Glycol filling practices - insight appreciated


Thread Starter



I would like to thank you in advance for any of your comments, insights, and opinions on my question; I will most certainly take it all into account. I would also like to say that I am using the word he over she, but am in no way making any statement with that other than for ease; there are quite a few lady instrument technician in my shop, and they are great!

So to start off, I am an Instrument Technician working at an Upgrading facility; basically upgrading bitumen to a barrel of oil for refineries abroad.

We have cold winters, and as such most of our flow, level and pressure transmitters require the legs to be glycol fill... and there are a lot of transmitters. My question mostly concerns DP flow transmitters

At my shop there are two current schools of thought, and the odd guy who does a bit of both schools.

The first school of thought from the more senior technician involves using a can filled about halfway full of glycol, and then filling that can with air. This is the method of deploying the glycol. The technician then goes to the transmitter, blocks in the transmitter at the 5 valve manifold, blocks in the root valves at the tap points, and opens the vent to atmosphere at the tap point below the root valve. He then proceeds with his glycol fill, pushing the glycol through the lines from the drain valve a the bottom of the 5 way manifold and out the vent at the top. Once the liquid coming out is clearly glycol and no longer process, he zeros the transmitter and recommissions it.

The second school of thought involves what we call a Baker pump. I am not sure the actual name of this style of pump, but it works off a manual lever, really a heavy duty siphon hand pump - looks like a red metal portable gas can, with a handle on top. This is capable of pumping thousands of pounds of PSI. Now what is being done with this is the instrument technician goes to the transmitter. He blocks in the high and low legs at the 5 way manifold, opens the equalize valve and connects to the drain valve on the 5 way manifold with his baker pump. He primes the pump a bit and then opens one side up and pumps continuously thereby overcoming process pressure and pumps until there is an excess amount into that line - maybe 20 pumps depending on the length; it's an estimation every time - and also into the process. He closes that side and keeps pumping until the valve is fully closed. He repeats it on the other side, going through the equalizer so as not to damage the transmitter cells.

Anyway I will boil it down to the points being contested. The pressurized can way of doing it ensures that the lines are indeed full of glycol, whereas the baker pump is an assumption that after " X " amount of pumps on either side. So it has a certainty to it and a confidence in it's accuracy.

The pressurized can people - PCP- (laughing at the title) feel that the glycol going into the process is bad practice, and contaminates the process. The baker pump people BPP, argue that the the vessels have water troughs, and that since the glycol boils off over time or is sucked out anyway, it makes little difference. The PCP also say that there is the potential for the process to push back into the can and blow the can up - to which the BPP say they have 2 check valves ensuring that doesn't happen, even if they were to stop pumping... to which the PCP respond that you can't trust your life with check valves, to which the BPP say you're trusting a block valve....etc...

The reason the BPP are not for going through the vent and pushing out the glycol at the top into atmosphere is because of the exposure to process which is often foul smelling, and unhealthy to breathe in if it comes out in a vapour cloud. And there is no guarantees that the process is exactly what the operators claim it is... as in it could be H2S rich without their knowledge. All that can be saved with supplied breathing air, but the volume of transmitters done, this is unfavourable and not cost effective. The BPP say, why bother taking the risk and exposing yourself to that danger.The baker pump also has the advantage of pushing through the root valve and clearing any blocks, doing away with the need for nitrogen bottle blowdowns.

The real question I have is what practice do you guys use or have heard of? What are the drawbacks of either method? Is there a better, safer way? Is there a happy medium... Any insight and help is very much appreciated.

Kind Regards.
Hi there,

In level, flow and pressure applications we just use a T-Piece on the bend just before the impulse line goes into the process. On top of the T-Piece there is a rated needle valve.

We block in the main isolation valves of the impulse line at the process end, close the equalizing valve and start filling the lines until it overflows at the needle valve open end. At that point all air bubbles will be out and the lines will be filled to the main isolation valves before the process. You can do some draining at the 5-way manifold to make sure. We now close the needle valves and turn in a safety plug in it's open side. We then open the main isolation valves to the process. The Glycol will now start to run into the process up to a point where the impulse lines is filled to the overflow point and have been compressed as much as it will be by the process pressure. At that point you can do your process zero, if the process conditions is right, otherwise you need to disconnect the lines from the process and do at least a atmospheric zero. Very safe and easy to do and pretty much the same for level, pressure and flow.
I also agree with you about using Glycol since it eliminate a lot of problems.