Thermocouple Pressure Drop in Pipe

[SuperG]

How do you calculate the pressure drop in a pipe due to a thermocouple in the middle of the fluid stream?

I have a 4" pipe with water flowing through it and a thermocouple will be permanently inserted into the stream. I see thermocouples can be inserted vertically or at an angle along the direction of flow.

 

[KernOily]

As a first cut I think I would ignore it and see where my overall system dP was. I think if you contact the vendors that make point contact flow switches, like FCI, they may have dP data that you could use as an approximation.

If I was doing this I would probably just neglect it. Thanks!
Pete
P. J. (Pete) Chandler, PE
Principal Engineer
Mechanical, Piping, Thermal, Hydraulics
Processes Unlimited International, Inc.
Bakersfield, California USA
pjchandl@prou.com

 

[SuperG]

Originally I did ignore it. But I just saw a diagram of it. The thermocouple appears to be 1.5-2" thick, taking up nearly half the width of the ID.

Anyway, I have never heard of anyone doing this calculation, but I have left word with one of the vendors. I'll let you know what they have...

 

[TD2K]

I still doubt the dP will be significant. An orifice plate beta ratio is typically anywhere from about 0.5 to 0.75. At 0.75, the area of the orifice is about 1/2 that of the pipe and unless you have relatively high velocities for your fluid, the dP across an orifice plate is usually about 100 to 200 inches of water with close to 1/2 of that being recovered downstream. You don't say what the length of the well is versus the ID, can you do a quick calculation assuming 2, 1/2 segment openings on each side? (I'd calculate an area for one side and then assume it's an orifice and split the total flow between each. Rough, yes).

Granted, this is not an orifice plate but I still doubt the dP will be that high. Interesting question, I've seen vibration calculations routinely done for thermowells but never a pressure drop. Sometimes I've seen them swedge up the line to insert the thermowell or alternatively, put it in a elbow facing into the flow.

 

[KernOily]

I've also seen swages or reducers at thermowells but in my experience that was to accomodate a TI/TE of the correct length on an otherwise too-small pipe. One time I had to swage up to a 3" from a 3/4" SS tubing line in a caustic dilution skid. We were trying to monitor the heat of reaction when the caustic was diluted and the only way to properly get a TI in that small tubing was to use a reducer.

Man that is a BIG TW. I've never seen one that large - a typical process plant standard TW is, like, 1/2" OD, no? Thanks!
Pete
P. J. (Pete) Chandler, PE
Principal Engineer
Mechanical, Piping, Thermal, Hydraulics
Processes Unlimited International, Inc.
Bakersfield, California USA
pjchandl@prou.com

 

[TD2K]

Corrosion allowance Pete

A BIG corrosion allowance.

 

[lewbbhp]

The following is a formula that was given to me that I have used for years.

K EQUIVALENT - THERMOWELL

K = [A (sub P) * A (sub TW)] / {[A (sub P) - A (sub TW)]^2}

A (sub P) = OPEN AREA OF PIPE (SQ. IN.)
L (subTWe) = ROJECTION LENGTH OF THERMOWELL END INTO PIPE (IN.)
W (subTWe) = PROJECTION WIDTH OF THERMOWELL END INTO PIPE (IN.)
L (subTWs) = PROJECTION LENGTH OF THERMOWELL SHANK INTO PIPE (IN.)
W (subTWs) = PROJECTION WIDTH OF THERMOWELL SHANK INTO PIPE (IN.)
A (subTW) = PROJECTION AREA OF THERMOWELL INTO PIPE (SQ. IN.)

The projection will vary depending on whether a straight -olet or latrolet is used.

 

[TD2K]

Where are subTWs and subTWe supposed to be in the formula?

 

[lewbbhp]

(sub xxx) is the abbreviation I use for the particlar subscript rather than trying to change the font which wasn't always successful.

 

[lewbbhp]

L x W = A
The area is calculated for the end and shank separately as they were different for the thermocouples we used. The two areas combined give the total area used for the calculation.

 

[vpl]

Hey Lewbbhp - How much of a pressure drop do you normally see? Like TD2K, I've normally ignored thermowells in doing pressure drop calcs. I'm wondering if I shouldn't. Patricia Lougheed

 

[SuperG]

Excuse my ignorance, LEWBBHP...is this equation the calculation of a new K, then use that in a pressure loss calculation?

Also, what do you mean by "THERMOWELL?" Is this the K for the Thermowell?

 

[KernOily]

Another great discussion... I think VPL's question is most pertinent here. I'm certainly not the most experienced guy out there, but I've yet to be up against a situation where I needed to account for the dP across a TW. Even in a pump suction with a critical NPSHa situation, I don't think I've ever accounted for the drop across a TW. Educate me please!

In my pursuit of accuracy, I constantly have to remind myself that the various friction loss equations out there (Darcy, BBM, etc., etc., etc.) are only close approximations to the 'real' friction loss. You will never be able to match your dP calcs to field-measured dPs. The best I've gotten is within about 8%, but that is plenty close enough, even on pump suctions, because you're always going to add 2' to the manufacturer's published NPSHr before you put your design to bed, right?

I guess my point is - I don't see that the loss across a TW is worth determining, given the big picture. Enter engineering judgment...

SuperG, just curious, why do you need to account for the dP across this device? Thanks!
Pete
pjchandl@prou.com

 

[SuperG]

I want to calculate it at the request of the client...it's on the back burner right now, but I still want to see what the loss will be...

 

[hacksaw]

SuperG,

The pressure drop across the thermowell is not the only problem.

Sounds like the customer is having problems with that measurement, hence the overly heavy design being used.

1. Calulate the pressure drop (only an estimate) of a sudden restriction imposed by the thermowell. Crane or Camron's hydraulic handbook will work.

2. Determine why such a heavy walled thermowell is being used.

3. None of the thermowell manufacturers can offer much help in such applications. Seems like there are related articles on thermowells in Hydrocarbon processing (nov. 2001)and in the oil & gas journal(april 2002).

What is your flow rate? I'll run a quick calc using some FEA thermowell software we use for identifying the flow induced stresses.

Measurement errors (response time)will be a problem if the well is as thick as you've stated.

 

[hacksaw]

Ran some stress calcs on a thermowell with U-dim of 4" (more or less) and a root diameter of 1.5"

Flow induced vibrations are not a problem in condensate, as long as you are keeping the fluid velocity with in reason.

However, wells of this size are typically intended for super-critical service.

You don't want the thermowell to block more than 10-15% of the flow area, if it does then you must resize the pipe to to keep the superficial velocity within the ratings established for that class of piping.

good luck

 

[inControl]

Dear all,

Perhaps these websites can help you a bit further:

http://www.pressure-drop-calculator.com/

(freeware)

or, even better:

http://www.pipeflow.co.uk/

(30-day trial)

For the thermowell I would say: use a smaller one or use a temperature sensor which is submerged in the pipe-wall. I know Yokogawa has them. Then you avoid any pressure drop problems!

Good luck & Happy New Year!