Steam noise mitigation 5

[itsmoked]

Hi, I'm working on a large extruder.

It's a 150HP twin screw job with about twenty modules. Each module is electrically heated with 2 to 4kW of electrical heater. The drive is also adding 150HP of shearing heat.

Each module has a PID running its temperature control. The PID runs the module's electrical heat AND a bang-bang water valve for cooling. It's all mind-bendingly dynamic as the modules get heated up to get things going then the extrusion process starts and now that dumps lots of heat so the electrical heat throttles way back. Meanwhile the water valves cycle to keep the module temps from going too high.

The issue at hand is that the module temps hover around 320°F. The cooling water is 80 to 110°F. When the extruder gets going
the noise coming out of all the modules is horrendous! It's something relating to the cooling water probably turning to steam then back to liquid.

The entire operation would be almost dead silent except for the screeching, popping, and hissing that occurs whenever a water valve is open and most are frequently open! You have to shout to be heard anywhere near the machine.

What is this exactly?
Any suggestions on how to mitigate this?

Top pipe is the cold manifold and the bottom is the return. Could those be backwards?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[georgeverghese]

At 320degF plus, return cooling water pressure would need to be at about 100psig min to keep it from boiling. May also help if those on - off valves for temp control were damped a bit during closing to avoid hammer ?

 

[itsmoked]

Thanks george; We were thinking 'raise the pressure' but are faced with the cooling tank (shiny rectangular metal tank seen at the far left of both pictures) is open to atmosphere. That's extruder 'factory' too. Could be the Chinese aren't too worried about workplace noise...?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Latexman]

Or change to a less volatile component, like a heat transfer fluid. I doubt it would be just a 'drop in'; other changes would be needed. An exchanger, a pump, and more controls.

Good Luck,
Latexman

 

[danschwind]

Keith,

Maybe adding a backpressure valve and an upstream air cooler (radiator) to avoid flashing the hot water through the valve would solve it?

Daniel
Rio de Janeiro - Brazil

 

[MintJulep]

Suddenly squirting "cold" water on a a hot surface is likely bad for a whole lot of other reasons beyond the noise.

Rather than on-off water a continuously modulated flow (that never goes to zero) would help.

An old-fashioned steam trap at the discharge would likely improve things as well, if there isn't one there already.

 

[Compositepro]

Where large masses of steel, like platten presses, must be cooled from high temperatures, water mist cooling is often used to avoid shock cooling, which can crack steel. In your case where you are not really goingg though thermocycles but rather start-up vs continuous operation, oil heating is commonly used. Your system is just the cheapest possible.

 

[georgeverghese]

There may be steam-condensate hammer also (which may explain all that noise) with the current setup as the trapped steam suddenly collapses as cold water rushes in, so agreed, changing the controls from on-off to continuous control would be safer. That (closed loop?) expansion tank looks small; could be replaced with a 10barg design pressure drum with an inert gas blanket?
This may be a plant availability issue - fatigue with thousands of thermal cycles may mean an early demise for these extruder modules.

 

[davefitz]

when cold water hits a metallic surface that is more than about 70 F hotter than the boiling temperature at that pressure then it is like spitting on a hot frying pan- a layer of steam film forms between the water jet and the steam on the metal. This is called DNB departure from nucleate boiling , so the noise is likely the steam film bubbles collapsing when the steam moves into the much colder water . In german it means the metal temperature has exceeded the "liedenfrost temperature" for that pressure.

"...when logic, and proportion, have fallen, sloppy dead..." Grace Slick

 

[EdStainless]

Our HT furnace has water cooled sections.
The flow control was set up so that there was always some minimum flow through a side branch.
The main valve then adjusted based on the outlet water temp.
These are purely thermo-mechanical devices with no electronics at all.
Very smooth and predictable.
Google them.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[itsmoked]

Latexman; What other "less volatile heat transfer fluids are there? This might work because there is a tube and shell heat exchanger with an outside cooling loop. Water could be used to cool a different "heat transfer fluid".

danschwind; You're thinking is that the water is flashing to steam back down 50cm of tubing?
I'm afraid if the pump is only delivering 50psi then the back-pressure valve would block the 100psi needed at 320F to prevent the steam in the first place?

