9 replies to this topic

[Anshul Jain]

Hi,

 

The problem is really serious. I have faced a weird problem. We are using a 2 stage vacuum pump for distillation purpose. The system was running fine from past 6 months last week all of a sudden I heard a sound of explosion. When I rush to see what it was exactly, I found that the gear side of the roots pump in vacuum system has exploded. I reported this case to manufacturer, but they do not have any clue about it and also I have been told that there are four-five more cases have been reported to them. They are one of the leading manufacturer and are in the business from more than 20 years. When I had discussed this case with than I have been told that this related to static charge but I was not convinced.

 

Another problem is that the manufacturer supply roots pump in two applications one is for positive pressure and one is in vacuum application and cases reported are for vacuum application.

 

The gear oil used is shell XP90. I dont know what precaution should i take to rectify this problem.

 

Regards

Anshul

[curious_cat]

What's the fluid being distilled?

 

What Temperature is the Vacuum Pump inlet at?

 

When was the oil last changed?

 

Was there any chance the exhaust was temporarily restricted somehow?

 

PS. How big was the pump? At what vacuum level?

Edited by curious_cat, 19 September 2013 - 12:08 PM.

[fseipel]

I take it this is a twin lobe roots blower, and the gearbox is separated from the gas/vapor end with only lip seals on the two shafts?

Typically the gearbox side has a breather; was it open to atmosphere or plugged/restricted?.  Exactly what is meant by 'exploded', was the gearbox housing deformed or ruptured?

 

I have seen static destroy glass-lined vessels; however, in this case, any charge should have grounded through the casing/shaft.  If static is truly suspected or needs eliminated, check the resistance to ground with a digital voltmeter between gearbox and a known good ground.  It should be less than ~20 Ohms if well grounded.  I do not see how gear oil 'explodes'; unless it has been contaminated with solvent even if gearbox wasn't grounded, gear oil does flash, but only at extreme temperatures, typ. 400-600 F; I don't see it getting that hot even in a mechanical malfunction (frictional heating).  What is the flash point of the gear oil?  Was the gearbox hot after the incident? Does the gearbox have metal shavings in it or are the bearings shot or does shaft freely rotate?  I take it this was a forepump to attain a higher vacuum level and that it backed another pump?  If you are running vacuum, I would expect leakage to be INTO the process FROM the gearbox if the lip seals fail, correct?  In that case, how would gear oil become contaminated with solvent? How much oil is/was in the gearbox?

Edited by fseipel, 19 September 2013 - 05:02 PM.

[Anshul Jain]

Thanks for your intrest to the post. (fseipel, curious_cat)

 

Curious_cat

 

What's the fluid being distilled? It’s not a particular type of fluid but generally it is IPA, MDC,Methanol, Toluene and sometime even water.

What Temperature is the Vacuum Pump inlet at? somewhere between 50 to 70 deg C

When was the oil last changed? The machined i was talking about it was commissioned 2 months back and since then it was the same oil. I mistaken the 6months that was the other machine and it is running fine.

Was there any chance the exhaust was temporarily restricted somehow? No way as i was getting a good vacuum level in the range of 5-10mbar.

PS. How big was the pump? At what vacuum level? this is somewhere 1500-1600 m3/hr capcity pump

 

 

 

Fseipel

 

Yes it is a twin lobe roots blower and yes the gear box is separated by lip seals. In this pump there was no breather. I want to share the pictures of that part but don’t know how to share on this forum.

 

The only way gear oil can be contaminated is leakages thru shaft, and that is only possible when lip seals fail. Assuming that lip seals fail then the gear chamber should come under vacuum.

What I assume for an explosion to take place it needs three basic things 1) Fuel 2) oxygen 3) spark. If my gear chamber is under vacuum then the 2 requirement i.e oxygen is not fulfilled.

 

The gear oil grade was shall XP90 which has a flash point of 180-200 deg C and during operation I have measured the temperature of the outside surface of gear case between 80-90 deg C.

I am not sure about whether the gear cover was hot after the incident as we were in a shocking condition and nobody bothered about to go near the machine.

