Burst Corrugated Metal Hose 1

[365che]

Hi,

We have a ~30,000 gal tank that we use to hold fatty waste. In order to avoid fat solidification in the tank, we apply steam through an internal heating coil. The steam is regulated via an on/off discrete block valve controlled by the DCS. When the temp reaches the high end of the range the valve is commanded to close, when temp reaches low end of the range the valve opens. Immediately following the steam valve, but prior to entering into the tanks coil inlet, is a PRV.

Note: An RTD mounted on the side of the tank sends temperature readings to the DCS.

Once the heating coil exits the tank, it is met by a corrugated flex line (~6' off the ground) followed by a bucket trap on ground level. The condensate that passes through the steam trap enters into a condensate return header and ultimately heads to the condensate return tank to be fed back through the boiler system.

This design setup is legacy. The corrugated metal flex line has busted more than once. I figure this is the work of water hammer.

I'm looking for general design practices to prevent this incident from reoccurring.

What I've come up with so far:
1) Determine source of water hammer + provide necessary design changes. Any ideas?
-Steam occasionally getting through the trap.
-Sudden temperature changes from transfer of fat into the tank shocking heating coil.
-Sudden temperature changes in the coil from steam valve opening 0 to 100% suddenly rather than gradually.
-Note: This issue has historically occurred in colder months. The tank is located outside. The tank and all associated piping are insulated.

2) Check sizing of PRV
3) Check condition of PRV (it's old and potentially damaged)
4) Is PRV in proper location.


If the overall picture is unclear, I'd be more than happy to draw out a simple diagram. Unfortunately, I was unable to do so at this particular time.

 

[LittleInch]

First define PRV Pressure Relief Valve or Pressure Regulating valve?

Definitely sounds like you're getting condensate forming and lying in the pipes somewhere between the on off valve and the steam trap.

Only you can see this and know if there is a constant slope or not.

Is the steam to the on off valve on a dead leg, so you get condensate forming there as the pipe cools in the colder months?

So many things to consider including heat tracing, installing some sort of modulating steam valve instead of on/off, installing a low flow bleed line to keep some flow going through everything and preventing condensate build up.

diagram or pictures will always get you better answers though.


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

 

[georgeverghese]

Winter temp below 0degC in Maryland, so it may be due to trapped water freezing up in this flex hose. Can you heat trace this section leading up to the trap ? - dont know if you can heat trace corrugated flex hose. If you have problems running this temp scheme in winter, it may be the bucket trap getting blocked up too with ice. So steam or electric heat trace all the way down to the return header ? Assuming the return header stays warm in winter with hot condensate from other heating loops.

 

[365che]

LittleInch-

The PRV is a pressure relief valve.

Yes- The steam line, coming from the steam header, that leads to the on/off valve is a dead leg. The steam header is ~12' off the ground. The line with the on/off valve that enters into the side of the tank is ~8' off the ground. This section is entirely steam.

Likewise- the line exiting the tank drops about ~6' to ground level, then travels ~1 1/2' horizontally prior to meeting the bucket trap. The flex line is at the ~6' height right where the line exits the tank.

georgeverghese-

I somehow missed pointing out that the entire line from the exit of the tank to the steam trap is steam heat traced and insulated.


I have attached a diagram for better communication purposes.

 

[LittleInch]

So the inlet line isn't steam traced?

Looks like you collect condensate against the on/off valve as it cools which then blasts through the coil when you open the valve.

SO looks to me like you need to either install a steam trap right before the valve or connect a small bleed line around the valve and send ~10% of the steam through the line constantly.

You can check your PRV but don't think that's the issue - it's the condensate that's being blasted around the coil that I think is the issue.

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

 

[georgeverghese]

Another hypothesis may be this fatty waste is leaking into the steam coil when the coil is not in operation and gone cold - partial vacuum conditions in the coil when not in operation. How corrosive is this fatty waste, and what is this flex hose made out of ? When was the last time you did a leak test on this coil? Or could the fatty waste be freezing up in the flex hose - does it expand or contract on freezing ?
What is the max operating pressure / temp of supply steam, and confirm this hose can handle this.
So this flex hose is steam traced - steam tracing may not be reliable. Checked?

 

[hacksaw]

What's the size of the corrugated line? Is it heat traced or insulated?

 

[pierreick]

Hi,
-I suppose you are using rated flex hose with proper certificate (P, T).
-Why don't you use hard pipe instead of flex hose?
-It should be easy to insulate the short pipe and provide some tracing (electrical preferably).
My 2 cents
Pierre

 

[365che]

LittleInch- There is maybe a foot between the ON/OFF valve and the inlet to the tank. That small section is not steam traced. I checked the layout and there is actually a steam trap on the dead leg prior to the ON/OFF valve. Not shown in the diagram.

Hacksaw/Pierreick- The corrugated line is 1 1/2". It is both heat traced (steam) and insulated. I checked the tracing, and it is active. The flex is rated for 800 psi. Our steam is typically around 115 PSI (350F). There must have been a large instantaneous pressure spike for this flex to have busted so severely. We use flex instead of hard pipe because the tank is on weigh cells.

georgeverghese- I hadn't thought of this. I'm not sure how often, if at all, the coil is pressure tested. We keep the tank at a fairly narrow temp range, so the "cold" condition of the coil would still be relatively hot.

