Presented to:
Instrument Society of America Powid '90
33rd Annual Power Industry Division Symposium
Toronto, Ontario, May 21-23, 1990

Achieving Maximum Benefits in Applying a DCS
to a Circulating Fluidized Combustion (CFC)
System Power Plant

Gordon L. Johnson
Engineering Supervisor
Bechtel Corporation
50 Beale Street
San Fransisco, CA 94419
Robert L. Chappel
Vice President
Applied Control Systems, Inc.
426 S. Main Street
Pittsburgh, PA 15220
Richard J. Velan
Manager
Cogeneration Control Systems
Leeds & Northrup Company
426 S. Main Street
Pittsburgh, PA 15220


INTRODUCTION
The Distributed Control System (DCS) purchased to control and monitor the circulating fluidized combustion (CFC) system and related auxiliaries has the potential to start-up, shutdown, control and monitor many of the subsystems required for a small (30 to 50mw) coal fired cogeneration or waste fuel fired plant.

These subsystems would include coal handling, ash handling, particulate removal, water treatment and steam condensing systems. Other systems unique to a specific plant such as a thermally enhanced oil recovery (TEOR) system, dust collecting and disposal system or an air cooled steam condenser could also be done in the DCS. Segments of the mainstream turbine control, other than speed control, could also be included.

ADVANTAGES OF THE INTEGRATED APPROACH
The overriding reason for integration is lowest overall cost. Many of these small plants have to pare costs to a minimum to make them feasible. The control and monitoring equipment is one place where it is possible to reduce costs without impairing plant performance if the decision is made early in the project to use this approach.

Another benefit is reduced control room size as space for dedicated control switches or pushbuttons and read out devices (indicators, recorders, lights, annunciator windows or CRT'S) will not be required. Not having to provide space for the free-standing control panels used in the unitized approach will significantly reduce the size of the control room. The size, however, should not be shrunk to an absolute minimum and approach the cockpit of a 747, as it complicates maintenance and does not allow effective use of additional personnel during start-up.

Overall maintenance costs are significantly reduced as only one vendor's hardware is involved for the logic part of the system. This minimizes training expenses and spare parts inventory. Additional savings could be realized if the transmitter and final actuator sections were standardized early in the project and were purchased from the same vendor.

Ease of operation is an inherent benefit as the operator is not required to learn various methods of transferring control from manual to automatic and vice versa.

DISADVANTAGES OF THE INTEGRATED APPROACH

To achieve lowest overall cost, specifications for the subsystems must be written so that the vendors understand exactly what they must furnish and the extent of their responsibility for the successful control and monitoring of their system. As these subsystems are usually ordered early in a project without adequate input from control system personnel, completely configuring the DCS in time for scheduled shipment cannot be done due to inadequate and unclear information from the subsystem vendors.

During the initial start-up of the plant, added cooperation is required from the start-up personnel of all the vendors, due to limited access to the devices to start, stop or modulate their equipment.

Equitable allotment of costs to add additional hardware or functions or to modify control strategy for various vendors after shipment of the DCS requires the understanding of all parties concerned of the true costs of such changes.

The distributed control system is made up of hardware modules that contain a certain amount of memory. This memory can be either in the form of control functions or in some systems, free from programming. When this memory is exhausted another module must be purchased. Some distributed control DCS systems available today allow the use of a programming language which means that many more functions can be packed into a given module before another module is required. This can cause problems since the major cost of the DCS contract is for hardware and a smaller portion is for programming. As changes are made through the course of the project, added configuration costs result since the purchase of additional hardware modules may not be required. The true cost of both hardware and software changes should be known and mutually agreed upon when the contract is placed with the DCS supplier. The notion that configuration changes are "free" must be dispensed with immediately.

Another disadvantage to the integrated approach is that the module size in the DCS system containing the software functions may be overly large for each particular subsystem. This means that there is a potential for purchasing too much hardware if an entire module is assigned to each subsystem. Again, a high degree of early cooperation with the DCS vendor should minimize the amount of hardware required to integrate the functions of the subsystem.