There were no metal shavings and to the beat of my knowledge shaft was running free after the incident. The gear oil qty is around 3 liters.

 

I am not sure or I am still in doubt about the condition that solvent mixed in gear oil has caused this. I was of the same view earlier that it could be the case until I got in touch with manufacturer. The manufacturer got a case reported where the client was using vacuum pump for removing water vapors, so this case made me confused.

 

Resistance between ground and the machine is 100 ohms (best) and 600 ohms (worst).  

[fseipel]

I would suggest putting a surface thermocouple on any remaining Roots Blowers, on the gearbox, tieing into SCADA/DCS for alarm/trending/automatic shutoff if possible.  Getting a history of what is happening with temperature on the gearbox, would seem to be extremely relevant.  If this isn't possible due to cost, and area classifications don't forbid it, consider purchasing a battery powered or AC powered standalone handheld datalogger that records max. temperature and beeps/alarms and leave it attached.  Typically under $150 with simple type J or K T/C (Cole Parmer, Omega, etc).  I don't think you need to get fancy with a thermowell; simply clamping a T/C junction to casting or pressed metal gearbox, with some insulation between clamp and T/C should suffice.  From a preventative standpoint, I suppose you could insert a Nitrogen blanketing regulator, set at VERY low pressure, say, 1 inch water column, and insert it into the threaded gearbox fill port -- that eliminates the Oxygen in the fire triangle.  Alternately buy a Dwyer or similar purge rotameter and inject 1-2 SCFH Nitrogen into gearcase every hour with breather open... Not conventional practice but may be a wise precaution... Does Roots recommend venting gearbox? I seem to recall they do on older models at least.  I'd send Roots the gear oil and ask they assess the dielectric constant just to help rule out static.  Static seems very unlikely to me as I doubt gearbox ever reached flashpoint temperature.  I would measure flashpoint (Roots or outside lab) IF you don't have a flashpoint test apparatus available.  If you can't vent gearbox for some reason, can you put a low pressure relief valve and pressure gauge on the vent tap so internal pressure can be monitored and limited? A standard U-RAI Roots Blower from Dresser has a breather fitted. Thermocouple could also go into a tee fitted up to breather tap on gearcase housing.

It seems to me a more likely scenario, especially given the additional information in second e-mail ('it happened to somebody else not compressing solvent vapors'), also leads me to ask you to consider a potentially more likely scenario -- the oil was contaminated with water, the gearcase was not vented properly/conventionally, and steam formed when gearbox reached some temperature above 100 C -- since it normally runs 80-90 C, an excursion, might take you above 100 C.  This can most certainly cause an explosion in hot oil systems in the CPI, e.g. water pocket exists in a vessel jacket, I inject hot oil at 200 deg C into jacket, steam and hot oil erupt out of hot oil expansion tank raining down on personnel...  It may not be a chemical 'explosion', but it can still cause an explosion nonetheless -- though I'd expect pressure to leak out through lip seals unless gearcase temperature rose rapidly.  What pressure can the gearbox withstand or does Roots have that data?  After the 'explosion' was there soot or any signs of combustion indicating a chemical reaction, or no such signs (potentially more indicative of steam)?  Is there rust or a waterline on gearbox casing at bottom of gearbox where water would settle by gravity?  When you drain oil from other Roots blowers do they have water layer at bottom? What is viscosity of gear oil?  Any signs of coking or exposure to high temperature for prolonged periods?

I'd still check flashpoint of residual oil and also moisture content of same.  You could also use an oil with an even higher flashpoint.  I question whether gearcase could ever even attain 180 - 200 C.  I'm assuming gearbox is uninsulated, as it gets that hot it's going to dissipate a lot of heat to surroundings.

Assuming the lip seals are in good condition and not leaking, then what is happening on the gas end is irrelevant.  I suppose it's possible some liquid solvent leaked through lip seals while blower wasn't running (especially if you 'flooded' it with liquid or ingested liquid), but then only leaked BACK to process side slowly upon pulling vacuum; this seems unlikely.  Is elastomer used for lip seals compatible with the solvent being compressed?  Does process ever have upsets where it goes under positive pressure?

