Answer / sam

I would not do a installation like this with a piped DP tx
and think that you should reconsider your position before
you do this.

In hazardous or corrosive applications you might need to
make a modification in order to install a instrument that
will contain the process permanently.

If this is not a safe application at the moment you need to
put in a chance proposal with a safety justification and a
proposal for a new instrument that will make the application
safe.

Do not do something just because you we instructed by
someone to do it, you need to satisfy yourself that the
inhalation is safe, reliable and accurate, not someone else.
If it is not, it is your right, duty and responsibility to
do something about it.

In your case the liquid you are working with is corrosive,
hazardous, flammable and will causes violent chemical
reactions with various other liquids. I would recommend you
modify the installation to a capillary DP tx with chemical
seals(exp-Teflon).

Find out from the liquid supplier what type of material is
resistant to the liquid and ask you Tx supplier to supply
you with a capillary type smart transmitter with this type
of material chemical seals on the pad cells.
The supplier will ask you the length capillary length
needed. The capillary lengths should always be as short as
possible to prevent drifting, so look at the installation
and decide where you are going to install the new cap DPT
and double the distance from the transmitter to the top tap
off point on your vessel to get the right capillary lengths.
Do the calibration as normal for a capability type DPT and
forget about the vacuum inside since it is irrelevant in
your calibration.

Just some important calibration information when working
with capillary DPT's.
After the installation, open both pad cell to atmosphere and
have look at the displayed value. When you do a process zero
you might find that the displayed zero have shifted slightly
from the atmospheric zero. It is therefore important to
check the process as well as the atmospheric zero especially
if you are working on a small span. This small shift will
cause big inaccuracies if you do not work from your a
process zero. (yes, similar thing to the old pneumatic DPT
static alignment problem)
To do the process zero you might need to install a piece of
tubing temporary between the HP and LP tap off points in
order to get a equal vacuum on HP and LP side of tx. That is
if you cannot drain your vessel and just keep the vacuum
pressure inside the vessel.
(We normally install 316SS flushing rings for this purpose
but you need to look at the corrosive properties again
before you install these SS flushing rings. You might have
to order them in a similar material as what you used on the
pad cells.)
Use the displayed value as your process zero reference and
just add your measured mm, multiply by 1,08(sg for acetic
anhydride), to get your L/URV's.
If there is a difference between atm zero and process zero
always use process zero for the calibration since the DPT
will measure the level while under vacuum and not at
atmospheric pressure.

Something else you might encounter in a application like
this is that this process zero could fluctuate slightly due
to the variance in vacuum and you might find it difficult to
read. I had similar experiences before on a FPSO where the
whole vessel moves all the time and therefore the process
zero keeps on changing all the time. I have compensated for
that by standing there and watch the reading fluctuating for
about 15 minutes and write done the lowest and highest
readings and then use the average in the middle as my final
process zero.
This is called splitting the error. Your final level
indication after your calibration will fluctuate as well,
but since you have split the process zero error, the reading
will be pretty accurate and you can apply some damping on
the tx to stabilize it some more.

Below is a previous discussion on a similar application but
for non hazardous liquid and no way to modify the piped DPT
installation to a better type of instrument. You can have a
look at it for information but again, in your case you
should not use a piped DP transmitter.

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

First of all there is no difference between a vessel with a
process pressure of 10 or 50Bar or a vessel with a vacuum.
These pressures cancel each other across the LP and HP legs
on our tx's, and are therefore not taken into account in our
calibration.
To explain in more detail:
If there is 100Bar on the LP side there is also 100Bar on
the HP side so the DP across the tx is still zero.
If there is 750mmHg vacuum on the LP side there is also a
750mmHg vacuum on the HP side so the DP across the tx is
still zero.
So work with the transmitter as if there is no pressure or
vacuum in the vessel.

The best transmitter to use in a vacuum application like
this is a capillary type, but according to your question it
seems you already have a piped transmitter in place. This
makes it a bit more troublesome to do, but by working
carefully and accurately you can achieve accurate and
reliable results.

