Selection of Instruments in petrochemical industry

A

Thread Starter

Andy

Hi,

I need to know how selection of instruments takes place during engineering in petrochemical industry?

Having different kind of process, how instrument engineer decide for a particular process (Liquid/gas/single phase/multiphase) a particular transmitter (say considering flow, temperature, pressure or level) shall be suitable for?

and do we have to study each instrument specification for this? but still the best measuring instrument is finally decided on what parameter??

There are many types of flow,temperature,pressure and level transmittes?.. how i am goona decide which one is best suitable for the current process application, i don't care if its sound strange, but i am deep into this, Could you please help?

of course the next question will be, the installation of instrument?? The position, elevation, placement/mounting on piping network or field especially the valves, how much distance shall we consider from pressure vessel/tank? how we take care of this to get best measurement from instrument without any error??

As they say, Good engineering practices lead to good engineering design. but i feel seniors are well habitual to engineering practices, so much that they unable to explain the reasons behind practice....

Waiting for your response....

Warm regards to all,
Andy,
 
B

Bob Peterson

The reality is that in many cases selection of instruments (and really most equipment) is deeply ingrained in what has worked well in the past and what people are familiar with. It is just too expensive to make a bad choice when a good choice is known.

The other thing is that most larger plants have standardized on particular brands for various reasons. That tends to nudge one toward the solutions offered by those venders.

The "seniors" have been around a while and run into all the problems and want to avoid the problems they experienced previously. It is difficult to put some of that into words sometimes.

There are standardized practices suggested by such organizations as ISA and API that also come into play.
 
Thanks.

But i m still searching my answer in the reply..
i would appreciate if some one can share his experience and a detail note on this very significant topic, that really nagging in my mind terribly these days.

Andy,
Instr engineer
 
Once upon a time, when end users had plenty of engineers and though there were multi-national suppliers, each company usually had its own sales and specialist staff.

A new sensor would involve the engineer searching through trade Journals, Kelly's directory etc for instruments and manufacturers. Most engineering groups maintained a library of data sheets etc. to search and Purchasing would have a list of existing suppliers, usually single product suppliers.

Then the specification would be sent to a short list of companies potentially able to supply.

Up to date information would be asked for and quotations.
Then the engineer might shortlist a number of companies, make telephone contact or receive visits from sales engineers who would be specialists for that product.

The engineer would then decide on the optimum solution and which company he wanted to buy from and send a requisition to purchasing. If Purchasing were independently minded the might also send out for multiple quotes. Engineers might then ensure their choice by specifying as essential some feature (often irrelevant) that only the preferred supplier could match.

However, this whole process is time consuming and expensive and purchasing ended up with a very long suppliers list. This too is expensive to maintain with credit checks, quality audits etc. and of course, spreading 1000 orders amongst 800-1000 suppliers dilutes their buying power.

SO many major companies would have, for "standard" items, a preferred vendors list. Note the word "preferred". It meant that if the engineer could justify a supplier not on the list he could buy from them.

Today there are very many fewer engineers in the end users engineering group. Purchasing want a much shorter vendor list and they want to focus their 1000 orders on as few suppliers as possible so they can extract discounts. But we also have the rise of the "single source supplier" and "strategic alliances". This means that the engineer simply defines the specification for what he wants, submits it to purchasing who send it to the single source supplier.
The supplier will then quote the best fit from their range, if it will do the job, or decline to quote if they cannot supply.

But do not expect that he instrument supplied will be a perfect fit. Nor even the optimum choice. It may do the job but it may also be the most expensive version of the selected technology. It may also not be the best version of that technology and that technology may not be the optimum technology for the application.

BUT this is a necessary consequence of cutting back engineering staff. It also means that general sales enquiries will be dealt with by a sales office which handles all the suppliers products. There are no specialists. The sales engineers will use a computer selection program. If further support is needed it is available but less easy to access.

The consequences are perhaps best illustrated in this article at Control Engineering:
Flowmeter Selection: Right size right design"
(http://www.controleng.com/index.php?id=483&cHash=081010&tx_ttnews[tt_news]=6047)

According to this 70% of flowmeters (and one assume similar for many instruments) are the wrong size and the wrong technology.
In my opinion they ought not to say "wrong" but "sub-optimal". The specified meters will do the job pretty well but they are not necessarily the ideal choice. But finding the ideal choice is not something most companies can afford to do.

I think this outcome is a consequence of the way instruments are specified and bought today and that despite often paying higher prices and getting a sub-optimum choice, the commercial benefits are such that this is the way the industry now works
.
This is fine if you want one instrument for a single application.
Note also that single source suppliers are more and more like supermarkets. They supply everything from biscuits to Computers. But don't expect to see the full range of any product. You will be offered a choice between fast sellers.

Manufacturers don't like to make multiple technologies with niche market applications when they can focus resources on a "universal" solution. This is why flowmeter manufacturers will often offer coriolis. It may be more expensive and it may often not be the ideal choice but it is safe.

The more instruments needed e.g. for multiple applications, the more the engineer can justify moving away from this model. Logically OEMs should always seek the optimum choice because the cost of finding it is justified. But for the single application and the small engineering team, you get what you are given.
 
Correct me if i am wrong you mean selection of right instrument is the job of product specialist/technologist?
not in scope of instrument engineer?

so you agree on this node?

whats your view on mounting/installation of instrument as per process requirement?
depending on the nature of stream?

Valves distance from tank/vessel? etc...

Regards,
Andy
 
Andy,

Please read book by LIPTALK for instrumentation engineers.

it is more or less practical applications.

regards
pan
 
You might consider looking at Applied Instrumentation in the Process Industries by Andrew and Williams.

It is a 3 volume set, the 3rd volume is reference tables and resource material.

The first edition was mid 1970's, 2nd edition early 1980's, so they're somewhat dated, but many aspects of instrumentation remain the same.
 
I simply mean that few engineers can now devote the time, officially, to acquiring the skills to make an independent assessment themselves of the right sensor for their application and as often as not, with individual instrument inquiries, the inquiry will not be processed in the manufacturer's general sales team by a specialist.

Both become dependent to some degree or other on the skills (and other influences) that go into the product selection tools.

Nor can one safely rely on simply reading then literature. Literature can be misleading since it usually talks about performance under calibration conditions. Performance may also be presented following a convention - and this can mean that two different technologies appear equivalent in the literature but in reality and in some applications one will perform far better than the other.

The sensor you get will rarely not do the job intended. But it will quite often now not be the optimum for the job. The compromises that get made (and even under the best of approaches every sensor is a compromise between what you really want - which is a custom designed sensor for the application - and what you will settle for, e.g. you really need 1.5" and you choose between 1" and 2"/ It is a compromise. It is just that the system is likely to make some further compromises.

Unless you have multiple requirements or the application is very challenging, getting the attention of the specialists is going to be more difficult than it used to be. And difficult to justify.

So how to get past this?
By taking an interest in sensor technology above and beyond that sufficient "to just do the job". Use opportunities like this to actually talk with whoever you can. Try and get references and take them up.
Read the trade journals.

There is a lot you can do as background but less that you can officially do in the course of sourcing a sensor because it costs money.
 
Top