All about diaphragm seal Transmitter 2

[Michael2009]

Helo. I have a project using a lot of diaphragm seal DP tx to measure level. But i have a lot of query regarding this. Because i nvr do it before. Any experience guy please help to advise me. Thanks

1. Anyone know why usually diaphragm seal transmitter did not have small range like -100 mbar to 100 mbar???

2. Any mounting concern about diaphragm seal transmitter???

3. how we gonna handle those extra long capilary???

4. Any one hear before the distance transmitter mount below H side cannot be more than distance H side tapping point and L side tapping point.

Thanks for your advise.

 

[djs]

Diaphram seals introduce a mechanical resistance between the fuild being measured and the sensor. This makes measuring small pressures with a seal impractical.

 

[danw2]

Be aware that remote seals constitute a closed, liquid filled system, which according to Boyle's or Charles' Lawmeans that the pressure of the filled system is directly proportional to the temperature.

The pressure seen by the sensors in the DP body is the combination of the applied process pressure and the pressure induced by temperature of the closed, filled system.

Which means that the filled capillaries are not only pressure conduits, but thermometers reflecting how hot or how cold they are.

Temperature variations of the capillaries at 100mbar will be reflected in the pressure at the DP cell is likely to be a considerable percentage of the applied process pressure.

 

[Michael2009]

[qoute]Be aware that remote seals constitute a closed, liquid filled system, which according to Boyle's or Charles' Lawmeans that the pressure of the filled system is directly proportional to the temperature.

The pressure seen by the sensors in the DP body is the combination of the applied process pressure and the pressure induced by temperature of the closed, filled system.

Which means that the filled capillaries are not only pressure conduits, but thermometers reflecting how hot or how cold they are. [/qoute]


How about if the measurement are just only -32.2 mbar to -10.7 mbar?? I should not go for transmitter range -100 mbar to 100 mbar??



Temperature variations of the capillaries at 100mbar will be reflected in the pressure at the DP cell is likely to be a considerable percentage of the applied process pressure.

 

[Michael2009]

Diaphram seals introduce a mechanical resistance between the fuild being measured and the sensor. This makes measuring small pressures with a seal impractical.

noted. .thanks

 

[roydm]

So why are you using a lot of filled capilary systems?
They are expensive and can cause all kinds of headaches for calibration and maintenance.
I suggest you re-evaluate if you really need these.
95% of the levels we do in petrochemical / Acid related process can be done with regular 3" diaphragm seal level transmitter and either dry or wet LP connection. In the past 10 years I think I have purchased 2 remote diaphragm seal transmitters.
Roy

 

[Sascalt]

I am uncertain as to what make of devices you have selected but you may wish to concact rosemount as they have a best practice available that will aid in installations!
Some things to ensure you consider
1-always have equal lengths of capillary lines or the device will NOT work! Changes in ambient temperature will result in constant zero shifts!!!
2- attention must be paid to installation of gaskets for bleed rings to ensure no damage is done to diaphrams. Far to often theses are not ordered with bleed rings and this results in a maintenace headache trying to preform calibrations.
3- the longer the capillary lines the slower the responce time so these devices typically will not prwform well in a control loop application unless they have large capillary lines and short runs.
4- I would suggest you consider insullating capillary lines for beter preformance. Care must be taken to ensure that the lines are protected from external temp. sources!! I have had good success with heat tracing lines with selflimiting tracing above ambient temps and insulating the lines!
Best of luck

 

[roydm]

Everything Sascalt says confirms my response, they are nothing but trouble. Having said that sometimes they are the only solution.

 

[Mass44]

Hi there,

Seems like you have a lot of concerns about these capillary transmitters and with good reason since unless you know what you are doing they can be very difficult to set up properly. Especially if the installation was done by someone that does not understand anything about capillary transmitters, and by the way if you want to measure crude or levels in high pressure vessels, the only way to go is capillary and not piped wet and dry leg installations.

I am not going to answer your questions one by one but will rather give you advice on the way they should be installed. You also seems to be inexperienced (I could see that with your comment on the long capillaries) with the calibrations of these capillaries so I will give you an example on how to do it and you can then use that and will be able to do your own calibrations on or offline.

