Abnormalities during oil pressure measurement cause frequent shutdown issues! The situation has been checked now; it’s 15L
Thread Content
This post was last edited by lk800811 on 2011-9-27 at 20:42. Yesterday, there was a sudden fluctuation in the medium-pressure nitrogen oil pressure at my workplace; the alarm records indicated that a low-low level trigger led to a cascade shutdown. However, no such low-low values could be seen in the trend data – perhaps it was too short a period for them to be recorded. Therefore, it was checked that the wiring at the EJA header was in good condition, there was no looseness in the intermediate junction box, and there was no looseness in the AI wiring channels of the DCS either. Then check the oil circuit drain valve; there is no blockage in the drain passage, and it seems that everything is normal. Therefore, restart the process. As a result, at 7 a.m. this morning, the same thing happened again – it jumped out of the vehicle. So I started checking the parking logic and found that there was no delay in its output; I then added an interlock block with a 1-second delay. I drove the vehicle again; it ran for just over 20 minutes before shutting down once more. Therefore, I decided to replace the pressure transmitter with a ROSEMENT 3051 model and carried out the configuration. We drove the vehicle again, but it stopped after less than 30 minutes of operation. So we checked the equipment to see if there were any issues, and so far the equipment is working properly; we are currently draining the oil from the system in order to replace it with new oil. I really don’t sense anything unusual there. I hope everyone can help and think about where the problem might lie. Is it possible that the oil pipeline is actually blocked? (The oil pipeline is quite long, about 10 meters), but it seems to be unobstructed to me when I check it manually.
In your description, the word that appears most often is “again,” which shows that this matter is quite troublesome. The transmitter is fine; if it’s electromagnetic interference, adding a delay should help prevent interlock. So, if the cable quality is poor and there are invisible joints in the middle (not referring to the junction box), could poor contact occur due to environmental vibrations?
Reply to 2# denghl: It should be impossible to have joints in the middle; we don’t allow any joints to be present there. . It really doesn’t work; tomorrow I’ll have to disconnect the oil lines and take another look
Reply to 1# lk800811: Is the lowest point in the trend close to the low value, and are the time points roughly consistent? You'll know once you take a look.
Switch to a pressure switch; it is recommended to use a two-out-of-three oil pressure interlock.
It can be considered whether the 3-out-of-2 interlock AI channel should be equipped with a SOE function and measurement accuracy at the mS level, so that real-time data during tripping can be recorded. Overall, it is advisable to check the oil pipeline to see if there are filters, whether they are clean, and what the quality of the oil is; it is likely that the issue is caused by sudden fluctuations in oil pressure
I’m not entirely clear about the requirements regarding your process. If the process permits it, the delay time can be increased; additionally, more measurement points should be used to obtain signals, and a comprehensive assessment should be made. If low-low alarms occur continuously during that delay period, the machine can be stopped; otherwise, it can be assumed that the fluctuations are normal and the alarm can be reset. I’m not sure whether the low-low alarms you’re monitoring are based on on-site analog signals that you evaluate yourself or on signals from pressure switches. It’s better to use signals from pressure switches, as instant interference with analog signals can cause problems and lead to incorrect operations
Today, I started another thorough inspection; an additional transmitter was added, and the cables were replaced. A 2-to-2 selection option was incorporated into the logic. We’ll see how it goes this time – if there are still problems, I really won’t know what to do! ! !
Reply to 1# lk800811: I think we should first determine whether the pressure does indeed fluctuate momentarily； If there is indeed an instantaneous fluctuation of 2 transmitters, it will also trigger a trip. The typical scanning cycle for DCS is 1 second or 500 milliseconds, while the cycle used for historical trends is usually around 10 minutes; as a result, something that is not visible in the historical trends can still be detected by the DCS. Whether your interlock is triggered by “0” or “1” determines the selection of “Off delay” and “On delay”. I think in DCS, if the signal in question appears to be normal, and all other instruments are also functioning properly, but an action was taken despite no indication in the historical trends, one should first check whether the delay block has been added correctly.
Is there electromagnetic interference? It is possible to check the grounding of the DCS as well as that of the large electrical equipment in the vicinity. Last year we experienced an electromagnetic interference issue; during such incidents, some gauges showed suddenly lower readings before returning to normal over time. Upon investigation, it was found that the connection cables passed through the air-conditioning room, and the start/stop of the air conditioner was causing the interference. After re-grounding the indoor and outdoor units of the air conditioner, everything returned to normal
Installing a buffer tube in the transmitter’s measurement pipeline should solve the problem.