Addressing Furnace Plug Return Bend (Mule-Ear) Leaks

If your furnace plug return bends are smoking, don’t blow your top!? There is a practical method to seal them.?This article describes the purpose of plug return bends (aka “mule-ears”), different types of plug return bends, and practical ways to seal them.

What is a plug type return bend?

Plug return bends in a furnace are typically cast 180-degree bends which have removable plugs in order to access the tube ID’s.?The plugs are typically intended to be removed for mechanical cleaning or hydro blasting.?The plugs form a metal-to-metal seal.?The tube connection to the plug return bend can either be expanded into the return bend or designed with a welded tube-to-return joint.

They are most commonly used in coker furnaces now-a-days (if utilized at all), but have been found in many older refinery and petrochemical furnace applications.

The plugs are sometimes referred to as “mule-ears”.?The origin of the nickname has two plausible origins:

1. Some of the older plug return bends are found to have the foundry’s name “Mueller” on the casting.?Mueller (sounds like “mule-ear”) has a long history of piping and foundry products dating back a century and have previously produced these types of plug return bends. ??https://en.wikipedia.org/wiki/Mueller_Co
2. One common design utilizes two long “ears” axially protruding which are used to jack the push-bolt against as can be seen in Figure 1.?These are compared to the ears of a mule.

?Figure 1 – Cast plug return bend (mule-ear style) with welded tube joint

The following configurations are known to exist:

• Mule-ear style (Figure 1)
• Horseshoe style (Figure 2 and 3)
• Threaded toggle style (Figure 4)
• One-off flanged designs (Figure 5)

?

Figure 2 – Cast plug return bend – horseshoe style with expanded tube joint.?Note the shape of a horseshoe.

?

Figure 3 – Cast plug return bend – horseshoe style with expanded tube joint

?

Figure 4 – Cast plug return bend – threaded toggle style with expanded tube joint

????

?Figure 5– Cast removable return bend – one-off flanged style with welded tube joint.?This one was used in a ~1000psi operating application.

Why would I ever want to rely on a metal-to-metal seal in a furnace application?

You wouldn’t.?The best way to avoid a leak on these plugs is to eliminate them altogether and replace with a more traditional 180degree return bends with a butt-welded joint.

This is old technology.?Back in the day before good furnace pass flow and firing control systems were employed, substantial or complete fouling/coking of furnace tubes was more common than today.?Removable plugs provided a method to mechanically clean or hydro blast coked tubes, especially before mechanical pigging technologies came about.

In at least one furnace rebuild in recent times, the argument was successfully made to keep the original plug return bend design (with rolled tube joint) as it was much quicker to expand the 9Chrome tubes into the plug return bends rather than weld traditional 180’s and PWHT every bend weld. ?

In general end users tend to eliminate the plug return bends during planned furnace retubes.?Plug return bends are believed to be a partial cause of less effective mechanical pigging as the cleaning pig can lose effectiveness as it passes through multiple box-like return bends.?Newer plug designs with radius plugs now exist and SHOULD be utilized whenever planning a purposeful retube intending to keep plug return bends – however, most end users would just prefer to remove the plugs altogether.

On some coker furnaces with small tubes (3” NPS), the small tube is already at a higher risk of complete blockage during spalling/decoking.?If considering to eliminate plug return bends on small tubes in coker furnaces, the end user should evaluate the past history to of plugging to determine if easy access via the plugs are worth the risk of a potential future seal leak.

What are the most practical methods to seal leaks if they develop?

There are at least four options:

Method 1:?Hope the leak goes away by coking itself off

A leakage acceptance criteria for coker heater plug return bends during an off-line pressure test can be one drip per minute (at operating pressure) – experience has shown that such a leak has been able to successfully seal (coke) off during normal operation – application of a steam lance may be required during operation until fully sealed. ?While this may work occasionally for coker heaters processing heavy pitch, it may not work for lighter HC furnaces.?In addition, steam air decoking (if employed) may remove coking which has previously “sealed off” any plug leaks.?

This method is not always effective.

Method 2:?Hone the plugs using a specialty machinist company

This method has been used numerous times and found to be timely and very effective, including on bends in a cat feed hydrotreater (~1150psig operating pressure), coker heaters, and crude heaters.?Generic procedure is as follows:

• Matchmark each plug component to the each return bend, such that each plug gets put back into the exact same hole.
• After disassembling jack-bolt and plugs, perform visual inspection of plugs, jack-bolts, and mating surfaces after pulling, looking for large dings/dents in seating surfaces.
• Clean threads on jack-bolt, clean plugs and mating surface. ??Jack-bolt threads should be able to be turned by hand after cleaned. Do not abrade sealing surfaces.?Use nothing more abrasive than Emory cloth on sealing surfaces.?
• Plug:?chuck up on a lathe and skim cut as little material as possible.?Skim cut the plug sealing area NO MORE than 0.005 inches, maintaining the same taper.?Preference is to skim cut 0.002” initially and then check to see if the surface is smooth.?If needed you may skim cut more, but DO NOT exceed a total of 0.005inches.?Preferred finish is 16μ”AA. Below is a photo of the good finish (completely clean and less than 0.005”, and fine shiny finish). ?