Some general info:
These capillary transmitters can be installed anywhere on your vessel. The HP padcell(remote chemical seal) can be installed on the Top tapping or the Bottom. The transmitter itself can be installed below or above the bottom tapping it will only be limited by the length of the capillary. We normally install the transmitter in a place that is easy to access by the technician that needs to do the calibration and repairs afterwards. All this can be done but the normal and standard way to do the installation is to connect the HP padcell to the bottom tapping and the LP padcell to the top tapping on the vessel and then install the transmitter below the bottom tapping somewhere close to the floor level where it is easy to reach. If this happen to be above the bottom tapping that's fine as well.

The second thing that needs to be done is to install flushing rings between the flange of the tapping point on the vessel and the flange of the padcell. These flushing rings must be specially rated for you application since they will form part of the vessel in the future. We normally do the sizing at the same time as for the transmitter needed for the application. There should also be a 1/2" threaded hole on the top and the bottom of the one flushing ring and only one 1/2" threaded hole in the other one, since you will be installing 1/2" straight instrument fittings in both.

If the gasket needed on the tapping of the vessel is RTJ the flushing ring needs to be RTJ as well.(Ask mechanical dept about RTJ's)
These flushing rings can be manufactured by any engineering company out of 316S/S and have a OD about the same size as the vessel flange diameter below the bolts and a ID the same as the ID of the tapping point pipe on the vessel.

Only once you have all these items and gaskets, can you start your installation.

Install the RTJ or gaskets, flushing rings and then the padcellls on the tapping points on the vessel. HP padcell to the bottom tapping and the LP padcell to the top tapping point on the vessel. Coil up the excess capillaries in a neat coil and tie it of somewhere. These excess coils can be tied off any where and will depend solely on where it will be the most out of the way of the transmitter and what will look the neatest.

One more thing to look at during the installation of the flushing rings, make sure the that one hole point straight up and one point straight down on the flushing rings during installation and that the ring with the two holes is installed at the bottom tapping. The one with only one hole should be installed on the top tapping and the hole must point downwards.

Finally mount the transmitter anywhere where it will be convenient to work on it later. Carry on and connect it up and switch on the power to it. Do not make any changes to it since it must stay as you received it from the factory.

Ok now with the padcells and flushing rings installed and the threaded 1/2" flushing ring holes in a vertical line and the transmitter mounted and connected up, you can carry on with the installation.

Take x3 1/2" straight connecting S/S fittings and turn then into all three holes in the flushing rings. Actually it might be better if you do this in the workshop before you mount the flushing rings.
Now connect a piece of 1/2" S/S tubing from the bottom fitting of the top mounted flushing ring to a needle valve about 200mm above the bottom tapping point. From the outlet of the needle valve carry on with the S/S tubing and connect to the top fitting of the bottom tapping flushing ring.
Connect a small piece of tubing and another needle valve to the bottom fitting of the bottom flushing ring.

Ok the installation is done but go back and make sure everything is done as described above.

Ok now to for the calibration. I obviously cannot know what your applications looks like, so I will created a simple and basic one to use as a example.

Say we have a vertical two dome vessel, standing on four legs, 2000mm from the floor to the outside bottom of the vessel and from the outside bottom of the vessel it is another 4060mm to the top of the outside top of the vessel.
The product pressure inside the vessel during normal operations can be about between 10 and 14Bar
There is no lagging on the vessel and the wall thickness is about 30mm so therefore the inside diameter of the vessel is from bottom to top is about 4000mm. The horizontal width of the vessel is also about 2300mm ID.
All this information as well as the exact zero and span positions on the vessel can be found in the design engineer's internal vessel drawings so you might want to get hold of those first before you start the calibration on your applications.
Let's say the bottom tapping point is located exactly 2400mm up from the floor (measured to the middle of the tapping point pipe) and the top tapping point middle is exactly 3200mm measured from the bottom tapping middle of the pipe.
Say in this example according to the design engineers drawings, zero should be 200mm above the bottom tapping point and span should be 150mm below the top tapping point. Start by making marks of these zero and span requirements on the sight glass if one is available on the vessel. After the calibration you will be able to use these marks on the sight glass as zero and 100% and the transmitter reading should give exactly the same % indication as red off the sight glass based on these markings.
The transmitter is mounted 1500mm measured from the floor up.
The information you need to do the calibration is the distance tron the middle of the diaphragm of the transmitter to the zero position on the vessel (LRV) as well as the distance fron the zero position to the 100% position on the vessel, which you already have. In this case it will be 3200 - 200 - 150 = 2850mm(called the span distance).
So to get the distance from the transmitter diaphragm to the zero mark on the vessel is just a mater of add and subtract and you will get it, since you already have the distances from the floor to the bottom tapping point as well as the distance to the transmitter from the floor up. Subtract the one from the other and add 200mm to get to the zero position distance from the transmitter. In this case 2400 - 1500 + 200 = 1100mm

When working with capillary transmitters you do not ever do a zero trim on them but rather use wherever it display when both padcells are open to atmosphere as the zero. This is called the atmospheric zero value of the transmitter and is only used for the initial calibration.
In this installation the atmospheric value will be about -3200 but will not be exactly that in real life since it is currently measuring the filled fluid of the capillaries but let's say it is exactly -3200mm. Write this value down.