Mint; You really think any less flow (smaller continuous) would limit the noise? Seems cold water hitting the hot metal in small or larger amounts all will equal a bunch of noise.

georgeverghese; Cheapest possible... I'm on board with you! LOL Oil. That would probably solve the noise problem but I fear the hazards that come with using it to cool something that has the potential to get way way hotter during some operational screw-up. I can see using carefully heated oil to heat your process because you know it will not get over heated somehow. But cooling the process what if one of those heaters (which are solid-state relays) sticks on? Or the circ pump dies mid job, Problem? Not a problem?

davefitz; Thanks much for that DNB explanation. It makes sense!

Ed; "The flow control was set up so that there was always some minimum flow through a side branch." So a continuous shunt?
And, you say "google them". What are "Them" called?




Keith Cress
kcress -

http://www.flaminsystems.com

 

[georgeverghese]

With the increased recirc pressure of 100psi or so, we could still operate at a constant 100degF exit water temp or so I presume. 150HP heat input = 112kW, so if water goes from 80degF to 100degF, we would be looking at approx 9m3/hr (40gpm) pumping rate. Presume 150HP is total extruder shearing heat input for the 20modules?
There probably is some temp cutout switch for cases when controls fail or if the CW pump drops out.
Some suitable hot oil may work just as well, if heat transfer or pumping constraints dont get in the way. Another concern would be compatibility with the material being extruded - what if this hot oil broke through a pin hole into this material ( or is min extrusion pressure > 100psi) ? Dont know if there is some food grade hot oil thats good for up to 350degF or so. On the other hand, this recirc CW may be some DI or low chloride RO water doped with oxygen scavenger or similar to minimise corrosion in the cooling circuit and extruder heat exchange surfaces?

 

[Latexman]

There are dozens of heat transfer fluids out there. Look at Dynalyne, Therminol, and Dowtherm. There's more. Pick one with a freezing point and viscosity low enough you have no problems, and a boiling point greater than your hottest temperature, 320[sup]o[/sup] F.

My first thought was ethylene glycol (FP 9[sup]o[/sup] F/BP 387[sup]o[/sup] F), but I'd get a HTF company involved in the decision.

But realize, they will be an inferior HTF than water on a lb-to-lb basis. But they won't flash when they hit the 320[sup]o[/sup] F in your extruder and collapse (bang!)/condense shortly thereafter. They will be inherently quieter.

Is your product used for/touches foodstuff? If so, maybe propylene glycol.

Good Luck,
Latexman

 

[EdStainless]

Sorry, like these

https://www.control-specialties.com...lves/temperature-control-valves.html?limit=24

Put the sense bulb in a 'tee' is the drain side.
I have used the Hoffman and RobertShaw valves.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[itsmoked]

Thank you gents for this fodder to chew on!


Let me present what I was thinking about at lunch.

Would it be possible that it's designed this way and actually expects enough coolant (H[sub]2[/sub]O) flow that the water can't actually reach 212°F?

I'm told they, "Closed all the manual valves then opened them 2 turns".

But they're needle valves! That may be a minuscule amount of water?



Keith Cress
kcress -

http://www.flaminsystems.com

 

[georgeverghese]

The relevant parameter to compute with continuous flow is the film ( or more conservatively surface temp) on the hot end of the CW side of the cooling surface of this extruder - some idea of surface geometry, CW velocity, hi and ho are required to compute this parameter. Say we keep surface temp at less than 190degF max. Physical props and flowing parameters of the extruded material will also be required for this estimate.

 

[EdStainless]

You did say that the inlets are on the bottom and drains out the top?
You need to do that in order to keep the system full, unless you have wide open inlets and are controlling flow at the outlets.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[MintJulep]

Mint; You really think any less flow (smaller continuous) would limit the noise?

I didn't say smaller. I said continuous.

Enough flow to keep the surface continually wet. It can boil, or not, just don't let it go dry.

 

[Compositepro]

https://www.chromalox.com/en/catalo...ystem/hot-oil-heaters---heat-transfer-systems

https://www.aecinternet.com/product-category/oil/

 

[itsmoked]

Ah Mint; That's just down this new road of maybe trying to keep the water cooled surface operating below 212F. Do that with the help of continuous flow. I get it now. Nice.

Ed; I believe all I said was the "feed manifold pipe is the top one" and the return is the bottom one. I think they both enter at the bottom of the cooling block. Which is probably eminently lame with regard to purging the air out of the space. Sheesh.

Comp; Thanks for those links. I'd not have expected those two to do cooling too.

Keith Cress
kcress -

http://www.flaminsystems.com