphasize easy access to all controllers.

Again, this is an item that must have the involvement of the plant's operational personnel.

PROCESS ALARMS

How are process alarms handled?

Alarms can be displayed by the system in several ways. Key to displaying any alarm is that the operator know exactly what the alarm means and what equipment is involved.

Display of alarm tables is very useful for time and date stamping but has little visual effect.

It is more visually effective to display the alarm on the operational graphic (the mechanical flow sheet display). This method allows the alarm to be visually associated with the equipment.

The alarm acknowledgment can be performed in a number of ways. Typically, it is designed so that the operator must go to the actual controller faceplate display or alarm page to acknowledge an alarm.

Acknowledgment should involve association with the actual device in alarm.

PARAMETERS

What are the parameters for process controllers and indicators?

Establishing parameters for process controllers and indicators requires the most time.

A detailed data sheet for each process controller and indicator must be generated. This data sheet should contain as a minimum the following information:

• Controller or indicator tag and description

• Measurement signal type: 4 to 20 ma, 1 to 5 y, thermocouple type, and similar data

• Measurement range

• Measurement engineering units: psig, gpm, %, etc.

• Controller action: direct or reverse

• Control-valve failure position: fail open or fail closed

• Self-tuning: yes or no

• PID, PI, or PD controller

• Initial proportional, integral, and deviated settings

• Default setpoint value 9 Is the controller cascaded or interfaced with other controllers? If so, how?

• Are any alarms associated with the controller or indicator? If so, what type and settings?

• A complete description of any advance control schemes.

Describe exactly how the controller is to interface with other controllers or variables.

If it is to be a multivariable control algorithm, there should be a complete and thorough explanation of the algorithm and the required results. If a process analyzer is a variable, the timing and sequencing of measurements and events must be provided to the configuration engineer.

Additional information may be required depending on the controller and the system, but the previously listed information should be provided as a minimum.

SHUTDOWNS

How are process shutdowns to affect the plant?

The most efficient way to provide information on the effect on the plant of process shutdowns is by developing a cause-and-effect report on the facility.

The DCS configuration engineer will be able to take the cause-and-effect report and develop the plant's shutdown logic. This logic may be developed in either ladder logic or Boolean Algebra form, depending on the system.

A thorough review of the logic should be made by the facility's engineering personnel to verify proper shutdown action.

Any DCS operator interaction required during a shutdown should be included with the cause-and-effect report given to the configuration engineer. This can be in the form of verbal explanations of what is required during and after a shutdown as answers to the following questions:

• What action is the DCS operator to take when a shutdown occurs?

• After a shutdown, are there devices that need to be manually reset by the DCS operator?

• Is it required that explanation text be displayed to assist the DCS operator during a plant shutdown or start-up?

EQUIPMENT CONTROL

Is the motor-driven equipment in the plant controlled from the DCS?

This question addresses more than the shutdown functions.

Are the pumps, fans, and other motor-driven equipment to be turned "on" or "off" by the DCS operator? Answering this question involves investigating the company's operating philosophy.

If the equipment is to be started and stopped by the DCS operator, other considerations include the following:

• The DCS and field operators must be trained on the safety procedures of remotely starting or stopping equipment.

• The company should develop a standard procedure for both DCS and field operators. This should be a verification procedure for all equipment starts or stops.

  A reliable and safe base radio transmission system must be employed to allow the DCS operator to communicate with the field operators. Care should be taken before purchasing the radio system to ensure that it will not exceed the DCS specifications for radio transmission interference.

  Also, ensure that the hand-held field radios meet all area classification requirements.

• The motor-driven equipment must be equipped with local "Hand-Off-Auto" control stations.

These local control stations should be designed to override any command from the DCS. In other words, if the field operator places the local "Hand-Off-Auto" station in the "Off" position, the equipment cannot be operated by the DCS.

The "Auto" position is the only one in which the DCS can control the equipment.

It should also be decided how the equipment's "Start-Stop" or "On-Off" operation will be performed from the DCS.

It is not recommended that the actual "On-Off" or "Start-Stop" control buttons be continually displayed on the screen. This allows a situation to exist in which equipment can be inadvertently turned "On" or "Off" by an invalid "pick" on the screen.

