Insulation for Flanges of Heat Exchangers 3

[etsen]

The medium through tube side is MP steam and shell side benzene. Channel barrel and shell barrel are insulated but the flanges beween them aren't. I think the flanges should be insulated for heat conservation and even personnel protection. But the equipment engineer said it would be not convenient for maintenance and not easy for finding leak and even increase the external corrosion if they were insulated. It seems a little reasonable and I am a little confused now. Could anyone give some suggestions for me? Thanks a lot.

 

[chief]

Yes , normally the flange connections would be insulated to conserve radiation heat and of course to protect personnel from contact with a hot surface.

Offshore Engineering&Design

 

[DG72]

We are a fabricator of heat exchangers, so we see a little bit of everything (everyone is different). Some insulate the flanges and some don't, but what we see most often is uninsulated body flanges.

 

[moltenmetal]

If flanges need to be inspected frequently, the thing to use is a removable insulation jacket. Glass fibre cloth impregnated with silicone, with a suitable batt-type insulation material underneath. Held together with drawstrings and/or hook and loop (Velcro) fasteners on the exterior (ie. where it's cold enough to not melt the Velcro). Insulation companies can have these made for you.

 

[etsen]

TO DG72
Why are most of them uninsulated? Could you provide appropriate reasons for that from the aspect of design? Thanks.

 

[TGS4]

The reason to keep flanges uninsulated is that doing so keeps the bolts cooler than the process fluid. The differential in thermal expansion caused by keeping the bolts cooler maintains a higher bolt tension.

Another reason that I'm aware of for keeping flanges uninsulated is that if the process fluid contains hydrogen, it will diffuse through the gasket. If there is an insulating cover, hydrogen can build up in the enclosure, leading to the potential of an accident when the cover is removed.

 

[DG72]

We don't insulate the exchangers--this is done after we fabricate them, so I am not involved in the decision process. The reasons for not insulating (that I am aware of) were the ones mentioned by TGS4.

 

[itdepends]

Chain of events

1) Process engineering specifies insulation on flanges for heat conservation
2) Heat exchanger is installed and in service with insulated flanges
3) Maintenance is performed that requires breaking of the said insulated flanges- If rigid insulation- it's normally damaged beyond repair at this stage.
4) Maintenance is complete and flanges bolted back up without insulation
5) Some time passes..............
6) Energy efficiency review identifies heat losses - potential to improve plant efficiency by insulating all those nasty hot flanges
7) Insulation is applied to flanges
8) Go to step 3)

 

[moltenmetal]

Wow TGS4: hydrogen diffuses straight through a gasket but is then magically retained by a flange's insulation cladding sufficiently to become a hazard?

itdepends: what are you claiming is damaged beyond repair: the insulation, or the exchanger?!

 

[TGS4]

moltenmetal - it sure does, because it's a function of the pressure. For example, inside the pipe, you have a hydrogen partial pressure of, say, 2,000psi. That diffuses through the gasket into the area contained by the flange insulation. However, there the pressure is essentially atmospheric. Therefore, there is no driving force to move the hydrogen out (short of some sort of osmotic action).

I've collected atmospheric pressure hydrogen (high school experiment with electrolysis), and even a few milliliters will create an unpleasant explosion.

Most of the refineries and upgraders that I've worked at or worked for have specific requirements for NOT insulating flanges in services with high hydrogen partial pressures. In the course of asking about these requirements, I have inevitably found that this has come about because of either a near miss or an injury incident involving hydrogen accumulation in an insulation "box".

 

[moltenmetal]

TGS4: you're correct that you no longer have a pressure driving force, but you've got one hell of a density driving force for the hydrogen to leave, along with the concentration (diffusion) driving force.

Your insulators are to be commended if their cladding is hydrogen-tight against diffusion losses THROUGH an intact gasket!

I can see the cladding concentrating gross leakage from flanges, but it would need to be quite a large leak, and I trust your normal hydrogen handling procedures would prepare workers to detect such a leak. This sounds like another argument for the sewn removable jackets I suggested to the OP- they're definitely more vapour-permeable than aluminum or stainless cladding. They're also far easier to remove, and far more likely to be correctly re-installed, than clad insulation covers.

 

[TGS4]

moltenmetal - as far as sewn removable covers go, I think we are in agreement. There would be enough leakage to avoid such a problem.

However, given the benefits of a cooler bolt, I would still tend to not want to insulate flanges that operate warm/hot. The additional gasket contact stress caused by the cooler bolts can often negate any gasket creep that may occur.

 

[zerosum]

When we run really hot service, we seldom insulate for the bolting integrity reason addressed by TGS4.

 

[SnTMan]

Ever seen any rain shields?

Regards,

Mike

 

[moltenmetal]

Dunno how many kW you're willing to waste for the benefit of cooler bolting...

We build small plants, and if we didn't insulate the flanges you couldn't keep things hot enough to operate properly. Then again, the consequences of a leak are less dire.

 

[itdepends]

moltenmetal- the insulation gets damaged (although I have seen heat exchanged flanges that required re-work because the guys cut the seized bolts off with an oxy-torch.......and cut into the flange at the same time).

Cheers.

 

[AND98]

Has anyone run into a problem with insulating boiler ends with extra blanket insulation? I installed it, the paint burned, and people are pointing fingers. It was a large heat loss - i believe either the paint was wrong or the refractory was previously damaged (inspection of the refractory was fine).

Has anyone dealt with this before?