If the blower is on a frequency drive monitoring, trending, historizing the amperage draw may be useful.  I concur per my first e-mail that leakage of process fluid into gearbox should not happen if lip seals are good, and even if/when they fail, during operation at least, you should draw gear oil INTO process NOT process solvents INTO gearcase.  In fact, that's how I've detected leaks in the past with roots blowers -- you see oil in your air or gas stream.  I strongly recommend reporting to manufacturer your findings/conclusions.

100-600 Ohms resistance to ground is not ideal. For gas end, resistance to ground ought to tbe < 20 Ohms.  If it isn't, running a copper wire to ground is recommended. Given that you have an unexplained explosion, I'd run a copper wire from gearbox to ground also just to be on safe side.  On process side (with solvents) you may also want to look into purging seal area with Nitrogen but again this does not relate to the gearbox explosion in my opinion.  Any chance bearings overheated?  Seems unlikely if shaft rotates freely but wanted to throw it out... How long between when blower was last turned on, and explosion?

Edited by fseipel, 20 September 2013 - 05:20 PM.

[curious_cat]

Really, I think a photo might help. Right now this is too much speculation. 

 

There's a link you can post (see menu above the box you are typing in). Post a photo online and share a link if you can. 

[DHRUV MALHOTRA]

Hello,

 

I am a colleague with Anshul who posted this problem, 

 

Following our the details available with us: 

 

1) The process is Distillation in a pharmaceutical company, the process solvent vapour is MDC, Toluene. 

2) We feel the case is of sudden explosion in controlled environment.

3) We feel the case might be of contamination of oil with process solvents, even thought they being seperated by lip seals which seperate the body from the gear chamber there might have been a possibility of diffusion of process vapour into oil over prolonged period. 

4) We have gone thru notes published by EDWARDS on net where they have cited a possibility of such an incident happening in ROOTS BLOWER used for vacuum duty application. 

I am attaching pictures of the damaged gear cover for ready reference and better understanding. 

a. https://www.dropbox.com/s/4zf0t75c26neimw/DSC_1106.jpg 

b. https://www.dropbox....0519_100316.jpg

c. https://www.dropbox..../13 x 18 cm.pdf

d. http://mmrc.caltech....Pump Safety.pdf     (Chapter 6 Page 29)

 

BR// Dhruv 

Edited by DHRUV MALHOTRA, 30 May 2014 - 12:29 PM.

[DHRUV MALHOTRA]

Am attaching some additional information i came across while exploring the internet. 

 

6.2 Elimination of stagnant volumes

A stagnant volume is any closed volume in a vacuum pipe or component which is not subjected to a through flow of 

gas. Examples are the gear box of a mechanical booster pump, or the gauge head of an instrument. Valved pipework 

and nitrogen gas inlet pipes can also become stagnant volumes when they are isolated

Stagnant volumes must be taken into account when you consider the mixture and reaction of process gases which are 

not normally present together in the process chamber. Pipes, pumps and process chambers generally transport gases 

linearly, with one gas or gas mixture followed by another. Gases transported in such linear flows are not normally 

mixed unless the velocity of the exhaust gas is reduced by a restriction or blockage. A stagnant volume is not purged 

and may be filled with process gases as the pressure in the system rises and falls. In this way, gases which pass through 

the system at one stage of the process can be retained. These may then react with gases from a subsequent phase 

of the process. Thorough evacuation of the chamber between the introduction of incompatible gases will guard 

against the risk of explosion.

You must take special care when you consider cross-contamination in stagnant volumes when the pressure is 

relatively high (close to atmospheric pressure) and the gases are potentially explosive. In particular, you should 

consider the hazard of build-up in filters and separators. Where appropriate, use high integrity, continuous flows of 

inert purge gas to reduce the probability of cross contamination.

When pumping flammables, it is possible for stagnant volumes to fill with potentially explosive gases or vapours that 

cannot be removed by normal purging. Where an ignition source could also be present, specific purging of the 

stagnant volume should be considered.

