12 replies to this topic

[u062268]

Dears,

 

 

We are Facing a problem in the Hydrogen Compressor Sealing. Both of the Sealing (Primary Hydrogen &

 

Secondary Nitrogen) are Leaking. Primary is leaking to the atmosphere by 4 Nm3/hr of Hydrogen to Flare

 

while Secondary is leaking by 1.63% of Hydrogen is going with Nitrogen (95.5%) to atmosphere.

 

Normally Hydrogen has to be flared whereas nitrogen only has to be vented to atm.

 

 

What will be the harm or HAZOP study in case if we run the compressor and both of the Seals are leaking?

 

 

Attached Files

•  Capture.JPG   51.66KB   31 downloads

[u062268]

????

[thorium90]

Those are labyrinth seals which are non contact seals. There will always be a certain amount of leakage flow through the seals. In fact, it is common to measure the dp across the seals to ensure that the supply of seal gas is at a sufficient pressure and trip the compressor if the dp is too low.

 

To summarise, both seals will always leak a certain amount.

Edited by thorium90, 19 May 2013 - 03:22 AM.

[ankur2061]

uo62268,

 

One way of ascertaining that the labyrinth seal is functioning properly is to measure the flow in the vent (primary and secondary). If the flow exceeds a certain value it indicates that the seal face could be damaged. A compressor trip function should be incorporated on flow alarm high-high in the vent for both primary and secondary seals. 

 

Regards,

Ankur.

[Art Montemayor]

Thorium is absolutely correct.  There will always be a certain amount of leakage flow through the seals - especially in the case of using a screw compressor on hydrogen service.   This fact should have been known and evaluated long before specifying and purchasing this type of compressor.

 

Every Chemical Engineer should know that the smallest molecules among conventional gases being processed are: hydrogen and helium.  Helium is rare; hydrogen is very common today.  This automatically should trigger your concerns on what to do with the hydrogen that is going to leak through a mechanical seal (a type of seal that is designed to leak) or through a labyrinth seal - that is also designed to leak for the obvious reason that you can't fabricate a successful seal that will have 100% seating surface composed of casing metal-on-rotating shaft metal.  That is not possible on our planet.  You have to sacrifice a % of seal gas to vent (or flare) in order to keep the seal cool, lubricated, and satisfy nature's law on seals around rotating or reciprocating components.

[Atttyub194]

Dear 062268

 

Good day !

 

What I have understood that you intend to run your with both primary and secondary seals leaking, probably up to the coming turnaround or first available opportunity and these seals were not leaking in past or at least you do not observe a leakage of such extent. Please

 correct if I am wrong ?

1.  I would simply say that for Hydrogen service it is neither desirable nor recommended to operate as such  as hydrogen forms explosive mixture with air at very low range and you can not shift the explosivity mixture range by introducing steam or nitrogen.
2. Probably all the securities related to seal gas system are also bye passed for such operation which is again not desirable
3. For detailed response you have to share P&I diagram of your seal system
4. This forum may , help you in analyzing prons and cons but can not give you the decision for which the consequences are not known without detailed investigation

 

Best regards and God Bless You

[u062268]

Thank You for your Reply..

I think I didnt clarify the issue well.

29FI746D: Primary Seal

29FI746B: Secondary Seal

 

 

According to the attached pic, 29FI746D normally is reading 1.4 Nm3/hr. After the T/A, the reading exceeds this value to reach 4 Nm3/hr. The unit was still under Hot circulation where H2 was in a recycle Loop.

Moreover, H2 has to go to Flare not to ATM. Also, 29FI746B was reading 0.5 Nm3/hr one year ago, before going to T/A the value increased to 0.88 Nm3/hr, We thought this is due to Instrumentation failure.

Tow months ago, we had a TurnAround (T/A) where the compressor was not included in the list of overhauling. After starting commissioning and cuttin the H2 to the unit, the compressor was leaking. This was figured out by the flowmeter readings (29FI746D = 4 Nm3/hr & 29FI746B = 0.88 Nm3/hr ). We started measuring the H2 in the outle tip of Venting line to find out there is a slight percentage of H2 ( 1.3%H2, 95% N2, 2.5 O2) is leaking to atm. Besides, the N2 is leaking to flare because the reading of 29FI746B = 0.88 Nm3/hr is higher than normal.

Once we started cutting the feed to the unit (Naphtha), the H2 was consumed and there was a recycle H2 returning back to the compressor combining with fresh H2 (Purity 99.99%). The reading of 29FI746D came down ( 1.5 Nm3/hr). We assumed the molecular of other traces which are coming from Recycle H2 line, are larger than H2 helping to control the leak.

 

My question:

Is this Normal??

Are we having a real Leak?

What is Hazard of this Leaking (Primary & Secondary) ??

Attached Files

•  Capture.JPG   96.46KB   25 downloads
•  Capture2.jpg   126.75KB   20 downloads

[thorium90]

The resolution of your P&ID is horrendous. You might want to reupload...

 

Just as a hunch, I would like to share the following tips.

