View RSS Feed

passandscore



Liquid Floodback at its Finest!

Rate this Entry

I first mentioned this story back in the summer of 2013. Since then the problem reoccurred and further measures were taken. This is an updated account from start to finish.

Click image for larger version. 

Name:	2013-08-08T05-52-12_0.jpg 
Views:	1727 
Size:	95.9 KB 
ID:	11179 Click image for larger version. 

Name:	2013-08-08T05-52-12_1.jpg 
Views:	1774 
Size:	94.6 KB 
ID:	11180

You're looking at a low stage Mycom compressor liquid-filled all the way up the discharge line! You are probably wondering how something like this can even happen? The story goes like this.

The system was low on ammonia so we decided to use the ammonia that was currently being removed from another location as it would be free to the customer and spare us a disposal cost. The tanker truck arrived and added 700lbs to the system. All is good, right?

Well it wasn't until the tanker truck was ready to depart that he decided to share some vital information with us. In his opinion, the ammonia that he had just added was dirty. That means oil! Possibly lots of oil as it is heaver than the liquid and sits at the bottom of all vessels, like the one on the back of his truck. So now what? There was no way to determine how much oil was or was not pumped into the system. The plan was to drain the oil pots regularly. Well as you already know, that didn't happen.

This particular system has two low stage compressors pulling from a low temperature vessel and that vessel maintains a constant liquid level within it by means of an eight foot probe located in a column that is connected to the vessel. If there is anything worth learning from this whole experience it's that oil and probes do not mix. When oil gets on a probe, the probe no longer responds properly. In this case the probe told its controller that the physical level inside the vessel was lower than the controllers set point. So a long story short, the vessel was 30% full but the oil on the probe told the controller that it was 10% full which led to the liquid line solenoid being opened to make up the difference. Now with the solenoid opened, liquid began to fill the vessel. This is where things went from bad to worse.

With the probe constantly reading 10%, the liquid line solenoid remained open. The liquid continued to fill the vessel. The mechanical high level control located on the vessel functioned properly. The problem was in the main control panel where the high level shutdown circuit had a faulty relay. So, with the liquid making its ascent to the top of the vessel,it began to enter the suction line and make its way to the two low stage compressors. Once in the compressor it diluted the oil, causing the machines to shut down on their oil failure controls. The controls saved the equipment from imminent failure but it didn't stop the liquid as the solenoid remained open and liquid continued to fill the compressor crank case, making its way up the oil return line, lifting the oil return float ball and entering the oil separator and discharge line. The picture truly tells the story!

It was at this point that the customer called and informed me that his freezer and cooler temperatures were too high. Unaware of the current condition of his plant room, I arrived and was in shock! I had never before seen such flood-back. This was certainly liquid flood-back at its finest! After photos were taken the repair work began.

Removing that much liquid from a compressor is not hard. I started the ammonia pumps and put the high stage compressors online. I isolated the flooded compressor and installed a hose from the discharge of the operating high stage machine to an access just below the suction isolation valve on the flooded compressor. I had to remove some existing tubing but it was critical that the hose be connected to the highest possible point. A hose was connected to the oil fill valve on the compressor crankcase, with the other end connected to the oil pot on the high temperature vessel. In order for this procedure to work the high temperature vessel pressure needed to be lower than that of the isolated, flooded compressor. With the high-side compressor running I achieved exactly that.

With all prep work complete it was time to initiate the procedure. The goal was to open the oil drain valve on the compressor and the valve on the oil pot, creating a path for the liquid. Once that was established it was only a matter of opening the discharge line from the high stage compressor and injecting the hot gas into the top of the suction on the flooded compressor. With over 100 psi of hot gas placed on top of 25 psi of liquid, it was forced through the oil valve, via the hose and into the oil pot, oil and all. It took about five minutes to transfer and empty the compressor. Once emptied I opened the isolation valves, allowing more liquid from the lines to fill the compressor and did it again. After 15 minutes all the liquid had been removed. I completed an oil change and put the compressor back online. Mission accomplished.
Tags: None Add / Edit Tags
Categories
Working in Refrigeration , Working Blogs


Comments

  1. RANGER1's Avatar
    Passandscore,
    Certainly a good floodback & lucky for it to cause no damage.
    Just a thought on the high level not working, is it fail safe?
    For example if a wire or plug falls off high level switch will it shut down plant.

    Depending on what was wrong with the relay, maybe it should be somehow made fail safe if possible.

    I'm only mentioning this as plants are left vulnerable, not just yours.
    Depending on programmer & how things are wired can cause problems months or years later.

    Pressure transducers are another example.
    If a HP transducer plug falls off will PLC system recognize this as a sensor failure & shut compressor down.
    If not plant is out of control waiting for an accident.
    We had a problem recently with a fault relay being held in by residual voltage. Had a controls electrician to investigate fix problem & also put in a bigger relay in case it happened again.
    so
  2. HVACRsaurus's Avatar
    Very informative thank you.
  3. passandscore's Avatar
    The high level is not fail safe. I agree with your view point and will mention it to others.
  4. RANGER1's Avatar
    Quote Originally Posted by HVACRsaurus
    Very informative thank you.
    Never sure if your taking the piss or not with me!
  5. HVACRsaurus's Avatar
    Hey Ranger, not taking the piss - save that for the guys who can't understand why thier tin can isn't giving 'em 1MW of cooling running off an ebay solar panel.
  6. RANGER1's Avatar
    Just to add, depending how good your PLC programmer or controls person is, further protection can also be added to this scenario.
    If compressor is selected in auto with no faults or alarms,liquid feed to vessel is in circuit.
    If all compressors on this vessel are in a fault condition & in auto then it can interlock with liquid makeup solenoid to prevent it opening no matter what the level.
    Liquid pumps have their own safety with low level or pump differential fault.

    How often does this happen & a truck load of liquid returns to accumulator causing high level after a compressor fault.
  7. passandscore's Avatar
    It only happened twice. Good point about the interlock. There are a couple of improvements that will have to be made but at least the main problem is resolved. I don't expect to see liquid flood back like this again. Well, not from this facility anyway.