Thermocouple drilling/extraction

[herthbn]

Hello. I’ve got a project on hand at work that I need some ideas for. My team and I have ran out of talent! We have a 1/4” dia. X 47” long thermocouple probe stuck and broken off inside of its 1/2” ID steel thermowell. This probe is used to measure internal metal temperature at the inside of our steam turbine. After several failed attempts at extraction, we are now left with the probe broken off about 8” up this steel pipe (thermowell). That leaves about 38” of probe still left in place. For a little bit of background, we tried pulling on it with a cable grip and slide hammer, tried pulling on it with a small come-a-long, tried twisting it, tried putting heat to it…. All to no avail. Now it’s broken off so far inside that we can’t get a bite on it.

The last plan of action I can come up with is to drill out the probe because I am assuming that the end of this wedged so tight way up in there it would not come out with normal extraction methods. It gets to over 1000 deg during normal operation so I’m sure there is some galling and debris wedging it in place. It hasn’t been removed since 2007 so that’s 16 years of thermal expansion and contraction seizing this probe in place.

This thermocouple is basically a 1/4” OD thin wall stainless steel tube with 6 small gauge wires and epoxy filler inside of it. It is mounted vertically inside of this steel well. Since the well is 1/2” ID and the probe is 1/4” OD, there is an air gap all the way around it with the exception of the very end where it has a slight taper according to the very vague drawing I have seen.

I’m on to our last ditch effort for removal. Since it’s mounted vertical, gravity will work in my favor if I can get some type of movement. I was thinking about ordering some extra long HSS drill bits to try drilling it out slowly. Maybe start with a 1/2” or 3/8” bit and see where thst goes. No one has cobalt drill bits that long readily available so I’m left with only HSS. Does this sound like a plan that might work, or would anyone else have a better idea? Are there any better tools or methods that could tackle this job?The hard thing is we cannot damage that outer steel casing well because it’s our barrier between 1000 deg steam at 2400psi and ambient conditions. Any tips on drilling it? It’s all going to be done with a cordless drill. It’s about 30’ up and I’ll be working off of scaffolding.

Any ideas would be much appreciated. I’m all ears on this one. This probe reads a very important temperature for starting our unit up. Without it, we are flying blind and guessing on temps.

 

[Gordon Heaton]

You want to drill out 38" of 'stuff' with a cordless drill and not damage the bore? If there is half a chance it would probably include special made tooling with pilot surfaces and some method of chip extraction. That thin wall SS tube isn't going to crumble into well-behaved chips.

Some epoxy compounds are quite abrasive as well, adding to your problems. A starting idea: a short 3/8" solid carbide 3 or 4 flute endmill on the end of a piloted shaft. If you've got 8" clear to start, the pilot can be fairly long for stability and segmented for reduced friction.

 

[EmGo]

Any tips on drilling it? It’s all going to be done with a cordless drill. It’s about 30’ up and I’ll be working off of scaffolding.

You'll be effed trying that, I think.

I'd make a device that held either a drill bit or maybe a center-cutting end mill, in a sort of tube, loose-ish but not too loose, that was guided by your 1/2" hole. Make like a thick washer on the end that your cutting tool mounts in, held with a setscrew or similar, then that is soldered, glued, shrunk, or otherwise fastened into a 5' long piece of tubing that locates on the larger bore. Then your cutting tool will stay centered, with no effort on the operator's part.

Or even a 1' long section of tube, welded to a 3/8" 4' long rod ... you get the idea. Guide your cutting tool.

The part you'll have to be careful about is, when you get to the end, where it is seated in some type of tapered bore I imagine ? You don't want to butcher that. Cleaning that up so the new piece seats might be your biggest problem.

Or maybe it's just got an o-ring, you didn't say. But if your cutting tool is guided by the bore, drilling out the old part up to where it is fastened, should be a picnic. If you try to do it freehand, good luck.

 

[herthbn]

I know this isn’t going to be an easy task by any means. The alternative is full disassembly of the turbine and that would take 40-45 days with more broken hardware than just this probe.

I like the idea proposed of an end mill or bit with a guide tube. Im thinking that the guide tube would have to slightly extend beyond the cutting edge in an effort to keep the probe centered on the bit and not allow it to walk to one side of the bore or the other. I’ll see what we can come up with tomorrow.

My hope is that with enough vibration and movement from this process, the probe would release itself and fall out before I make it all the way in. It is not mechanically held in place by any means. When I said it looked tapered at the end, I meant that it looks like it has closer sidewall clearances that are likely filled up with debris or rust acting as the binder.

 

[Mark Rand]

Get a gun drill of a diameter to fit the thermowell and use compressed air to blow out the chips. It'll need a carbide tip (most have that as standard) because the thermocouple insulators will be ceramic, not epoxy. The station workshop, or an outside supplier will need to make the air feed fitting and adaptor to whatever drill motor you use.
The gun drill will be self piloting and shouldn't damage the thermowell, but it'll need feeding gently so it doesn't grab on the sheath/insulator/wire nonhomogeneous mixture.

Spent 7 years putting thermocouples into thermowells as part of the job when doing power station turbine performance tests, but luckily didn't get one stuck

 

[jim rozen]

Brass tube brazed on the end of a steel tube. Teeth filed on the end of the brass. Rotary coupling on the top so the it can be spun, and abrasive slurry pumped down under pressure while downforce applied. Renew the teeth periodically.