 

Your primary seal takes the hydrogen from the discharge of the compressor right? That means there will only be flow (and pressure) when it is running right?

 

Your secondary seal gas is nitrogen and will always be flowing even when the compressor is not running right?

 

The leakage of seal gas will can go both ways right?

 

Flowmeters calibrated for hydrogen will read a larger value when nitrogen flows through it (MW difference) even when the flow is lesser since you control the pressure and not the flow.

 

Your primary seal gas goes to a flare but your secondary seal gas exits to a vent which I assume is high enough and away from any walkways?

You could always modify it to go to a flare but the backpressure in the flare might be a problem and cause flow and pressure fluctuations in the seal gas, especially since the flow and pressure is already so small, the effect might have been significant enough to cause the original designers to choose atmospheric vent instead of a flare.

Edited by thorium90, 20 May 2013 - 01:16 AM.

[ankur2061]

A Dry Gas seal has normally three components. The primary seal, secondary seal and the separation seal (carbon barrier rings). 

 

Process Gas is supplied to the primary seal with a vent connecting from the seal assembly to the flare. In modern dry gas seal designs no separate gas injection is needed for the secondary seal and only one gas injection point is required. The small amount of leakage from the primary seal to the secondarys seal is enought to form a gas film barrier to cool and protect the seal faces. Thus in modern designs, the vent is common for both the primary and secondary seals.  

 

In older dry gas seal designs two separate process gas injection points, one for the primary seal and another for the secondary seal used to be provided wiith separate vents for primary and secondary seals.

 

The third part of the dry gas seal is the "separation seal" also sometimes known as tertiary seal. This seal requires a very small amount of inert gas (Nitrogen) or dry Instrument air, depending on the design. The separation seal generally has carbon barrier rings as shown in the seal schematic in the OPs post. The reason for the small amount of Nitrogen or dry Instrument air purge for the separation seal is to form a barrier to prevent any lube oil ingress into the dry gas seal assembly.    

 

Since the quantity of this purge gas to the separation seal is very small and it is predominantly Nitrogen (or dry instrument air depending on the design), it is preferred to vent it to the atmosphere.

 

From the sketch provided by the OP in post #1 it is evident that the system has one process gas supply and vent for the primary / secondary seal and one Nitrogen supply and one vent for the separation seal.

 

The modern dry gas seal assembly is provided such that the separation seal part can be separately dismantled from the seal assembly.

 

Regards,

Ankur.

[u062268]

Thorium90:

 

 

1- Yes it takes from the Discharge to add a little Delta Pressure in order to have a sealing

2- N2 gas is injected before injecting H2

3- Yes it is going in two ways

4- It is not high just 3m meters high

[u062268]

Ankur:

 

You are totally right

 

But you didnt answer me in the Leaking issue??

[ankur2061]

u062268,

 

If you are venting more than the design vent rate then that is definitely not normal. It probably indicates the the gap between the seal stationary and rotating faces has increased. Your seal vendor is the right authority to comment on what are the consequences of increased vent flow rate and whether there is a problem with the seal.

 

Analyzing it from the HSE angle it certainly is totally undesirable and dangerous to vent hydrogen from an open vent considering that the ignition energy for hydrogen-air mixtures is very low and the hydrogen-air mixtures have a wide flammability range. Refer the link below for hydrogen safety:

 

http://en.wikipedia....Hydrogen_safety

 

To summarize, you probably need to stop your compressor as soon as possible and take up maintenance for the dry gas seal in order to resore the leakage rate as prescribed by the dry gas seal and compressor vendor. The best way to perform this exercise would be under the direct supervision and guidance of the dry gas seal vendor specialist.

 

Regards,

Ankur.

[thorium90]

The vent flow rate is very low and if the design was done well, there should be no problems.

There is no problem with venting small amounts of hydrogen to the atmosphere.

Although hydrogen has a wide flammability range, once it goes to the atmosphere, the wind dissipates the very buoyant gas almost immediately to a concentration too low to ignite elsewhere. This is just a very small continuous flow, not a large uncontrolled release.

The ignition energy is low and can thus be easily ignited by lightning, but a simple N2 line can extinguish it.

 

However, some guidelines should be followed for the vent.

 

It should vent to a safe area, that means not facing ladders or walkways or buildings or other equipment and piping or collecting under roofs.

Pipe diameter should not be too large to avoid air ingress.

It should be high enough to avoid thermal radiation in case it ignites. (In case it does ignites, the N2 line can be opened to blow out the flame)

It should be a minimum of 3 to 5m in height depending on pipe size.

It should be at least 1m above nearby equipment and 1.5m above a rooftop

 

The part on hydrogen safety in wikipedia is not very well done and can lead some people to sort of overreact when it comes to its safety aspects.

Safety isnt just about how dangerous something is.

http://en.wikipedia....iki/Risk_Matrix

 

As for the seal leakage flow rate, as Ankur has suggested, the compressor vendor would have better advice and might even recommend an overhaul.

Edited by thorium90, 20 May 2013 - 09:30 AM.