VERSIONS OF THE INTEGRATED APPROACH
The first method is to have all of the DCS hardware purchased by the owner. Capability to add all the functions required by the subsystem vendors could be bought initially or added in segments at agreed upon prices.

The second method is to have the owner buy the hardware only for the CFC system and miscellaneous equipment not part of any vendor's system. Specifications for the subsystems would require the bidders to obtain firm pricing from a small selected group of DCS suppliers for added hardware and configuration for the functions necessary to start-up, shutdown, control and monitor the bidder's system. Once the DCS supplier is selected, the appropriate cost is known for final evaluation of subsystem bids if time allows. Obviously the DCS supplier must be selected early in the project to obtain the greatest benefit from this approach and also simplify the procedures.

The first method is the simplest but not necessarily the least costly, nor does it usually clearly define the complete responsibility for the success of the control system. The owner's engineers tend to become involved in the control strategy and in reviewing the functional block diagrams (FBDS) and the logic diagrams which can dilute the vendors responsibility to provide a completely workable system.

The second method requires a section in each subsystem specification which clearly defines all of the information required from the vendors to configure the DCS. This includes not only FBDS and logic diagrams but complete system descriptions and sections of the operating manual which explain thoroughly the objectives of the control and monitoring system. A schedule for submitting final information and documentation must be included with stated penalties for non-compliance.

INGREDIENTS FOR SUCCESS
The success of either method depends on total cooperation between the owner's engineers and the DCS vendor. The ability to obtain total cooperation and a reasonable "comfort factor" should be a major component in the evaluation of the DCS vendor. Not only must there be "comfort" in the vendor's technical and commercial offering, but also, the vendors track record must be taken into account regarding delivery schedules, contract change costs, and satisfied customers on past projects.

The first method requires reaching an equitable pricing structure for all the added hardware and configuration costs necessary for the subsystems.

The second method requires the DCS vendor(s) to submit firm prices to subsystem bidders to supply configured hardware to perform all of the required functions-delineated by the bidder.

For either method, the subsystem vendors should have a basic knowledge of the capabilities of a DCS in order to reduce overall costs. In particular, they should know what specific information they must submit to the DCS vendor in order for him to efficiently configure the system to their needs. They should also realize that the overall control strategy must be finalized early on and not be subject to continual revisions simply because it is so easy to change the DCS as compared to the old analog systems and hard-wired relays.

For the second method, a formula must be developed to equitably allocate the initial start-up charges for the DCS vendors site engineer between the owner and the various subsystem suppliers. Costs for changes initiated after the initial start-up must similarly be allocated.

PROTECTIVE INTERLOCKS
A DCS usually has the ability to provide all of the protective interlocks required for this type of plant at the least cost. However, each interlock should be reviewed to determine whether it should be included in the DCS or independently wired for greater protection. An example for the latter would be the interlocks for the turbine lube oil pumps, both AC and DC powered.

The furnace protection system (FPS) for the start-up burners can be done successfully in the DCS in accordance with the applicable NFPA 85 standard for multiple burners particularly if the energize to trip concept is used. Critical logic functions must have maximum security so that inadvertent changes to it cannot occur.

One method of providing this maximum security is the use of redundant hardware in a "one for one" backup scheme that totally backs up the furnace protection logic. An even more secure approach is to have that backed-up logic hard coded on proms so that the program cannot be changed inadvertently. Overall costs are reduced by requiring the boiler manufacturer to include his FPS in the distributed control system.

The typical sequence of events (SOE) function can be done in the DCS at significant cost savings and does not require a dedicated recorder. Some distributed control systems today contain the ability to do on4ine SOE. This requires that the distributed control system be able to scan at speeds of one millisecond, store and time tag all of the events that occur, and also that a trip report be generated from the distributed control system. This results in cost savings because additional SOE hardware and recording is not required.

