Oil level problems

[gerry79]

I have a Weka (Tab) level indicator on a vertical oil seperator, the column it's attached to is piped externally from the seperator vessel to which is also attached the oil level proxi switch. This level indicator column has been heat traced & until recently there haven't been any problems, since the heat trace failed I have experienced regular oil level trips inspite of the WEKA indicator showing there to be plenty of oil (Seperator not fitted with sight glasses).
I replaced the heat trace to see what would happen & haven't had any further problems, the heat trace seems to be critical to the effective operation of the oil level switch but can anybody tell me why this is?
The system in question is flooded NH3 providing cooling for a coldstore which is not occupied overnight, this also seems to be when the majority of trips occurr. Any feedback would be appreciated.

[RANGER1]

jerry79,
Maybe the float inside column is weighted for the specific gravity of oil, not liquid ammonia.
The float will not float with liqid ammonia in it, only oil.
Without the heat trace it acts like a mini condenser, as it is cooler than the oil seperator, condensing liquid in it displacing the oil.
The heat trace keeps column dry of liquid refrigerant in the column.

Maybe someone else can add or has a different idea.

[gerry79]

Thanks RANGER1,
That sounds logical but it seems there has alway been enough oil in the seperator to satisfy the level switch inspite of low level faults & I have no reason to suspect the proxi switch is faulty. I would have thought the oil, having the greater specific gravity would have found its level in the column with the addition of condensed vapor above this, which although it would falsely indicate a higher level would still have a satisfactory oil level as detected by the proxi switch.
I'm sure there's something i'm overlooking!

[Nh34life]

Sounds bang on to me Ranger , would also happen more at night as mentioned. Strange design of an oil seperator without a sight glass pat the design engineer on the back make him feel special.

[RANGER1]

Thanks RANGER1,
That sounds logical but it seems there has alway been enough oil in the seperator to satisfy the level switch inspite of low level faults & I have no reason to suspect the proxi switch is faulty. I would have thought the oil, having the greater specific gravity would have found its level in the column with the addition of condensed vapor above this, which although it would falsely indicate a higher level would still have a satisfactory oil level as detected by the proxi switch.
I'm sure there's something i'm overlooking!

We build oil seperator sight glasses with heaters for similar reason as you cannot see where oil level is as liquid is clear.
Have also seen similar setup that you have on petro chem or mining packs as they want high quality stuff.

A simple test is to close top isolating valve & gently warm column with warm water or similar.
It will force refrigerant/oil mix out of column.
Open up again & I'm sure it will read normal again, for a little while anyway.