Figure 6– Trued surface of male plug after skim cutting

• Paint Dyechem machinist’s layout paint on plug female seating surface. Do NOT confuse machinist’s layout fluid with Prussian blue used in a later step. The intent is to ensure an even lapping is performed. See attached video to understand the layout fluid: https://www.youtube.com/watch?v=n5laGi3GO7M
• Work with machinist specialty company to produce an aluminum dummy plug to be used to chuck up to a drill, and to accept emery cloth strips. Use machinist specialty company to perform the honing. Holding the drill as steady as possible, hone the female portion of the plug return bend until all Dyechem layout paint is completely removed. If some areas are not being removed, it is an indication that the seating surface is not true, and requires additional honing.

Figure 7– Setup for honing the female part of the plug return bend

• After all skim cutting of plug and honing of female seating surface is complete, specialty machinist company to perform a Prussian blue check of every plug. Do not confuse Prussian blue with machinist’s layout paint used in a previous step. Apply Prussian blue to mating surfaces, torque plugs to full tightening values specified, and then carefully remove plug for observation. Plant inspector to spot witness the blue-checking. Figure 8 shows examples of blue checking, showing a clean sealing band all the way around. In general, it is desirable to have at least ?” wide continuous and clean (absence of blue) band all the way around – which indicates a true metal-to-metal contact. If a continuous band is not present, additional honing is required. Note, the previous skim cutting and honing steps above generally allow for a good Prussian blue check.

Figure 8– Results of the Prussian blue check – a clean 360 band indicates a positive seal.

• Clean any residual Prussian blue from the plugs.?Procure Loctite Heavy Duty Anti-Seize (or equivalent) and graphite powder.?Mix graphite powder with anti-seize compound (25% powder by volume).?It was found that craft sometimes forget the addition of graphite powder, so it is suggested to add an inspection hold point to ensure graphite powder has been mixed with the anti-seize compound.?
• Do NOT apply Grafoil tape to the plugs.?Although this has been used to effectively seal plugs in some cases, experience has shown the Grafoil can relax over time and can lead to significant future unexpected leaks years down the road.
• Apply anti-seize mixture (75% anti-seize, 25% graphite powder by volume) to seating surface.?Liberally apply anti-seize to jack-bolt threads and bottom of jack-bolt.?Reinstall plugs and torque to XXXft-lbs (job specific).
• Leak check per local plant procedures.

Method 3: Lap the sealing surfaces using specialty made cast iron lapping tools and lapping compound.

This method can take much longer to perform and is manually intensive but is reported to be the most effective if time is available to be performed properly.?However, most maintenance organizations do not have the time needed to perform this method properly.?Contact the plug return bend supplier to purchase manual lapping hardware (cast iron) and recommended procedures.?A high-level overview is discussed below, not to replace the plug manufacturer’s detailed lapping procedures:

• Remove any imperfections by lapping.?The steel plugs shall NOT be used for lapping as steel-on-steel does not give the proper finish.?Use the manufacture supplied cast iron tools.??Figure 9 depicts a typical cast iron tool for lapping the female part of the seating surface – similar cast iron tools can be provided to lap the male portion of the plug. The cast iron tool should be used to lap in short strokes and follow manufacturer’s instructions.?The initial lap should be performed with a course grade (80-120 grit) compound to get through any build up, and finish lap performed with a medium fine grade (220-240 grit) lapping compound.

??

Figure 9– Depiction of cast iron lapping tool for lapping plug female surface.

• When done, inspect sealing surfaces, clean away lapping compound thoroughly.
• ?Follow manufacturer’s instructions on re-assembly.

?Method 4: Seal Weld the Plugs:

This option ALWAYS gets asked by plant management when a leak arises.?It comes with concerns which need to be addressed and understood:

• Is the casting (potentially overheated) weldable?
• If a low-chrome return bend, is PWHT required?
• If PWHT is required, is it a good idea to perform a local PWHT on this casting?
• If PWHT is performed, might it relax an expanded tube joint (if the tube attachment is expanded) thus causing an unintended leak? FYI, once the plugs are welded, a tube expansion cannot be performed.
• The plugs will need to be pulled anyway to be cleaned for welding.?At this point, does it just make better sense to remove the plug return bend and install a welded 180 bend?

Despite these concerns, in some cases seal welding has been successful.?Actually, the plugs in Figure 2 are seal welded if you look closely.

Other Methods:

One end user reported sealing by use of induction heating on the return bends for an interference fit, then applying the plugs and tightening, but no further information was available.

What methods have you used to successfully seal your plug return bends?