Start of the liquid oxygen pump in the internal compression process

In the internal compression process, the liquid oxygen pump is started; this is how we operate in our factory. Since we use the Air Liquide process, which involves nitrogen expansion, the setup differs from the current process. Before starting the oxygen pump, it is necessary to cool the oxygen side of the high-pressure oxygen heat exchanger thoroughly using liquid nitrogen – the temperature at the tube sheet should be reduced to around -160 degrees. The start-up of the oxygen pump involves two steps: air cooling and liquid cooling, each taking at least one hour. The pump must be rotated manually until gas in the outlet pipeline changes to liquid form and is discharged. All these steps must be completed before any amount of nitrogen is introduced into the nitrogen side of the high-pressure oxygen heat exchanger in order to start the liquid oxygen pump. Nothing like what you mentioned has happened to us, but after you start the pump, both pressure and flow rate experience significant fluctuations, which indicates that there is too much gas inside the pump. This could be due to incomplete cooling, or it might be that the pressure of the sealing gas is too high, or there are leaks in the inlet pipelines. There is a relationship regarding the sealing gas: the pressure of this gas must be higher than the pressure at the pump inlet, and the pressure at the pump inlet must in turn be higher than the pressure at the vent. This ensures that the sealing gas does not enter the pump, while allowing the gas inside the pump to be discharged smoothly through the vent. Lastly, the insulation of the liquid oxygen pump’s cryogenic tank must be thorough; there should be no vacuum gaps.

We manually set the opening degree of the vent valve to around 20–30%, then gradually increase it to the rated speed, and subsequently close it manually. The inlet return valve and the outlet return valve are set to around 10% opening; the outlet return valve is switched from manual to automatic mode. Then, the load is adjusted to 80–90% through the vent valve, and this process is carried out gradually, with both system pressure and flow remaining relatively stable.

We use variable-frequency control, so this situation does not occur

After increasing the pump pressure, ensure full pressure in all oxygen pipelines before raising the oxygen flow rate

Large fluctuations in pressure and flow rate behind the pump occur when liquid is supplied to the high-pressure plate type system, which indicates that there is an issue with the operator’s handling of the process. Although the liquid pump is operating properly, the channels in the heat exchanger and those leading to the cold box are empty. The outlet valves must be opened slowly, just as they would be during the initial setup of the piping system.

What was said on the third floor makes a lot of sense; we plan to proceed in that way from now on as well///

This is a typical case of cavitation – such a situation occurs when the pump isn’t cooled properly!

Cavitation phenomenon: 1 Insufficient cooling before pump startup leads to a large amount of liquid vaporization during startup, resulting in cavitation. When liquid oxygen enters the heat exchange equipment, the lack of pressure in the pipelines, combined with heat exchange with the feed air flowing in the forward direction, causes a large amount of the liquid to vaporize. This vaporized liquid then flows back rapidly to the pump outlet, resulting in cavitation. Solution: Ensure adequate cooling before starting the pump, and perform air release on the pump. It is best to use variable-frequency control to cool the outlet pipeline slowly. Prevent the gas that has been vaporized from flowing back to the inlet. If variable frequency control is not available, the outlet valve can be adjusted slowly, while ensuring that the air release valve does not open too wide. At the initial stage of installation, the pump was in a fixed position when it was first exposed to cold, which meant that it was subjected to minimal stress and this was not noticed. Now, due to repeated operations, the pump and its pipelines are affected by changes in stress, resulting in unstable operation of the pump; this could also explain the phenomenon described by the original poster. I’m not sure if anyone has different opinions on what I’ve said; I hope everyone will offer their corrections and criticism

I think it’s safer to pressurize the oxygen supply pipeline first. If there is no pressure in the pipeline, accidents can easily occur due to friction between the high-pressure oxygen and the pipeline. When supplying oxygen, we first use nitrogen to create pressure in the pipeline before delivering oxygen – safety comes first, after all

Agree with what was said on floor 8! But I also don’t recommend turning on the vent valve!