I have a question regarding establishing a proactive calibration program of critical instruments-particularly Rosemount pressure and flow transmitters. I come from the water utility industry. Wouldnt establishing a reliability and maintenance program for these types of instruments require first that the end users/operators determine/specify the expected performance criteria (tollerance allowed) for instrument readings i.e. how much drift is allowed per State/Fed water quality regs ect..and share that with the Instrument techs.

I think that companies like Rosemount have enough research information (for their pressure and flow transmitters) that knowing expected performance criteria (allowable drift)and failure data (operating temperature and static pressure) the end-user can possibly determine the calibration interval. This would help optimize the calibration frequency. Does anyone have any specific or general thoughts on how to optimize or proactively manage an instrument calibration program? How do the top performers out there manage their instrument calibration programs? Time-based intervals? Calculations using known statistical data?

I'm basing most of my info from the web address: http://www.rosemount.com/document/pds/tpe.pdf

Thanks ahead

Robert

Hullo Robert,

Instruments fail e.g., they drift or go out-of-span due to
1. The design and quality of the instrument
2. The way you operate it

Companies like Rosemount can tell you a lot about item 1, but only you, the operator, knows about how you opertae it and what are the consequences of failure.

For this reason while the vendor can tell you what to do in respect of his equipment, such recommendations are less useful than one that you have worked out for yourself, based on your knowledge of the equipment AND how you operate it AND the consequences of failure in each case.

There is an internationla standard IEC 61508 which can help you both ine the design stage, and in the opertaing stage to identify the correct maintenance activity.

V.Narayan

There are 23 standards documents under the IEC 61508 title: which is the one to get for calibration interval advice in the operating stage?

Ray,

For Process Plants, incl.Power Plants, the applicable derivative std. is IEC 61511, but it will refer to IEC 61508. Calibration of sensing devices is one aspect of the whole issue of managing risks relating to protective instrumentation. But as you are also concerned with logic 'black boxes' and executing elements, the picture becomes somewhat larger.

Once you establish the existing SIL levels and the reqd. PFD (prob. of failure on demand), the test frequency can be derived.

V.Narayan.

A related question is the calibration interval for accelerometers. There are different theories about the required interval, but the truth is a little surprising.

Frequently when we think of the need for calibration, we think of drift and aging in sensors and the asociated electronics. In the case of piezo-electric accelerometers, this concern is actually negliable. The sensitivity of the crystal falls off at a known rate after it has been polarized. There is a falloff of 0.5% sensitivity after 1 day.....another 0.5% after ten days, another 0.5 percent after 100 days and so on. Most vendors age the crystal and calibrate after day 100. As far as the user is concerned, the sensitivity will decrease by 0.5% after 1000 days (nearly 3 years) and another 0.5% after 10,000 days (over 27 years!). So basically a 1% maximum reduction in sensitivity after 27 years from crystal aging. The gain of the associated electronics is set by a single capacitor. Capacitors with NPO dielectric are used here, with essentially no change in properties from either time or temperature. Less than a 1% change in sensitivity over 27 years seems to reduce the need for calibration.

So why calibrate? As Vee said, it's a mater of how an item is used. In the case of an accelerometer, the crystal can crack or break from shock which can also damage the integral amplifier. The incredibly stable properties of the crytal can also change if operated at high temperatures.

Accelerometers used with portable data collectors suffer a tough life and are banged around constantly. For this application we recommend a yearly calibration. The chances of having a damaged accelerometer are just too great and the consequences of erronious measurements too important to skimp on this. Permanetly istalled accelerometers on the other hand lead a pretty easy life for the most part. They are unlikely to face the high shock of getting banged around, so there is little possibility of damage. Calibrating these accelerometers every 5 years is reasonable. or even when a machine is pulled for overhaul. The exception is accelerometers operated near their temperature limit. Yearly or more frequently if their temperature limit is exceeded.

Spintelligent Labs represents CTC, a manufacturer of accelerometers. These accelerometers offer an exceptional value. First, they have a lifetime warranty. If they don't meet specs they will be replaced. Period. No fine print required. Second, CTC will calibrate them annually at no charge. I can see you thinking "Ah hah....that's when they charge me for a replacement." Nope. Free calibration and if an accelerometer doesn't meet spec, you'll receive a new one in its place. Can't beat the value there!

Jon
http://www.spintelligentlabs.com

Jon,

Several good points and advice, thanks. If I may add, consider such devices in e.g., military vehicles such as battle tanks or trucks, most civilian vehicles, mining equipment and one can see that the shock loading can 'jolt' the piezo-electric devices to go out of calibration.

However my point was also about considering the whole end-to end system. It is not enough to worry about sesnsing devices alone.

V. Narayan.

Hi Jon,

Take it easy on the promotion messages.

That was a long road to make a plug.

Terry O

Terry,

I beg to differ and would say it was a short plug on quite a bit of information. I don't believe you will find that concise of an analysis of accelerometer calibration anywhere else.

If you'd like to continue this conversation, please feel free to e-mail me directly: jon@spintelligentlabs.com.

Jon

Jon -

I got a few emails from some of the members - hence the warning.

The spirit of the list has always been set by the members and we are simply the hosts.

You may wish to invite the members who took issue to email you off list to continue the discussion.

Terry O

This is again regarding the proof test intervals of SIS Instruments.
Recently, we conducted RCM analysis of all SIS Instruments and arrived at 2, 3Y PM frequency for field transmitters,against the existing 1Y PM Frequency.
But understand that the SIS Instruments need to be proof tested at frequency ASSUMED in the SIL calculation (which is 1 to 2 Year) to maintain the safety integriity level.

Kindly advice.
Regards,

Dear Sir,

As I dont know what SIS intruments are and how they operate, perhaps you can tell us some details.

V.Narayan.