Take your time, this is one of the MOST DIFFICULT DP level
setup's you will ever came across in any industry.

It is always better to use a wet-leg since condensation will
cause your DP to chance in time. If I look at the sg of the
product you most probably are working with hydrocarbon
condensate, so you would want to install a wet leg in a
application like this.

Process zero with LP leg filled:
1st open both legs to atm and do zero trim. Even better if
you know how to do a factory reset and then do a zero trim.

Connect a 1/2" T-piece to top of LP leg just on the bend
before it goes to the top tap-off point with a needle valve
pointing upwards.
With main process isolation valves still close, fill LP leg
to max and close needle valve.
Try to use glycol since it's density is higher than water's,
and will prevent contamination of the wet leg.
You can also use glycerin or diesel.

Close 5-way manifold equalization valve and open main
process isolation valves and then only open both isolation
valves on manifold.
Open needle valve and fill again LP leg as much as possible,
start closing needle valve slowly and keep filling
to make sure lp leg is filled properly. Playing with the
main LP isolation valve and the needle valve will give best
results to get the LP leg filled to max.
This should put you in the situation that your tx has been
zeroed at atmospheric pressure, so the vacuum pulled
on both sides of diaphragm now and LP leg filled to max,
should now give a accurate process zero to work from.
Write down this displayed value.

What ever this value is is not important you will use this
reading as your process zero reference point to work from
so it can be anything as long as you are sure this is a
accurate and reliable process zero.
To double check if the process zero is good, isolate the tx
again and open it up to atm again not draining the LP leg.
It should still give a zero indication with both sides open
to atm.
Put it back on line and make sure the LP leg is still filled
to max by making use of the needle valve and LP main
isolation valve again.
You should be back at the previously displayed value. Do
this a couple of times to make sure you get to the same
values every time. Only then can you be sure that your
process zero value is reliable and accurate.

To calculate the LRV and URV:
Like I said I am assuming now that the tx had been installed
300mm below the bottom tap-off point.
I am also assuming that the bottom tap-off point is zero
position and the top tap-off point is 100%

You should now have something like (+/-) -750mmH2o on the tx
display.
Let's say the value is exactly -750mmH2o.

Measure from the middle of the tx's diaphragm to the bottom
tap-off point. We make this say 300mm
Measure from the bottom to the top tap-off points. We make
this say 500mm.

Calculation:
LRV is -750 + 300 = -450mm x .95 = -427,5 mmH2o
URV is -750 + 300 + 500 = +50mmH2o x .95 = +47,5 mmH2o

Modify your L/URV's to these new values. There is no need to
use a hand pump when working with smart tx's, just modify
the values with the HART, make sure it's on line and give it
back to production.

Since it is such a small span the level might be to
sensitive so you might want to increase the damping as well
on the Tx.
Good luck.

please send me hpcl placement papers

13 Answers   HPCL, L&T,

---

please can any one tell me clearly about pid tuning i am confusing about this .how to tune pid easly?

0 Answers   HP, Siemens,

---

why we use 24v dc in instrument rahter than 230vac????

4 Answers  

---

What is hook's law

4 Answers  

---

WHY THE DEGREE OF VALVE PLUG AND VALVE SEAT IS SLIGTLY DIFFERENT I CONTROL VALVE TRIM DESIGN

0 Answers  

---

---

I have one DP transmitter of range 0-500mmWC i have to calculate its range in M3/hr for flow measurement so whats the formula.

4 Answers   ABB, TICB,

---

how to do dynolab wind speed sensor loop test, how to simulate value,i have simulate mA 4-20mA but not responding in DCS faceplate, please kindly advice me any one.

0 Answers  

---

deference between DCS and PLC including date processing and architacture.

11 Answers   Jindal, Shell,

---

i feel sq. pipe is better than round pipe pls give comment for this. advance thanks for reply

3 Answers  

---

what is flow

2 Answers  

---

Explain Three way boiler.

2 Answers   L&T,

---

Why are the uses 24vdc in industries

0 Answers  

---