Say the sg of the product inside the vessel is 0,93.

Now calculate the LRV and URV as follows in this example:
LRV = the atmospheric value + (200mm x sg of the product)
= -3200 + 186
= -3014mmH2O = 4mA
URV = the atmospheric value + (2850 x sg of the product)
= -3200 + 2650,5
= -549,5mmH2O = 20ma
Input these values with the HART and the initial atmospheric calibration is done.

Because we have a high product pressure this atmospheric zero value can change the moment you put the transmitter on line so we need to do a process zero check as well and if the process zero differs from the atmospheric zero you need to recalculate the calibration values.

To do the process zero the transmitter needs to be online and in contact with the vessel pressure but not the liquid so keep the bottom main isolation valve close and by making use of the flushing rings and needle valve fill both the padcells and tapping point pipes with product gas pressure and leave the needle valve and top main isolation valve open. Now check the value displayed on the transmitter. It should be the same as the atmospheric value but as the product pressure increases in a application the bigger this zero shift will become so the calibration values have to be recalculated since it is more important to have the calibration right for when the transmitter is online that at atmospheric pressure.
To make the correction is a simple matter of redoing the calculations and replace the atmospheric zero value with the process zero value. After this is done put the transmitter on line and do some draining and flushing with the flushing rings and it should be fine and should read exactly as per you makings on the sight glass. You also might want to increase the damping since with long capillaries like in this case, the reading tends to drift up and down a bit.
Good luck.

 

[Mass44]

Hi again,

Sorry made a mistake with the URV calculation

URV = the atmospheric value + (2850 x sg of the product)
= -3200 + 2650,5
= -549,5mmH2O = 20ma

It should be the distance from the bottom tapping point to 150mm below the top tapping point and not the span value so it must be 3200 - 150 = 3050mm,

URV = the atmospheric value + (3050 x sg of the product)
= -3200 + 2836,5
= -363,5mmH2O = 20ma

 

[danw2]

Nice writeup on installing remote seals.

I suspect that Michael1983 will need some luck.

This thread seems to be a separate but follow-up thread to his thread

http://www.eng-tips.com/viewthread.cfm?qid=277416&page=1

with a request for advice on a suitable DP transmitter range to monitor pressures of -32.6 to -10.7 mbar, approximately -13.1 to -4.3 in H2O. That's a fairly narrow span to tolerate temperature drift from long capillaries.

 

[Michael2009]

One more thing to look at during the installation of the flushing rings, make sure the that one hole point straight up and one point straight down on the flushing rings during installation and that the ring with the two holes is installed at the bottom tapping. The one with only one hole should be installed on the top tapping and the hole must point downwards.

Hi. Mass44

Thanks for your explaination it really helpful for me... Ya. you are rite. i an quite new for remote seal system, the system was choose because of the like of my customer. I did a lot impulse line now is the 1st time doing remote seal.

i did not understand about the flushing ring drain. Why should hv to drain hole for flushing ring High side or bottom? and why should the drain hole one facing downward and one upward???

for extra long capilary i see a lot of people are not only coil. They also twist the long capilary both side together and coil it. Why they do that? to get the same temperature effect???

 

[Michael2009]

Nice writeup on installing remote seals.

I suspect that Michael1983 will need some luck.

This thread seems to be a separate but follow-up thread to his thread

http://www.eng-tips.com/viewthread.cfm?qid=277416&page=1

with a request for advice on a suitable DP transmitter range to monitor pressures of -32.6 to -10.7 mbar, approximately -13.1 to -4.3 in H2O. That's a fairly narrow span to tolerate temperature drift from long capillaries.

Hi danw2

This two question actually was come from same project. I follow up this project from one living colleague.
the project remote seal system was seem to be problems.
1. The calibration range are just only -32.6 t0 -10.7 mbar but the transmitter range are -1000 mbar to 1000 mbar.
it was too big to the span. I was worry about the accuracy of the transmitter.