NOTE: It was relatively short time ago that we automated the valves to open/close based on a temp range. Prior to this, the valves were open 100% at all times & the tank was uninsulated. The condensate exiting the coil was sent to the trench. The changes we made were automating the block valve to ON/OFF based on temp range, insulating the tank, and connecting the condensate, post bucket trap, to an existing condensate return header.

I'm still unsure on next steps in trouble shooting this issue. I know that the flex line hardly ever had issues prior to these changes. It has busted twice now since the changes ~6 months ago.

 

[Latexman]

How corrosive is this fatty waste, and what is this flex hose made out of ?

I am interested in this too. We know the flex hose is kept hot. One SWAG that could explain it, is if the waste has chloride ions and the flex hose is a 300 series SS. After some time chloride stress corrosion cracking can cause catastropic failure. I would send the failed flex hose to a good metallurgist/materials engineer.

Good Luck,
Latexman

 

[LittleInch]

So is this a dead end steam heating line?
Even a foot of line with water in it could do the damage. Also there is only 2 feet difference in height between entry and exist nozzles. Could the internal piping have developed a low spot or just not have a big slope on it?

what pressure does it fall to when the main steam valve is off?
Also what are your open and close temperatures?

So is the heating coil 1 1..2 or 2" size?
Opening speed of the valve?

My best idea is to have some sort of min flow bypass around the on / off valve or change out your on / off valve for a modulating valve.

Or maybe have the valve not fully close?

Slug / particles of water at velocity have a big impact on the pipes and connections and can easily break flexible / corrugated sections. The rating will be a non shock rating.

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

 

[365che]

Latexman- I have confirmed that we do not have any possibility for chlorides in our fatty waste. Sending the hose out for inspection analysis could prove a good idea anyways. The flex line failed at such a catastrophic level that it actually looked like it turned itself inside out.

LittleInch- I suppose it's possible that there is a low spot and inadequate slope but it's not obvious here and would prove difficult to confirm.

We don't have reliable gauges for pres/temp on either the upstream or downstream portions of the system here. Our steam is typically 115 psi (350F) out the boiler. Would be slightly less by the time it reaches the fat tank.

I'm glad you brought up the question of pressure and temperature. This does raise a very interesting possibility to me. The condensate return header picks up condensate from a couple nearby tanks in addition to the fat tank. It then runs fairly far throughout the plant picking up condensate in a couple other busy areas before making its way to the condensate return tank. Where this fat tank is located is at the end of the line of the condensate collection header. There is no pump in the condensate return header system. The pressure that pushes the condensate over this long distance is entirely back pressure from steam traps along the way. If this is unclear, I could draw another simple diagram.

Is it possible that the condensate pressure near the fat tank trap is at extremely high pressure. When steam is closed off to the coil and temperature/pressure drop, the condensate header pressure is high enough to backflow condensate through the inverted bucket steam trap and into the flexline/coil. Then when the tank calls for steam and opens the ON/OFF valve the condensate inside the line is met by steam and flashes causing a massive pressure spike? The valve opens immediately rather than gradually.

Perhaps I should install a pressure gauge on the nearby condensate collection header to see existing pressure conditions. Perhaps a simple fix would be to install a check valve after the steam trap. I was under the impression that a bucket trap should act like a check valve. Maybe this trap is damaged from previous freezing conditions or leaks through at high enough upstream pressure? If this solution does not work, than I could install bleed line around the steam valve or have it always partially open. A more expensive solution would be a control valve that gradually opens and closes.

How high of pressure spikes are realistically possible when slugs of water at high velocity occur in pipeline in situations like these?

Is the above theory unlikely?

 

[Latexman]

Your theory is called transient flow or surge or water hammer. 100's of psi can be developed when a column/pipeline of water runs into a quick closing valve or steam trap.

Good Luck,
Latexman

 

[LittleInch]

365,

I'll happily admit I'm not really familiar with steam systems, but assume your tank is at the start of the condensate return ( i.e. furthest away?)

Liquid o liquid high velocity impacts can generate enormous pressure spikes, many times the normal operating pressure.

It's not usually one time which kills systems but repeated shocks to the system creating fatigue cracks until the final straw breaks the camel.

Looking at some pictures of a bucket trap it looks to me like if the steam in your heating coil etc cools the trap open and can allow condensate back into the system?

I think your system may just need to return somewhat to its original design intent and is perhaps an object lesson in either a poor MOC system or just you've been caught by an unintended consequence.

Modulating valve with a min opening sounds like a better bet.

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

 

[365che]

Thanks for your input. Yes, the fat tank is the furthest tank away - the start of the condensate return.

This certainly was an unintended consequence of the modifications made.

https://www.tlv.com/global/US/steam...outlet piping,installed at the trap's outlet.

"If the trap outlet piping is connected to a common condensate collection line, condensate discharged from equipment in operation may backflow into equipment that is out of service unless there is a check valve installed at the trap's outlet."

This caught my eye in the above check valve article, as to my theory from the previous post.

 

[georgeverghese]

Stainless steel ( SS304 / 316) pipes and other sundry fittings with insulation are prone to corrosion under insulation (CUI) with chloride accumulation from the atmosphere also when surface temp is more than 55degC or so. To counter this, a thermal aluminum sprayed coating (TSA) is applied to the SS surface. Else go for standard duplex SS (22 Cr, 6Mo) flex hose, or change out to hard pipe.
Agreed, a spring loaded check valve at the trap exit may help to prevent condensate hammer.