It is recommended that the equipment's "On-Off" or "Start-Stop" control require at least two operator commands. This decreases the likelihood of an invalid command.

Experience indicates that having the equipment's symbol "pickable" from the mechanical flow sheet graphic, and then the "Start-Stop" control appearing, works well.

Depending on the complexity of the system requirements, the amount of information required by the configuration engineer can greatly increase.

The key to a successful system configuration is to provide this information as completely as possible to the configuration engineers early in the project.

TRAINING

Before the system is applied to the actual control of the facility, it is imperative that the operational personnel be adequately trained on the DCS.

This should be started as soon as possible in the project. The more time the operators have to learn and adapt to the system, the easier will be the actual start-up and commissioning of the system.

How the operator training is provided can determine the success of the training for the operators.

Often, operators are sent to a system software-configuration course. Although it is important that someone from the facility be trained to perform system configuration, it is not recommended or needed for the plant operators.

Typically, these courses are in-depth studies into the system's software configuration and generally address no actual tasks involved in operating the plant with the DCS.

One or two people from the facility should attend the manufacturer's software configuration training courses. Personnel with a basic understanding of personal computers (PCs) excel in these training courses and seem to understand the system configuration concepts quickly.

How the plant is operated by the DCS depends on the specific configuration for the facility. Therefore, the operator training should be performed by or involve both the facility personnel involved with the DCS project and the configuration engineering personnel.

Operator training should include the following:

• A general explanation of the system hardware

• A definition of terms, associated with the DCS

• An explanation of the operational platform (the mechanical flow sheet graphics)

• How controllers and indicators are accessed

• How alarms are displayed and acknowledged

• How shutdown and start-up procedures are to be performed

• How to start and stop equipment with the system

• How to recognize system malfunctions.

Training must be developed to meet the facility's specific DCS requirements.

A general knowledge of the system is helpful, but that alone is insufficient for the operators. They must be trained on the specific configuration of their facility's DCS.

Typically, a 2-3 day course at the facility, with the DCS, is effective.

Another effective training tool is a simulation system. Having at least one DCS operator station in service will allow the operators to get hands-on experience with the system.

It will require some effort on the part of the system integrator before the operator training to ensure that the system's configuration is near completion and that the necessary hardware is on site to set up the simulation station.

POWERING; SPARE PARTS

Once all the system's hardware is on site and installed, power up the system as soon as possible.

System component failures are more likely during the initial powering-UP, or shortly after. Therefore, the more time power is applied to the hardware before actual start-up makes component failure at a critical point during start-up less likely.

Before start-up, the system must be completely checked out from each I/O point to the end device.

As calibrations are being performed on the end devices, verification should be made through the DCS. All shutdowns and alarms must be checked and verified.

Involving the operators and the facility maintenance personnel in the system checkout will provide learning experience and build confidence.

Another key item before start-up is system spare parts.

It is advisable to have on hand during the start-up some spare parts for the system. Most systems are designed with redundancy for the critical hardware, but not all have redundancy down to the 1/0 component level.

It is recommended as a minimum, therefore, that spare 1/0 modules or cards be on site.

OPERATION

Once the DCS is in operation, close observation should be made of all the graphics, controllers, alarms, shutdowns, and displays that are configured,

The operators should begin a punch list of all detected system errors or desired changes. This list will be used by the configuration engineer to make the necessary corrections.

This type of system-correction procedure can save configuration-engineering time and money.

Application of report packages in the DCS should be carefully considered.

Decide what is needed on the reports and at what intervals. This information should be provided to the configuration engineer in a format displaying exactly how the report is to look.

These reports are usually configured after the system has been in operation for some time. This allows the operational personnel time to adapt to the system and become familiar with what the system can produce.

DCS also offers application of advance control operations, but consideration should be given to the timing of implementation of advance control schemes. Depending on the advance control application and algorithms, it may be advantageous to delay implementation until the operators have more time to become familiar with the system.

If the advance control schemes are as straight-forward as, for example, a simple feed-forward loop or cascade control, it may present no problem.

But if multiple variable applications are to be implemented involving chromato-graphs or other process analyzers, it may be advisable to delay this portion of the systems application for a short time.

Once the operators feel comfortable with the system, it will be much easier to implement the more difficult control schemes.

Copyright 1993 Oil & Gas Journal. All Rights Reserved.