2. Some more it was a remote seal system some capiliary was long about 15m. because the chamber are short and high no place for mounting on top and hv to be ground level. Because it was a chamber on top of vessel to delect high high level of a vessel.

 

[Mass44]

Hi again,

It seems you have a very small level you need to measure very high up on a vessel.

Transmitter ATM value will be the distance from the transmitter to the top tapping point minus the distance from the transmitter to the bottom tapping point, so the max capability of the transmitter only needs to be slightly bigger than the distance between the top and bottom tapping points. In this case 223,30mmH3O

So ideally you need a transmitter with a max range of about 250mmH20 so that is -125 to +125mmH20. Unfortunately we don’t live in a perfect ideal world so you need to find the closest range to that.

Here is one that is pretty close so give Emerson a call and ask if they can supply the Rosemount 3051CD1A with 15m Capillaries. If not another possibility is to replace the capillaries with tubing and then do a wet leg installation with a normal piped 3051CD1A

The max range of the 3051CS1A is 1275mmH20 (compare to your current 10200mmH2O transmitter) so that is -637,5 to +637,5mmH2O (-62,2 to +62,2mBar)

Your calibration seems to be in the region of,
-332,4mmH2O to -109,1mmH20 (-32,6mbar to -10,7mbar)

---------------------------------

About your question about the flushing rings, you will see in the writeup I was trying to explain that you need to connect a piece of tubing between the top and bottom flushing rings with a needle valve in the middle or close the bottom tapping. Later in the writeup you will see when and how we do a process zero and I think you will be able to understand if you read the writeup a couple of times carefully. I cannot explain it any better about why the holes need to face one up and one down.
The drain valve at the bottom of the bottom flushing ring is there for when you close the bottom tapping main isolation valve and needs to drain the liquid out of the pipe before you do a process zero.

-------------------------------------

About the question about the twisting of the capillaries I cannot say what the person had in mind when he did that. I can only assume he was concerned about temperature variations but remember if there is a ambient temperature increase or decrease the effect will be exactly the same on both capabilities and therefore cancel each other out. So the capillaries is not the cause of drifting in capillary installations. You will find that ambient temperature changes have a bigger effect on the electronics of the transmitter and after long periods of direct sunlight on the transmitter you start getting these drifting problems. This happen on normal wet and dry leg installations as well. If you replace the electronic module, the problem disappear until a couple of months later, same story.
Where you do pick up problems is when you work on these capillaries on a moving vessel like a FPSO or ship. Then you will find out what drifting really is and you need to still get stable readings to the CCR. That's when your hair starts to change color.
One trick we came up with to minimize drifting problems is to split the error. If you ever get into a situation and you need information on how to do this let me know since it will be a big one again to explain.

 

[Michael2009]

Hi Mass.


Regarding the flushing ring, I just order one for high and low tapping point with one hole drain. the purpose was drain function like wat you say. I will install the both flushing ring with the drain facing downward for drain liquid function.
Are you flushing ring suggested with 2 drain hole are on downward and upward are for vent and drain function?

do you think one hole for the flushing downward is enough for both side?

for the coiling capilary functhion i also did not so understand why they do like that... But as i know my capilary was SS armound type. Maybe will difficult to twist together. I will follow as what you say to keep coil.


well, mass thanks for your reply..

 

[Mass44]

Hi again,

Please read these sections again of my writeup since it seems you are still not sure why we need the flushing rings.
Don't stress about it, just keep on asking questions until you get there, I don't mind. Remember I also had to learn these procedures from someone else. Once you get the first one right you will be able to carry on on your own from there. Keep on reading the writeup over and over and it will become more clear after a while.

"The second thing that needs to be done is to install flushing rings between the flange of the tapping point on the vessel and the flange of the padcell. These flushing rings must be specially rated for you application since they will form part of the vessel in the future. We normally do the sizing at the same time as for the transmitter needed for the application. There should also be a 1/2" threaded hole on the top AND the bottom of the one flushing ring and only ONE 1/2" threaded hole in the other one, since you will be installing 1/2" straight instrument fittings in both."

