Typical Problems
No steam balance control system
The consultant may define the equipment specifications and expect each supplier to provide the control concept for its own scope. In practice, the boiler supplier develops the boiler controls, the turbine supplier develops the turbine controls, and an older part of the mill may remain in operation as well. As a result, no one takes responsibility for integrating all parts into one system that operates safely, optimally, and energy-efficiently under all conditions. Without a complete steam balance control system, startup becomes more difficult, takes longer, and the mill’s profitability is harder to achieve.
The steam balance control system is incomplete
The control system must automatically handle process and equipment limits without exceeding them. When a limit is reached, the system should transition smoothly rather than switch abruptly
to another control component or operating mode. In an incomplete system, operator intervention is required whenever a limit is reached. An operator can never be as precise as a well-tuned
automated system, which leads to conservative limits and less efficient operation. Confidence in the system then declines, more control loops are left in manual mode, and optimal performance is lost.
Sizing and equipment problems:
Requirements are not specified accurately enough. Guaranteed values must be clearly defined. An oversized reduction station may be specified. Two 50% reduction stations usually provide much better pressure and temperature control, improved backup in problem situations, and easier maintenance without shutting down the mill. The actuator type (electric, pneumatic, or hydraulic) or actuator size may not be suitable for the application. After a turbine trip, the reduction station may open too slowly. Reduction stations are often equipped with a quick-opening solenoid that operates with a binary signal, but this is usually incorrect. The station should open to a calculated position rather than fully open. That position should be based on the steam flow through the turbine immediately before the trip. The piping material downstream of the reduction station may not be specified for adequate temperatures. Because reduction stations typically leak some steam, the line may not cool
properly when steam velocity is low and water-to-steam mixing is insufficient. This can raise the temperature enough to trigger a high-temperature interlock and take the station out of service. The spray water line may lack shut-off valves, allowing spray water to leak into the steam pipe. Position signals from important steam valves may not be brought into the automation system. This makes it much harder to identify failures and disturbances caused by malfunctioning equipment. Large pneumatic actuators may have air supply lines that are too small.
Uneconomical operation modes
The auxiliary condenser is used continuously. Steam venting is continuous. The control strategy uses oil or gas instead of lower-cost biofuel. Steam bypasses the turbine. Boiler and turbine trips occur repeatedly.
Uneconomical or incomplete use of the steam accumulator
The accumulator operates as a reduction station, bypassing the turbine and causing a loss of electrical energy generation. When excess steam is available, it is not charged into the accumulator. The accumulator water level is not optimal. Energy remains stored in the water at saturation temperature, and although the accumulator may be correctly sized for optimal use, only part of its capacity can be used if the water level is too low. Charging the accumulator disturbs other parts of the process.
Reduction station control logic problems
Transfers between automatic and manual mode are not bumpless. During an operating mode change, a controller may jump or immediately begin driving in the wrong direction when it becomes active, for example when a limit controller takes over. This causes oscillation. In split-range valve applications, the control may be programmed too simply, which limits optimal valve use. In a correct design, both valves should be freely operable without depending on whether the other valve is in automatic or manual mode. Switching between manual and automatic should also always be possible without disturbing the process. After a turbine trip, the reduction valve may fail to respond as calculated. This can easily cause a boiler trip and, in the worst case, a mill shutdown. The temperature controllers are not configured according to the cascade principle.
Insufficient factory acceptance testing (FAT)
The FAT team may lack sufficient knowledge of steam systems to understand the overall control structure, which can result in an incomplete FAT. In such cases, software faults may remain undetected until delivery to the mill. Because the circuits are complex, development and testing are far more difficult at the plant than in the FAT environment.
The FAT schedule is too short.
Inadequate site supervision
The location of the temperature elements is not optimal. The elements are getting wet and causing so an oscillation of the closed loop temperature control. This is stressing the pipe and disturbing the production.In a dP flow measurement the steam impulse line direction is not downward.In a dP flow measurement the steam condensate pots are isolated.In a dP flow measurement the square root is not systematically applied always in the automation system or always on the flow transmitter.The calibration of the transmitter is not correct. E.g. it is not differentiated between the density difference of the impulse line and the measured
material (caused by temperature difference). In a steam accumulator this causes a too high water level and the demister can't separate all water from steam anymore.
Operator displays
A central operator display for steam balance control, including recipes, is missing. Too many control modes are allowed for the steam balance controllers, such as manual, automatic, local, and remote.
If a recipe display is used, local mode should not be permitted. Alarms, including the suppression of unnecessary alarms, are handled unsystematically. As a result, important alarms may be lost among less important ones
Optimization
By the end of a project, it is often forgotten that system and component design may have taken up to two years. As a result, a system that merely works is accepted even if it is not optimized. Final optimization can only be completed after production has started, when the system can be tuned to operate in the most efficient way.
|