 

Your comments shall be appreciated, can we pin down our case to this explanation? 

 

BR// Dhruv

[Art Montemayor]

The forum still doesn't know the extent or type of the "explosion" that took place.  We don't have a photo of the vacuum blower - nor the identity of its manufacturer.  That this is a serious incident is certainly of concern to us all.

 

The Roots-Connersville company in Indiana, USA invented and manufactured the famous twin-lobed blower for many years.  I have bought and operated many of their models in the past.  Roots is no longer in existence and other companies - some in Europe - manufacture copies of this type of blower and vacuum device.  Your blower is probably not a true Roots-Connersville blower.  There are basically TWO types of models for this type of blower: one has the timing gear box connected directly to the compression chamber and separated by a "lip" seal.  This was the most commonly used model because it was meant for transporting low pressure air.  For special applications where the pressures were elevated or where the contact of the pumped gas was detrimental to the gear box - or a possible hazard - the gear box was segregated from the compression chamber and had packed glands on the lobe shafts.  I used this type extensively in the past on high temperature, higher pressures - such as regenerating adsorption beds.  I also used it as a booster to compressors.

 

From what you describe, I would not recommend using a lobe blower to develop a vacuum - and I would certainly not use the type of blower you have employed.   To me, this is clearly a case of bad engineering specifications in the process design.  This is where process experience and total process knowledge is required in selecting the appropriate equipment and type.  Simply asking a blower vendor to supply a blower for vacuum service is not an engineering function.  A complete data sheet, complete with process description and requirements should be developed and experienced process engineers should review and accept the selected equipment for the specific application.

[DHRUV MALHOTRA]

Dear Sir,

 

We value your suggestions and wish to draw your focus on the issue to keep the technical discussion alive. 

 

Twin lobe roots type blowers are extensively used world all over for vacuum pumping, working in the range of 0.001-1 mbar where they have large displacement with very low power consumption. They offer an ideal non contact, non sliding pump and are able to sustain a large differential pressure across its suction and discharge. In pressure duty applications these blowers rarely work beyond a compression ration of 2:1 while in vacuum duty the compression ratio can go as high as 50:1. The work done that is PV in vacuum is very very small even for relatively a large displacement. Hundreds of machines have been used and our running trouble free but in some rare cases explosion in gear chamber have been witnessed. Explosion is like a sudden blast wherein the gear cover shatters into pieces and explodes outward. The intensity is so strong that the gear cover made out of cast iron rips apart. There is never a fire even though the oil spills all over. This indicates that there is a limited combustion resulting in explosion like in a fire cracker or bomb. The construction of a vacuum booster and air blower is similar except for leak tightness and for the fact that in vacuum boosters the pumping gas/ vapour is low boiler, highly volatile and flammable. In a air blower this explosion is never being reported and has been observed only in vacuum boosters. This has led us to believe firmly that there is a minor leakage/ diffusion of vapours into the gear chamber which form a potential vapourised fuel and in the presence of limited air has a tendency to explode if a spark is made available. The spark may be due to hot spots on the gear sliding surface for which we are unable to give any explanation. In a rare event when a volatile mixture, limited oxygen supply and spark are all available, the probability of explosion would be high. The gear cover has various weak points like oil sight glass, oil fill and drain plugs, which under normal condition of pressure rise would fail and release pressure, but due to sudden release of large energy instantaneously the cover explodes before any other weak link breaks. In some design of booster where the gear chamber runs under vacuum the probability of explosion is less due to fact that there is no oxygen available.  

 

We propose to fill up the gear chamber with Nitrogen maintaining a pressure of 0.5-1.0 bar (G) so as to insure that there is no vapour diffusion from the gas chamber and also no oxygen to support combustion. This is achieved by connecting a compressed nitrogen line into the gear chamber thru a pressure regulating valve.

 

I request all for their view and comments and shall appreciate all inputs.

 

 

Best Regards,

Dhruv  

P.S:

1) In my previous link i have attached photos of some damaged component, can add some more photos of the pump as a whole in case required. 

2) These machines and manufactured and consumed in INDIA.