 

[Tyrone Shoelaces]

And I thought I had a some shit jobs ! Best of luck with that.

Regards Tyrone

 

[memphisjed]

Condolences. And a small hammer with a drift pin. If that doesn’t work a bigger hammer drill and drift pin with an abrasive slurry so you do not damage anything. If that doesn’t work a bigger hammer.

 

[Peter from Holland]

Like always in cases like this The hydraulic way
Get a tube that fits tight in your "well" and over your thermocouple
Fill the "well" with some grease Insert the tube into the hole and over the thermocouple
And then smack it with a hammer as hard as need be
It might need some refilling but if it is all full you notice you can get a solid blow
Trapped air makes it springy But just keek hammering
Ad some seals if need be or use clay or thick sealant
Let us know if it works if you go this route


Peter

 

[Bill D]

Like always in cases like this The hydraulic way
Get a tube that fits tight in your "well" and over your thermocouple
Fill the "well" with some grease Insert the tube into the hole and over the thermocouple
And then smack it with a hammer as hard as need be
It might need some refilling but if it is all full you notice you can get a solid blow
Trapped air makes it springy But just keek hammering
Ad some seals if need be or use clay or thick sealant
Let us know if it works if you go this route.


Peter

Peter has the best idea. I think he stole it from me. What PSI is this well sitting inside? If it is much the hydralic method should not do any harm to the well. If it does not work it will help lube any drilling.
Use a piece of welding rod to measure the depth to check if it is moving. You need to push it out far enough it moves out 8+ inches so you can grab it.
They make needle tips for grease guns. Solder on a 48" long piece of small copper tube to the needle. Use that to get the grease in deep.


Bill D

 

[Peter from Holland]

Peter has the best idea. I think he stole it from me.

Bill D

No I stole it from my neighbour Not all good ideas are yours

Peter

 

[Garwood]

1/2" holesaw or annular cutter brazed or welded to end of 7/16" steel tube.

 

[guythatbrews]

Man, that sounds like a problem! The biggest issue is the unknown tapered bottom of the well.

Call a mobile spark erosion service. They may be able to do it.

Spark Erosion - AP Autodiagnostics Ltd

AP Autodiagnostics is a company using the latest and best equiptment to allow us to provide the very best in service for your vehicle, specialising in the diagnostic sector of the motor trade giving us the benifit of concentrating our efforts in becoming the best diagnostic service provider in...

www.apautodiagnostics.com

If you cut it out chip evacuation will be a big problem for conventional cutting. You may have to cut just a bit and retract the tool often to remove chips. If the tube is not supported it will flop around and damage the cutter. The cutting torque might twist the tube making your situation worse.

The hydraulic method is likely the best DIY choice since it has the best chance of dislodging the end of the TC which is almost certainly what is stuck. You probably can't remove the end any other way. And as has been said you can still try the drill later.

 

[guythatbrews]

Not sure of your location. The spark erosion link I posted is in the UK. I know there are services in the states because we were approached about running a field location a long while ago.

 

[SteveEx30]

Spark erosion definitely not good idea since the thermocouple is epoxy filled.

And personally I would stay the hell away from Carbide drill/cutters if you break them then you really are in trouble.

Sounds like slow drilling and lots of air blowing with a deep gun might be only way.

 

[guythatbrews]

Spark erosion definitely not good idea since the thermocouple is epoxy filled.

And personally I would stay the hell away from Carbide drill/cutters if you break them then you really are in trouble.

Sounds like slow drilling and lots of air blowing with a deep gun might be only way.

Might still be worth a call to the spark erosion guys. As somebody said likely ceramic not epoxy due to high temperature. Maybe they can erode the tube a ways and break the ceramic and start over.

One thing for certain it is a nasty business.

 

[Hightemp]

Anyway you could take advantage of resistive heating? Maybe the probe is in electrical contact (with some corrosion and other crap) with the well. Hook up a welder and see if you can get that contact point to get hot enough to loosen things up?

Or, it melts a hole in the well and, well, sorry about that...

 

[herthbn]

That’s an awesome idea Peter. I like it. Sounds like it might work the best and would be the most accessible method yet. It actually gave me another idea. I’m hoping that this thing is only partially locked up inside the well. If I can find a tube that slides over the thermocouple and inside the well, then they tube may be able to act as a sort of sleeve where the thermocouple could slide down inside and out of the well. I didn’t realize it but the insertion point is at an angle at the start so it’s not a straight shot at drilling the first 8” (see photo). That means this whole probe could be at a slight bend and when it’s pulled out, it would have a continuous bend in it and not straight like it should be. I’m getting welders tomorrow to cut the crooked tube off in effort to get me a straight shot at a drill or hydraulic method. Wish me luck.

 

[boslab]

Porcelain tile drill, 1/4 tubular silver soldered onto a length of 1/4 bundy or steel hydraulic tube, if a swivel available possibly pump coolant up the tube, the drills are tubular and designed to go in a guide ( the thermocouples hole.
It’s going to be slow but doable
Just an idea
Mark

 

[Bill D]

So much heat. Upside down, no good way to pour in some Liquid N to freeze thaw the rusty bits.
Bill D