what could cause a pressure gauge to read low? 1

[amazing azza]

Hello friends!
I have a deaerator. It is fitted with a 0-1 barg analog dial gauge (wika, not sure of model right now). It is reading low. Digital sensor (4-20mA) reads ~0.22barg, consistent with thermometer reading 105-106 C. But the analog is reading ~0.12barg. What could be the cause? Analog gauge is brand new.

 

[roydm]

Could it be the impulse line is water filled? If the gauge is mounted above the connection point it will read lower but by the tables it would need to be 40" above connection for 0.1 Bar
What is the height difference between transmitter and gauge, do they share the same connection point?

 

[zdas04]

Roydm has a very good point, if the two instruments are at the same height, and the impulse line is self draining (and free of traps), and the digital instrument gives you a reading consistent with the pressure then I would check the calibration on the analog instrument.

Many people do not check the calibration (especially the zero) out of the box. I often find that practice to be a mistake. Instruments get pretty beaten up in shipping and in warehouses. When I'm dealing with a 0-10,000 psi gauge I typically don't care about a few psi intrinsic error. If I have to differentiate 0.1 bar from 0.22 bar then I always verify the calibration on a new analog instrument prior to use.

[bold]David Simpson, PE[/bold]
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[danw2]

A liquid filled gauge needs an open vent to atmosphere. Temperature changes will cause the fill fluid to expand/contract and cause an error in the indicated reading.

Wika can have an open/close vent valve on the top of the housing on its higher end gauges.

Others have a rubber plug that keeps the fill fluid inside the case during transport and one is supposed to poke a hole through the rubber plug when the gauge is put into operation.

 

[roydm]

Danw2 has a good point, I never thought it might be a liquid filled gauge because I would use a dry gauge in that application, I only use liquid filled on pump discharge where it's subject to high vibration or pulsation.

 

[amazing azza]

Here is a pic of both the digital transmitter and the analog gauge. Similar installation height. I did not see a hole for liquid filling, also no obvious way to adjust calibration.

Is it possible that the analog gauge is defective out of the box? Or could it be lying at the low end of the range?

 

[roydm]

I assume that's the transmitter fixed to the column at the left, the tapping point is 5 feet above the floor.
The gauge is connected below the floor, what makes you think the two points are at the same pressure?

Sorry it's a long time since I worked on a de-aerator .there are a bunch of trays inside the column so I assume some pressure difference.

If you suspect the gauge is out test it against a column of water 88.44" high (hose full of water) (0.22 bar) and reposition the pointer.
Note - the gauge has a syphon which may give a slight negative effect.
Wika are a reputable company, they won't sell you a defective gauge but it's possible it got bumped or stressed during installation and has a zero shift, check it against a known pressure e.g. water or Hg column, re zero.
Do not turn the gauge by the case, use a wrench on the square stem.

 

[amazing azza]

I assume that's the transmitter fixed to the column at the left, the tapping point is 5 feet above the floor.
The gauge is connected below the floor, what makes you think the two points are at the same pressure?

Under normal operating conditions neither is submerged in water. What they should be sensing is the pressure of the steam in the headspace above the water. I always assumed that gas pressure would read about equal in such an arrangement. But wait, there's more. The top of that dome is vented, so if anything, it should be the digital sensor that reads low due to proximity to the vent...

I guess you're right, the only way to really tell is to test it against a standard. I will think how to rig up a water column for this.

How do you go about re-positioning the pointer? Take off the case and look for adjustment screw? Nothing is showing on the outside and I am hesitant to start disassembling it.

 

[amazing azza]

So I ranged both the analog and the digital gauges today. Used a couple of lengths of pipe filled with water, and the sensor attached at the bottom of the pipe. The digital was right on, and the analog had a consistent 12% low reading across the 0-1 barg range. This still doesn't explain discrepancy reported earlier (the analog measured close to 50% off). I am planning to move the analog gauge next to the digital, use a tee and some elbows on the same siphon tube in hopes of getting them to display similar values.

 

[waross]

An old instrument calibration tech will have a tool for pulling the needle of the shaft.
The tool looks like a miniature version of a tie rod end remover.

