Thread Content
The SIS system is short for Safety Instrumented System; it evolved from ESD (Emergency Shutdown Device) and is designed to provide comprehensive safety protection throughout the entire start-up and shutdown process of production facilities. With the enlargement of petrochemical plants and the increasing complexity of interlock protection systems, SIS systems are being used more and more often. SOE is an abbreviation for Sequence Of Event, which refers to the recording of the sequence in which events occur; it is used to document the time and type of failures (shutdowns). Typically, SIS systems come equipped with their own SOE software for accident analysis and fault identification. As modern chemical plants become increasingly complex, there are many reasons that can lead to plant shutdowns; as a result, both the process operations and instrumentation rely more and more on SOE records to help identify the causes of these shutdowns. Original address: yunrun.com.cn/tech/1934.html 1. Reasons why SOE functions are virtually non-existent in current SIS systems. In many production environments, compared to the emphasis placed on complete, rigorous, and detailed interlock logic, most projects do not pay enough attention to the configuration of SOE, as a result of which the SOE system becomes a mere formality that fails to provide effective assistance in analyzing and identifying the causes of accidents. It is mainly manifested in the following aspects: ① The SOE events are chaotic and disorganized ； ②There are many SOE variables, and some variables that are not related to parking incidents are mixed in among them ； ③The SOE recording time does not match the DCS (Distributed Control System) trend recording time ； ④Locked when the event is triggered, but failed to lock upon recovery ； ⑤Input variables and output variables are mixed together ； All of this is mainly due to the lack of an overall plan for SOE during the engineering configuration process, which hinders both subsequent logical testing and event analysis after the system goes into operation. Changhui Instruments yunrun.com.cn 2. SOE configuration strategies in the SIS system. In response to the above issues, Changhui Instruments has summarized the following SOE configuration strategies based on its many years of maintenance experience. ①Optimize the SOE variable. The purpose of SOE is to help analyze the sequence in which events occur, especially in the analysis of parking accidents. Not all SIS variables need to be SOE variables. Carefully selecting SOE variables to minimize their number is the most important preliminary task in SOE configuration. The selection criterion is whether it is a necessary variable to help us analyze the accident; variables that are involved in the logic but are unrelated to the parking event, such as time timers, channel status, driving conditions, prompts to the DCS, and alarm indicator lights on the auxiliary control panel, should be avoided from being included in the SOE variables. It is ensured that all SOE variables are important data necessary for our analysis of accidents and identification of faults. ②Optimizing the operation cycle of the SIS system: SIS systems are all developed based on PLC (Programmable Logic Controller) hardware, and the system scanning cycle is determined by the actual program execution time plus an appropriate margin. The task of the SIS system is to monitor process parameters and exert final \"control\" over any \"out-of-limit\" conditions that could lead to severe consequences, with most of its actions being irreversible. Once the stop condition is activated, the device to be protected is required to stop quickly, safely, and stably, reaching a safe state. Therefore, we require the system to have a fast response time and low system load, with the goal of achieving high safety and reliability. The shorter the scanning cycle of the system, the fewer records appear in a single cycle within the SOE records. This helps to reduce the occurrence of multiple stop signals within the same scanning cycle, thereby facilitating the identification of accidents. In accordance with the requirements of GB/T50770-2013 Code for Design of Safety Instrumented Systems in Petrochemical Industries, the response time of the controller (input/output scanning plus computation time) should be 100–300 ms ； The load on the controller does not exceed 50%. Another important task in engineering configuration is to reduce the program execution time while ensuring logical integrity, reliability, and readability. a. Reduce the preprocessing of input data: Tasks such as signal squaring, linearization, noise reduction, and damping – which are common in DCS systems – should be carried out outside the SIS system whenever possible. For example, performing signal squaring and damping processing at the field transmitter, as well as converting temperature signals into standard signals in advance, in order to minimize the load on the SIS system as much as possible. b. Make flexible use of simple data types and data structures. Using data types flexibly during engineering configuration can effectively reduce memory usage and shorten computation time. For example, when performing logical configuration, use arithmetic operations as little as possible and replace them with logical operations ； When arithmetic operations are necessary, use addition and subtraction more often, and multiplication and division less often ； When there are no operations such as multiplication, division, or square roots, integer types should be used preferentially ； For analog values used solely for interlocking, if an integer type is chosen the range is too small resulting in insufficient precision; this range can be appropriately increased (by integer multiples). When performing numerical conditional operations, use comparison modules as much as possible, and use equality checks sparingly or not at all ； Furthermore, different treatment should be given to operations on various data types in order to prevent errors such as overflow and large deviations in the controller; for example, when performing division, it is necessary to check whether the divisor is zero ； The square root operation checks whether the number for which a square root is sought is negative ； When using operations such as multiplication, division, and square root, floating-point numbers must be used; otherwise, significant calculation errors will occur. c. Use complex data types and algorithms cautiously. Modern SIS systems support various configuration languages, and many function libraries have been developed for reasons such as intellectual property protection and readability. These configuration tools greatly facilitate the implementation of logical functions. However, when using it, one should avoid excessive complexity in logic to prevent increasing the load on the controller. For example, use system function libraries as much as possible and rely less on constructors ； Use Function Blocks (FB) less often and functions (FC) more often, as Function