ALSO THIS SECTION

"Take x3 1/2" straight connecting S/S fittings and turn then into all three holes in the flushing rings. Actually it might be better if you do this in the workshop before you mount the flushing rings.
Now connect a piece of 1/2" S/S tubing from the bottom fitting of the top mounted flushing ring to a needle valve about 200mm above the bottom tapping point. From the outlet of the needle valve carry on with the S/S tubing and connect to the top fitting of the bottom tapping flushing ring.
Connect a small piece of tubing and another needle valve to the bottom fitting of the bottom flushing ring.(The Drain)"

ALSO THIS SECTION

"To do the process zero the transmitter needs to be online and in contact with the vessel pressure but not the liquid so keep the bottom main isolation valve close and by making use of the flushing rings and needle valve fill both the padcells and tapping point pipes with product gas pressure and leave the needle valve and top main isolation valve open. Now check the value displayed on the transmitter. It should be the same as the atmospheric value but as the product pressure increases in a application the bigger this zero shift will become so the calibration values have to be recalculated since it is more important to have the calibration right for when the transmitter is online that at atmospheric pressure"

Good luck and give me a shout if you get stuck.

 

[Michael2009]

Noted...
and thanks for reply.... =)

 

[roydm]

Mass44,
Very good writeup.
Confirms my belief you need to have your wits about you when using capillaries.
Thanks
Roy

 

[Mass44]

Thanks roydm,

Just a few other pointers on level application that all instrument technicians and engineers should know,

What I have found in the past is that everyone knows how to setup a level with a dryleg and in a open tank and then try to use that knowlage and apply it to all level setups.
It is important to realize that open tanks and dry legs can be setup with the normal sg x g x h formula (correct way is for exp: 1 x 9,81 x 3m = 29,43Kpa, so your UOM needs to be in meter and the answer will be in Kpa directly - you need to make sure the UOF is right otherwise the answer will be wrong)if you need a pressure answer in order to pump up the transmitter to the span value.

I personally don't use the formula anymore since I only have been working with smart transmitters the past couple of years and you don't need to pump up a smart transmitter. You just calculate the LRV and URV with the mm x sg = mmH2O formula as well, and input them with the HART and the job is done.

For any wet leg or capillary applications pressure or vacuum you need to use the method described above in my big writeup, and start to work with the ATM zero and process zero values.

Incidentally it is also not necessary to change the LP and HP leg around for wet leg applications anymore since smart transmitters can all measure in the negative which means that a wet leg and a capillary setup are pretty much the same thing these days, if you use the HP at the bottom and the LP on the top as normal.
Some people working on the boiler drum level applications have a bit of a hard time to accept this since they always had to change the transmitter legs around in the past so you might get a bit of a argument from them, but that is just because they do not understand the difference between a old 4 to 20 electronic transmitters and a new smart transmitter. Old electronic and pneumatic DP transmitters could only measure in the positive.

I have also found it to be mush simpler to start working in mmH2O since you can take measurements in the field and apply them directly to your calculations
(mm x sg = mmH2O)so I never work in any other UOF than mmH2O anymore on all applications. Afterwords you can just change the UOM on the transmitter if you need it to display some other UOM.

Something else I have found in the past couple of years is someone will say they cannot setup the level in some or other application since no one can tell them what the sg value is of the product inside.
There is a very simple way to determine the sg of any product very accurately yourself.

Use the same transmitter (tubing not capillary type)you will be using to measure the level, or an extra one, and measure the exact distance from the transmitter diaphragm to the position where the tubing is going into the fitting on the top tapping on the vessel. Say this distance is 2000mm.
Do a zero trim on the transmitter and then recalibrate your transmitter for LRV = 0mmH2O and URV = 2000mmH2O. If you should now fill this impulse line with clean water the transmitter will read exactly 2000mmH2O. We all know this, so fill the impulse line with the product inside the vessel and it will read a different value. Take the value it now reads and divide it by the 2000mmH2O, it would have red with water, and the answer is the sg value of the product inside the vessel. For exp: Say the transmitter only reads 1970mmH2O when you fill the line with vessel product, 1970 / 2000 = 0,985, so the sg of the product is 0,985.

Have fun - it's not a job, it's an adventure!!

 

[Zinskie]

Mass44,

Your posts have been very well stated and informative. The only addition I might make for ease of understanding is: the tubing that connects the high and low pressure flushing rings through the needle valve acts similarly to the equalizing valve on the standard three-valve manifold of most transmitters, used to calibrate with an on-line process zero.

While I have no direct application for your above mentioned offer, I am curious about your procedure for calibration on a moving vessel.