The end of the screw is turned down to a diameter a little less than the diameter of a meter shaft.
Link
You are using a 1 barg meter to measure .22 barg. Some meters are very accurate over the entire span but I never expect good accuracy in the lower quarter of the span.
If the meter is accurate at the low end, great, but I am never surprised if it is not. Your error is 10% of full scale. Not great but....check the calibration.
The pigtail may be adding a few percent error also but not much.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[danw2]

It's called a pointer puller.

 

[danw2]

Any metropolitan area has one or more instrument distributors, at least one of whom does gauge work: a shop where they calibrate gauges and/or fit remote seals.

They might charge quite a nominal fee to check out gauge and correct it if its a slipped pointer.

You might want to find out who your instrument vendors are and ask.

 

[roydm]

The gauge you have (0-1 bar Bronze tube) is quite sensitive to rough handling, it may have got bumped during installation giving it a zero shift. Sometimes they have a small screw at the pointer shaft for adjusting the zero however I don't think this one has that so the pointer will need to be pulled off with a pointer puller and moved to the correct position (don't try and lever the pointer off).
Any bourdon tube gauge has adjustments for zero, span and linearity, I'm not sure if they still teach the technicians how to calibrate those.

 

[amazing azza]

so the pointer will need to be pulled off with a pointer puller and moved to the correct position (don't try and lever the pointer off)

Pointer puller... Ok. Is there any way to make one myself if I fail to locate one?

Any bourdon tube gauge has adjustments for zero, span and linearity, I'm not sure if they still teach the technicians how to calibrate those.

If you had any info on this, I would be very grateful. My searches so far have failed to turn up any instructions on how to do this.

I did find one video where the poster bends slightly the omega shaped link inside the gauge to reset a bad zero.

 

[Compositepro]

This is about the most inexpensive type of gauge that would be used in a process plant. It is cheaper to replace than to recalibrate. Field calibration of such gauges is to apply a known pressure and if the gauge reads within acceptable tolerance it is stamped as calibrated. If not the gauge is replaced. Only the manufacturer is set up to calibrate them in an economical way. In plants that have a formal calibration program, a sticker will often be applied to such gauges that says "Indicator only. Calibration not required."

 

[danw2]

How much time can you spend on this? A Wika 0-15 psig 4" solid front liquid filled process gauge on a US web site that lists prices is $122 plus tax and shipping. I don't know what your time is worth, but fabricating a pointer puller and coming up with reasonable pressure source 'standard' (something you can trust) and then experimenting with making the gain/span and zero adjustments is shoveling a lot of time at something that your boss might not view in the best light.

But you asked, so:

1. Homemade pointer puller. This site touts rolling your own, but requires downloading a freeware (read that Malware) app to get the pdf. Caveat emptor.

2. Mechanical pressure gauge calibration procedure:

If Wika has a calibration procedure, they do not publish it in their gauge manual.

Attached is a section from the 1976 Ametek US Gauge manual with illustration and an explanation.

Ashcroft publishes a method for adjusting linkage:

In essence there is a gain/range/fast/slow pointer adjustment and a zero adjustment (pointer).

 

[amazing azza]

Compositepro, thank you for sharing. In your practice, how much of a slope deviation would be cause for concern? By slope I mean - if you plot a calibration curve (true vs measured), how much the slope of the measured line deviates from ideal (slope=1) line (in %).

danw2, thank you so much for the documents!

 

[waross]

Put the meter on the test stand.
Apply a test pressure of 0.22 barg.
Remove the needle and install it indicating 0.22 barg.
If you want better than that, buy a good quality indicating, electronic DP instrument.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Compositepro]

Acceptable tolerance is completely up to the user. How accurately must you know the pressure to keep the process in acceptable control? However, you must define this tolerance in order to select the right gauge to begin with.

 

[MikeHalloran]

After several incidents of bogus low pressures reported because people who should have known better failed to pierce or remove body plugs from liquid filled gages, and/or similarly failed to remove a plug from the 'other' chamber of Dwyer Magnehelics, I now make it a policy to personally witness all pressure tests with which I am involved.



Mike Halloran
Pembroke Pines, FL, USA