Blocks require separate storage space ； Reduce or avoid PID computation ； Reduce the use of loop statements ； The use of recursive algorithms and similar methods is strictly prohibited. While maintaining logical correctness, safety, and readability, the logical structure should be simplified as much as possible in order to reduce the load on the controller. ③Grouping SOE variables: Setting SOE variables is primarily intended to help analyze the causes of events and the sequence in which actions are carried out. The above queries are not used simultaneously; for example, when we look up the reason for parking, we only need to enter the variables in order ； To analyze the action sequence, it is only necessary to arrange the timing of the output variables. If the SOE variables are not grouped in advance, SOE records may mix input variables, output variables, and intermediate variables, which hinders the identification and analysis of events. Furthermore, sometimes a single SIS system incorporates the logic of multiple relatively independent devices or functional areas, which also requires grouping of SOE variables. It should be noted that, depending on the system, the SOE variables in many systems need to be grouped at the time of definition. Therefore, it is necessary to group SOE variables in advance at the initial stage of project configuration. a. First, group them according to different functional areas within the system, such as several parallel production lines, or the control of several relatively independent units. Even if the control of these production trains or units is distributed across different SIS controllers, it is necessary to group SOE variables, as these SIS controllers may share the same SOE server. b. Group by input/output type; do not include intermediate variables ； For the analog values involved in interlocking, the boolean values resulting from comparison by the voting module are regarded as DI points, and not as intermediate variables ； c. For projects with strict timing control, the sequence numbers of the sequential control steps are also important parameters for analyzing the event process, and should be grouped separately. ④Time Synchronization: To ensure consistency in the timing of historical trends and SOE data across different systems within the same device, it is necessary to synchronize the clocks between these systems. For example, in the SOE records, the time of the variable is the time stamp provided by the controller, which is based on the controller’s clock ； The clock at the SOE station may be the clock of the engineer’s PC; the two are not necessarily identical ； Similarly, it may not be the same as the operator station clock either. Only on the basis of a time reference is it possible to compare and evaluate the trend records of analog values with the SOE event records. There are many methods for clock synchronization, such as timing via the Global Positioning System, NTP (Network Time Protocol), and DCS correction-based timing; sometimes it is necessary to use several of these timing methods together. When performing clock synchronization, the following points should be noted: a. The uniqueness of the clock source. With the coexistence of multiple systems such as DCS, SIS, and ITCC, the operation stations, control stations, engineer stations, SOE stations, etc. of different systems maintain the same time reference to avoid redundant time synchronization. b. For a single system, pay attention to the consistency and unidirectionality of time synchronization between the upper and lower layers of the operation station and control station. c. When using correction-based time synchronization, it is necessary to avoid synchronizing at whole-hour marks (the end of a day, week, etc.), in order to prevent large errors caused by signal jumps. d. In order to ensure the stability and resistance to interference of the timekeeping signal, corrective timekeeping should involve the proper use of delay modules to avoid any delays in time delivery. e. Regularly check the consistency of the clock signals to detect timing errors in a timely manner. Due to interference, switch failures, cable faults, IP address conflicts, and other reasons, the time synchronization signal can be interrupted; therefore, maintenance personnel need to conduct regular checks, restart the time synchronization service promptly, or manually correct any timing errors. ⑤The start and stop recordings of SOEs: In a SOE server setup that includes multiple SIS controllers or several relatively independent control units or devices, the SOE recordings for each unit (device) in such a system are not necessarily simultaneous. For example, there are production facilities for multiple series; some of these series are operating normally, while others are undergoing routine maintenance ； Some units (components, devices) will experience short interruptions in operation. For these units that are under maintenance or have short shutdowns, the SOE records of their systems should be stopped; otherwise, during the shutdown period, a large number of invalid records will be generated as a result of interlock tests, valve tests, system debugging, etc. Such records are useless for analyzing accidents and failures, but they do take up a lot of storage space. In fact, the default setting of most system manufacturers is to automatically start SOE recording once the DCS system is powered on. So there is a lot of room for optimization in this part. For example, with Tricon’s TS3000, the system provides up to 16 sets of SOE records, with each set capable of storing up to 20,000 records (in a buffer, first-in-first-out format). Each group of SOE records can have start, stop, and clear functions set. This allows for the reasonable grouping of SOE variables in SOE stations that have many relatively independent devices, units, series, or assemblies, and enables the setting of appropriate conditions for starting (stopping, clearing) SOE recordings, thereby ensuring that the SOE station records only the data we need. For some SIS systems whose SOE functionality relies on third-party software, the start/stop recording of SOE events is handled by the upper-level computer; it is important to communicate with the manufacturer in advance to ensure that these functions can be implemented. Furthermore, when configuring SOE, it is also necessary to note that some system software defaults to recording only the alarm time and not the recovery time; therefore, such important variables need to be set manually. The points mentioned above are the main issues in SOE configuration, as well as the areas where problems frequently occur during routine maintenance. In summary, the SIS system can fully leverage the value of SOE records and turn them into a genuine tool for analyzing and diagnosing faults only through configuration strategies such as rational optimization of logic, reduction of SOE variables, lowering the load on controllers, and effective grouping. Author: Xu Qiang, Operations and Maintenance Center, Sinopec Qilu Branch