OPERATOR INTERFACE
Using the integrated approach for the described plant requires additional information to be presented to the operator on the CRT's. This does not necessarily require that CRT's or operator stations be added if the basic design incorporated adequate hardwired recorders for critical variables such as total fuel flow, total air flow, feedwater flow, mainstream flow and temperature, drum water level, etc., which should be available to the operator at all times without any action on his part. A selectable trend recorder should be considered for those plants where integration is most extensive.

In addition, the incorporation of the subsystems does not require additional CRT's. This is one area where total cost savings can be realized. The main criteria for selection of number of CRT's is the number of operators that are necessary to startup the unit quickly after a unit trip. Each operator should have at least his own keyboard & CRT. Another important criteria is the data that is presented and the speed with which they are presented to the operator.

Careful selection of the distributed control system should include calculations of the point and screen update times under worse-case conditions.

Another DCS feature which is helpful to the operator is the incorporation of historical trending information. This information can be used by the operator as well as forming the basis for some of the report generation that is required.

Inclusion of the subsystems results not only in cost savings by eliminating all of the subsystems individual readout and control equipment usually located on free standing panels, but when these functions are included in the distributed control system they are much more "tightly coupled" to the system and therefore to the operator. Much better alarming capabilities as well as alarming speed are achieved. In addition to that, the trendability of variables in these subsystems is relatively simple. This results not only in reduced costs but also in a higher degree of operator comfort.

MAJOR PITFALLS
With any new approach there are pitfalls to be avoided. The major one is to insure that the subsystem vendors fully recognize that they are completely responsible for the success of their control and monitoring system regardless of who implements the functions they have deemed necessary. Care should be taken so that this understanding of responsibility is included in the subsystem supplier's contract.

A second problem which must be avoided is that of an adversarial relationship developing between the engineers and the DCS supplier because of the numerous changes which may be required due to misunderstandings during configuration. Regardless of how the specification on the contract is written, the integrated approach taken appreciable "give and take" to be successful in meeting the lowest cost objective.

With any new approach there are pitfalls to be avoided. The major one is to insure that the subsystem vendors fully recognize that they are completely responsible for the success of their control and monitoring system regardless of who implements the functions they have deemed necessary. Care should be taken so that this understanding of responsibility is included in the subsystem supplier's contract.

A second problem which must be avoided is that of an adversarial relationship developing between the engineers and the DCS supplier because of the numerous changes which may be required due to misunderstandings during configuration. Regardless of how the specification on the contract is written, the integrated approach taken appreciable "give and take" to be successful in meeting the lowest cost objective.

Another major consideration is that because the engineering is really on a quasi continual basis due to the short shipping schedules of the distributed control system, only one person should be responsible for configuring the system from beginning to end of the project. This person should be available at the DCS vendor's factory and also available in the field to make field changes that are continually occurring leading up to and during start-up. Since the start-up periods for CFB boilers are much shorter than the start-ups for major control stations, DCS vendors should be flexible in adhering to this requirement.

While we are recommending the inclusion of subsystem functions in the distributed control system, a significant exception to this recommendation is a separate system for continuous emissions monitoring. This should not be part of the distributed control system because of the specialized programming knowledge required in this application which includes report formatting, adherence to environmental codes, etc. By excluding the continuous emissions monitoring system from the DCS, lower total overall costs will result because of the specialized nature of this subsystem.

Another pitfall to be avoided is that because of the distributed control systems' ability to sense transmitter failures, dual transmitters are not required on all loops. They should only be included on critical loops, and care should be taken in system design so that these non-critical loops can be identified. Fewer transmitters are required, further reducing overall costs.

CONCLUSION
While either integrated approach has some risks, the potential cost savings and other benefits more than make up for the additional engineering effort and cooperation that is required when compared to a unitized approach.

An early decision to use this approach is an absolute must and has to be in place before any subsystems are specified if maximum savings are to be realized.


1990, Leeds & Northrup