Type of Temperature Sensor to use for measuring Pipe Temp during Hydrotest 1

[Smitty532]

I'm a young engineer and I've been working on the validation of Hydrotests. I've used mass balance calculations to assist in validating the Hydrotest. However, this technique requires accurate temperature readings to ensure the calculations are correct. As of right now, most Test Logs I've reviewed show an accuracy of either +-1% or +-1 degree F. Also, the readings only go to single digits, no decimal places. I want to update our procedure to require higher accuracy and a higher resolution so that my calculations can be trusted. I don't want to validate a test when I'm uncertain of whether or not there's a leak in the system. Could someone tell me what type of temperature sensor could achieve what I'm asking?

NOTE: I'm completely willing to accept advice from anyone who believes I'm going about this the wrong way or just wants to give some professional advice.

Thanks,

Smitty

 

[btrueblood]

You need a platinum RTD to get below 1 F accuracy (thermocouples are generally only good to +/- 2 F). When you find the right grade of RTD, you also need to look at the precision of the measurement circuitry. While generally I trust Fluke, I do know that their "RTD reader" has crappy precision relative to high grade RTDs. All of this costs money.

 

[Smitty532]

It looks like those will work just fine. I'll start looking at which ones would be most practical and make most economical sense.

Thank you @btrueblood,

Smitty



Smitty

 

[LittleInch]

Hydrotest of what exactly?

And to what code?

If you can get a copy BS PD 8010 has a good section on validating test results.

Normally you get strength hydrotests which really don't need much info other than pressure - pump it up to the required test pressure, hold it, If it doesn't break then job done.

Leak tests on the other hand are much more difficult as often the leak will be a very small amount of the test medium and often a long time period is required to eliminate temperature deviations.

Things like long pipelines will never get a uniform temperature so the accuracy to this level just isn't needed, plus you're only ever getting temperature at one location.

So whilst I truly admire your search for greater accuracy, you might need to take a step back and see if it is actually going to help or not.

Validating a test where the pressure falls or even rises is often very difficult and whilst the temperature change can be very significant, the practicality of life tends to come into play and in many circumstances people just leave the test on for longer to try and get a uniform pressure or allow any diurnal effects to even themselves out.

but good luck and hope this helps.

LI

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Smitty532]

@Little Inch
The Hydrotests are conducted on Natural Gas Transmission pipelines.

My pipelines are regulated by 49 CFR Part 192. I've been relying on the Pipeline Rules of Thumb Handbook for my volume calculations for validating that there are no leaks. With that, I've used a method I learned for calculating the propagation of uncertainty called the Kline Mcclintock method. Also, it is the only method I know of to calculate uncertainty given a single data point. Maybe it is naive and impractical to achieve what I'm trying, but in pursuit of knowledge I've decided I'd at least try to accomplish it. I've searched these forums for months for this answer and no one even mentions uncertainty associated with measuring equipment. API 1110 references AS/NZS 2885:5:2012 for methodologies but my superior struck me down saying they don't want to use a foreign standard (like how do you fight that logic??). I've actually never heard of BS PD 8010, but I'll try finding it. Unfortunately, it may end the same way as me trying to acquire AS/NZS 2885:5:2012.

As for the long pipelines will not achieve uniform temperature, I will admit that I don't have any working experience with actual Hydrotests, but many documents and information shared with me suggests that variation of temperature for buried pipe is very small (most of my pipelines are buried by the way, I should've mentioned that). If there were notable temperature variations (going through a lake), could a second temperature probe be applied and then a weight average be used? Now practically, that may be a waste of time if the pipe in the lake is a small section of the test section.

I've acquired a testing procedure from the California State Lands Commission that has very stringent requirements for Hydrotesting. Much of what it says mirrors what I'm saying, and if I remember correctly they require all Hydrotests to be approved by them. Now that procedure is much stricter than what is required of us, but if there isn't too much of a cost burden, then I'd like to improve our process. They also have a spreadsheet that they use to validate the test. If you ask for it nicely, they will provide a free copy of it to you (or if you curious as to what it looks like I can send you a copy @LI). I'll attach the procedure so you can see what I'm talking about.

Perhaps the whole idea of what I'm doing doesn't conform to the idea of just pumping out as much successful Hydrotests so we can keep the project moving forward, but I'd like to believe that operators should do everything within their power to make sure they're operating safely. If it's just a couple hundred bucks more per Hydrotests (cost of improved temperature sensors), I feel like it should be expected to take that extra step (from what I've been told these Hydrotests can cost around $500k).

I've been reading a lot of your posts @LI and I'm glad you replied to mine. I respect the engineering thought process you go through and I've learned a lot these past few months just reading through these forums.

NOTE: I don't have any practical experience with Hydrotests. All I know is what I've been told and what I've been able to ascertain while reading through these forums and other online resources. Unfortunately that doesn't make me quite qualified to come up with a solution on my own. Let me know if any ideas I have may be incorrect or based on false assumptions.

Thanks,

Smitty

Smitty

 

[LittleInch]

Smitty532,

That was a very measured response and I apologise if I came across a bit narky, but you never know what the background is a lot if times to the issue or the person.

I've never seen such a tight measurement of temperature on a pressure testing spec before, but what you're really after is repeatability which is sometimes quite different.

It is really the change in temperature you need not so much the absolute number.

As said before, the key issue is simply with the number of locations you can actually monitor and whether these are representative of the entire pipeline. They need to be a reasonable distance from the end to avoid end effects and buried to get the most accurate number.

Some pipelines just go through too many different soil types to get a good reading.

The post below has some good points, which include, after a stabilization period of say 24 hours, pressure up and monitor and keep water in and water out to maintain pressure and over 24 hours if the volume ends up the same then you're good to go.

https://www.eng-tips.com/viewthread.cfm?qid=393924

The relevant bits from PD 8010 are below.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[pennpiper]

Go to your local Hardware Store and buy a Pistol Grip, Lazar Temperature Gage.
Such as:

https://www.amazon.com/Etekcity-Lasergrip-774-Non-contact-Thermometer/dp/B00837ZGRY

More:

https://www.google.com/search?q=pis...jAhWPjFkKHXIvCR0QsAR6BAgFEAE&biw=1024&bih=657

Sometimes its possible to do all the right things and still get bad results

 

[btrueblood]

Yeah, those are super accurate, and not at all skewed by surface condition, ambient light or user stupidity.

 

[Smitty532]

Thanks @Little Inch for the advice. Really good points and I'm going to take them on board. I'll get back to work I start getting things done!

@pennpiper I forgot to mention in my OP but this is NG transmission piping that is buried.


Smitty