(im)possible temperature measurement - Any ideas?

[rdrast]

Okay, I've been searching around, and not coming up with many good solutions.

Problem: I need to run a temperature profile of a cable core, through a closed system. The overall length of the measurement is a minimum of 700 feet, optimum measurement would be up to 1800 feet. The temperature range is 70 to 450 F.

Possible solutions I've thought of:
1) Bury a TC in the core, and run a LOT of extension wire - concerned about losses in the fine gauge wire, and the difficulty of the physical hookup.

2) 'create' a fake TC Junction with a single wire, use that and the outer cable core as the conductors. Here, I'd have to profile my synthetic junction to calibrate it, but I still have attachment problems, loss problems, and the influence of temperature on non-standard conductors (Cu and something else).

3) Find some ultra-miniature, self-contained, temperature datalogger, that would log every second or so, and bury that in the cable. The problems here seem to be size, and the actual temperature limits of the dataloggers. I could perhaps get around the size (and case-melt) issues by cutting the electronics out of the case, and burying just them in the cable.

4) Put a specific defect in the cable, like a dissimilar metal crimp, and attempting to use some sort of TDR measurement...but I have no idea if that would be at all sensitive to temperature, or if my equipment would be sensitive enough to qualify it later.

In all cases, I can run an ambient temperature baseline test, before the at temperature test, to generate baseline losses and such.

Any thoughts? I'm still mostly looking at going with option 3, if I can find a suitable device. In this case, the device can be completely sacrificial, as long as I can read the data out one time.

Regards,

 

[leitmotif]

Was using light bulbs in a class setting. Did all calcs based on label ratings. However when cold the resistance of light bulbs was very low. They reaquainted me with the temperature and resistance relationship.

I wonder if you can measure voltage drop across cable. If you know load amps and you know cold resistance could you then derive a temp to Resistance table to determine average temp of cable?

Dan Bentler

 

[BryanG]

I am probably not understanding your question, my excuse is that it is getting late and I am about to head home.

The resistance of cable is affected by temperature, so can you measure the resistance of the cable. It would need to be very very accurate resistance or maybe the volt drop along the cable would be easier.

http://www.allaboutcircuits.com/vol_1/chpt_12/6.html

Bryan

edit: leitmotif types faster than I do!!!!

 

[rdrast]

The problem is I need to measure the actual temperature, of a moving point, every foot as the cable is being conveyed down the steam tube. No overall average will do, we are trying to determine the actual (and not theoretical) transfer characteristics of the steam tube profile vs. the cable profile.

So, I need to have a sensor at a single point in the cable, and then either in real time, or after the fact (datalogger) see what that point temperature profile is through a 600 to 1200 foot run.

 

[harryting]

Hard to visualize what you are trying to do. how about a picture?

 

[OkiePC]

http://www.phaseivengr.com/p4main/S...Default.aspx?gclid=CKTnx7ztmqECFRUeDQodkmTAOw


Just a thought...

EDIT: the ones in the link don't appear to have the temp range you need, but perhaps there are other RFID sensors that could meet your specs...

 

[Brownhat]

What about an embedded RTD? If you've three conductors in your cable which you could tie it to, you would be able to compensate for conductor resistance. You can get sensing element packaged as thin film or a glass bulb.

 

[leitmotif]

Why do all this?

Is this a test setup to determine operation parameters for cable in an installation?

Every foot is going to be measured? How and where are you going to put 1200 sensor leads in a cable?

Could you do this with ten foot or so chunk with RTD (or??) every foot or so. You know travel speed of cable thru steam tunnel, should be able to get temp rise as it travels thru. Then correlate to actual cable?

What is intended application? Down a hot geothermal well for submersible pump?
Intriging
Another thought you could put a tube inside the cable and then pull the temp sensor thru it to keep the sensor in the right spot (ie distance) in the cable as it travels thru.

Dan Bentler

 

[Brian123]

Why do all this?

Is this a test setup to determine operation parameters for cable in an installation?

Every foot is going to be measured? How and where are you going to put 1200 sensor leads in a cable?

Could you do this with ten foot or so chunk with RTD (or??) every foot or so. You know travel speed of cable thru steam tunnel, should be able to get temp rise as it travels thru. Then correlate to actual cable?

What is intended application? Down a hot geothermal well for submersible pump?
Intriging
Another thought you could put a tube inside the cable and then pull the temp sensor thru it to keep the sensor in the right spot (ie distance) in the cable as it travels thru.

Dan Bentler

The problem is I need to measure the actual temperature, of a moving point, every foot as the cable is being conveyed down the steam tube. No overall average will do, we are trying to determine the actual (and not theoretical) transfer characteristics of the steam tube profile vs. the cable profile.

So, I need to have a sensor at a single point in the cable, and then either in real time, or after the fact (datalogger) see what that point temperature profile is through a 600 to 1200 foot run.

4 wire RTD should completely compensate for any lead wire resistance. RTDs also look to operate in the range you need to measure.

Brian

 

[Gil47]

Something I read a while ago may be of interest and that was a fibre optic temperature sensor I think it worked like an infra red sensor it just did the sensing at the end of the fibre.

 

[rdrast]

Why do all this?

Is this a test setup to determine operation parameters for cable in an installation?

Every foot is going to be measured? How and where are you going to put 1200 sensor leads in a cable?

Could you do this with ten foot or so chunk with RTD (or??) every foot or so. You know travel speed of cable thru steam tunnel, should be able to get temp rise as it travels thru. Then correlate to actual cable?

What is intended application? Down a hot geothermal well for submersible pump?
Intriging
Another thought you could put a tube inside the cable and then pull the temp sensor thru it to keep the sensor in the right spot (ie distance) in the cable as it travels thru.

Dan Bentler

Let me try to clarify...
It's a continuous cure process, with the cable moving down through the steam tube. We would like to at least profile the absolute core temperature (at one point on the cable, moving with it) from its journey from the entry into the tube, to the exit (and then through the cooling legs as well).

An embedded, retrievable data-logger would do that, I can later correlate the position to the readings based on time.

Basically, I want to profile the temperature at one point in the cable, as it moves through the curing system.

If I must physically construct a cable, with a signal conducting center-wire, I can do that, but it would definitely make for difficulties in the sense-cable connections, and extend the length of the overall testing cable significantly.

 

[milldrone]

rdrast,

See if this works for you.

http://www.opsens.com/en/oilandgas/solutions/sensing/

If your using google to search, try including downhole as a search term. In the drilling industry, they use cables for different measurements all the time.

 

[jimfun71]

I have to ask why it's so imporant that you use only analog data? We transmit critical real time data over many thousands of feet. You can use any HEART speaking device. We use MTS gauges but there is a multitude of HEART devices out there. Rosemount is probably the best but also the most expensive.
The gauges take real measurements convert to a digital signal and then send the data via a HEART network that you can read at will from any other device that can speak HEART.
This might be a more viable solution for your needs.

 

[OkiePC]

I believe jimfun71 meant HART.

 

[rdrast]

I have to ask why it's so imporant that you use only analog data? We transmit critical real time data over many thousands of feet. You can use any HEART speaking device. We use MTS gauges but there is a multitude of HEART devices out there. Rosemount is probably the best but also the most expensive.

Cost isn't the issue, SIZE is. I need to bury the sensing device inside of a cable, RF devices that transmit are also impossible, as there is no way to receive the data.

That fiber system would be good, but again, size is an issue.

/sigh.