'SKIN' temprature measurement

[AYODEJI]

I work in an LNG/LPG/condensate production plant in Nigeria. I am presently faced with the challenge of specifying a surface mounted temperature transmitter for a 6" line in our plant. operating temp is between 20 and 80 deg C. I am considering a type J thermocouple (range 0 to 550dec). Will this work ok?


I need help on manufacturers of this type of equipment. Can you also tell me about possible operating and reliability problems as this is the first use in our plant? We are going to use the signal for plant safety (shutdown) purpose.
Thanks for your help
Isaac

 

[jbartos]

Suggestion: Normally, manufacturer hardware is selected and Representative is contacted for:
1. Availability of products
2. Lead time
3. Product literature
4. The conceptual approval of the application
5. Other product or interface compatibilities
6. etc.
Which manufacturer are you dealing with or preferring?

 

[cbarn24050]

Hi, thermocouples are useless at this temperature use a pt100 sensor.

 

[AYODEJI]

Thanks for the guidance. I really do not have a manufacturer in mind. Can you suggest 1 or 2 for me to contact?
Thanks again

 

[jbartos]

Suggestion: Visit

http://www.thomasregister.com

and type Transmitters: Temperature under Product or Service, which will return 188 companies to contact

 

[jOmega]

jbartos:

Josef,

How helpful do you think the Thomas Register will be
to someone in ....."I work in an LNG/LPG/condensate production plant in Nigeria."...... ????

 

[AYODEJI]

Very helpful!!!
Thanks jibarto.
Ayodeji

 

[jbartos]

Comment: I saw a beautiful picture showing people in a jungle using cell phones (no weak signals there).

 

[jOmega]

Spot on, JB

 

[SS7]

Suggestion:
Please evaluate your need, i.e. Do you need measurement only? Field display? Remote display? In case of failure of sensor what do you need to get indicated etc.
After which you can find plenty of various ways to tackle it. Post here such details we cn help you out.

Best luck

 

[jbartos]

Suggestion: Operating is dependent on the product type and manufacturer support. The reliability problems can be judged by the product Mean Time Between Failure (MBTF), Mean Time To Repair (MTTR), life-cycle, etc. Some manufacturers have these data available to their products.

 

[ScottyUK]

Thermocouple will work ok, but an RTD will give you better resolution over your range of temperatures. If you want ultimate accuracy, use a 4-wire type, otherwise I suggest a 3-wire type is your best option.

 

[cbarn24050]

Hi Scotty, your a bit late with this post. A thermocouple doesn't work at this temperatur because there is no temp difference between the hot and cold junctions and therefore no output. The instrument then has to rely on the cold junction compensator to produce a reading. There is no reason for a 4 wire pt100 to be any more accurate than a 3 wire.

 

[jbartos]

Suggestion: Visit

http://www.miinet.com/whatsnew/articles/wutt_1000.pdf

etc. for more info

 

[don01]

hi all
I just note with interest cbarn that thermo couples will not work at temps between say 20 ~ 80 C Im sorry to say that this is not the case. in fact they will, BUT and this is THE BUT, the linearity is not as good at the ends of the curves and it is normal to select thte TC or RTD according to the range you are measuring. I would suggest that for a limit of 80C you could find a number of linear devices
Any way not with-standing this, I would go for the pt100 device but I have used type j to indicate temperatures in our work shop ( air conditioned) and drink fridges. and while they were not NATA cert they did agree with most test gear we had.

Regards
Don

 

[cbarn24050]

Hi don, Its not the temp, but the temp diff that counts. In this application there is not much, therefore allmost all the reading comes from the cold junction compensator which is not usually very accurate.

 

[ScottyUK]

Hi cbarn,

I thought we were agreeing on too much lately!

A themocouple will *definitely* work at this temperature. The CJC should be capable of providing accuracy down to the 0.1K level, which is adequate for most applications. The cost of providing a really good CJC starts to become excessive if you want better accuracy than 0.1K.

The RTD is a better choice for the application because it will almost certainly work out cheaper than a high-end thermocouple amplifier with a good CJC for a given level of accuracy.

I agree, 4-wire RTDs should not be better than 3-wire, but this assumes that the circuit has perfect symmetry, and that goes to hell if a slightly high resistance termination creeps in on one leg or the other. One of the downsides of living in an imperfect world! 3-wire is fine for most control applications.

 

[cbarn24050]

Hi scotty, allmost all temp controllers these days use a forward biased diode for cjc. You'll go a long way before you find 1 as good as 0.1 deg.

 

[jbartos]

Suggestion: Visit

http://www.omega.com/temperature/Z/zsection.asp

for useful info about temperature measurements