Thread Content
This post was last edited by 328104062 on 2016-6-29 00:07. As the title indicates, the issue that needs to be resolved is: taking the CS3000R system as an example, implementing classification of alarms within a DCS system (different HIS for different types of alarms) as well as hierarchical organization (different sound levels, sounds, frequencies, screen flashing, pop-up windows, etc., for different alarm levels). The current situation is that alarms are not organized hierarchically nor by operation station; whenever an alarm occurs in a certain domain, it is displayed on all operation stations within that domain, regardless of which device or position is involved. For example, in the FAR03 domain, there are devices such as reformers, hydrogenation units, hydrogen production units, and benzene extraction units. Each of these devices has its own HIS, and different personnel operate these HISs, but the alarms are identical – the personnel working in the hydrogen production unit receive the same alarms as those working in the hydrogenation unit, etc.). The consequences of this are that alarms become ineffective; if the alarms are set to sound continuously, it creates noise that interferes with work. Additionally, there are too many alarms, which can lead to complacency among operators. In the FAR03 domain, each person receives alarms from multiple devices, far beyond what the human mind can handle. Moreover, since alarms are not categorized by position, it’s difficult to determine which operator should pay attention to a given alarm or who should handle it; 3. There is no attention paid to this issue; the alarm system does not take into account the importance of actual operational parameters, as there is no classification system at all. As a result, less important alarm messages are sent to operators frequently, causing them to become indifferent to them. For example, the liquid level alarm thresholds for underground oil tanks are set at LL30% and HH70%, which are identical to those for important alarms such as those related to the liquid levels in the high-pressure tanks of hydrogenation units and the liquid separation tanks at the compressor inlet. The alerts provided after an alarm occurs (via sound and light signals) lead operators to ignore the more important alarms. 4. There is no distinction between system alarms and process alarms: in the operator’s HIS, system alarms are displayed, but ordinary operators usually do not understand the system, while process alarms appear on the DCS engineer’s station; DCS engineers, on the other hand, do not understand production processes. When all information is sent to everyone, it leads to ambiguity in job responsibilities, and it is necessary to make distinctions (for example, a DCS engineer might be responsible for inspections but fails to do so properly, then shifts responsibility by claiming that the operator failed to report system alarms). The issues that need to be addressed are: 1. Alarms within the same domain should be categorized by device. For example, in the FAR03 domain, there are devices such as reformers, hydrogenation units, hydrogen production units, and benzene extraction units. The reformer unit has 9 HIS units, numbered HIS0121 to HIS0129; these 9 HIS units should only receive alarms from the reformer unit itself. 2. Alarms within the same domain should be categorized by job role. For example, if HIS0121 belongs to the operation station for the reaction unit, then it should only receive alarms related to that specific job role, although it can also be adjusted manually to display alarms from the entire reformer unit. 3. Alarms should be classified by severity: for instance, the liquid level in the compressor inlet liquid separation tank could be classified as a level 1 alarm, with the screen parameters flashing and a buzzer sounding at a high frequency and volume ; The liquid level of the high-pressure tank is considered the second level of alarm: the screen flashes and a low-frequency, loud beep is emitted…… The temperature of the underground sludge tank constitutes the highest level of alarm; the screen flashes but no sound is produced. Additionally, the latest alarm records should be displayed below the HIS screen (3–7 items, 4 items are recommended). Of course, process engineers need to be involved in this. 4. System alarms and process alarms should be completely separate; system alarms should not appear on the operator’s HIS screen. Request for assistance: 1. How can these requirements listed in points 1, 2, 3, and 4 be implemented? Does anyone have experience in implementing such solutions? 2. How are alarms generated in DCS? Is it the controller that sends the signal or HIS? And how is the signal transmitted to the buzzer? 3. What is the difference between system alarms and process alarms in the CS3000? Which parameters are included in system alarms? (To my knowledge, these include communication between redundant optical fibers and cards, the CPU usage of the controller, the status of downloads, system time, etc.; please add any other parameters you know.) What information are recorded in system alarms?
I’m asking this because of the needs related to driverless operation and screen-off mode
This is a problem related to programs and systems, right?
Of course, otherwise why post it here?
1. Regarding the requirements in 1, 2, 3, and 4 above, how should this solution be implemented? Does anyone have prior experience in implementing it successfully? Classifying alarms is not difficult to accomplish. 2. How are alarms generated in a DCS? Is it the controller that sends the signal or HIS, and how is the signal transmitted to the buzzer? DCS alerts are mostly generated by the HIS; I’ve never seen controllers with audio functions – most of them use flashing alarm lights to convey messages. 3. What is the difference between system alarms and process alarms in the CS3000? Which parameters are included in system alarms? (To my knowledge, these include communication between redundant optical fibers and cards, the CPU usage of the controller, its operational status, system time, etc.; please add any other parameters you know.) What information are recorded in system alarms? Communication and hardware failures require someone familiar with the CS3000 to provide guidance
For the alarms in our company, operators generally don’t pay attention to them (except for abnormal feedback such as flashing graphics). For important parameters such as level, temperature, and pressure, voice alarms are used (with additional amplifiers); the DO signals are sent to these amplifiers, and pre-recorded MP3 audio files are played to deliver the corresponding voice messages
I’m not familiar with the CS3000, but alarms in HMI systems are a common issue. The problems you mentioned actually revolve around the classification of alarms. There are more than one standard for such classification; at least the following categories exist: 1. Classification by alarm severity, such as critical faults, minor faults, events, and system alarms. Alarms in the last category generally do not need to be displayed on the operator station. 2. Classification by alarm area – usually based on the process equipment, with each operator terminal displaying alarms from one or several specific areas. In Siemens’ Wincc, as long as these details are taken into account when defining alarm information, setting up different filtering conditions in the alarm controls will suffice to meet the requirements. In PKS, alarms are classified as urgent, high, low, or Journal, while areas are defined using assets; this approach also makes it easy to meet the requirements
As mentioned on floor 9, I’m not familiar with the specifics of CS3000, but in most monitoring software systems, it’s possible to set priorities for alerts, as well as assign them to different areas or to the overall area. Regions can be different devices, different systems, different physical locations, or different logical locations. However, for configuration software, it is possible to achieve this as long as some time and effort are invested in setting up/configuring the data points for this function (of course, with the user’s cooperation). Depending on the scale and complexity of different systems, zoning and grading of alarm functions is, and should be, an issue worthy of attention. The issue raised by the original poster is quite common in practical applications, and neither users nor developers pay enough attention to it. As an important function of human-computer interaction, in addition to timely and accurate recording of alarms, proper configuration/settings are essential so that different operators can effectively be aware of the warning messages related to their respective areas, which is of great significance for safe production.