Roots type blower with high vibrations

[hirschaplin]

Hello,

Have some problems with a roots blower. Twelve inch gears, 762 mm long bi-lobes. Horizontal gas flow.

Operating with a 80% hydrogen mixture.
Inlet about 0,6 barg / 30 deg C
Outlet about 1 barg / <70 deg C

Machine max. allowable speed as per the OEM is 1800 rpm.

The blower is driven by an electric motor with a VFD for speed control.

Normal vibration level RMS for 1100 rpm is around 3 mm/s. When we increased the rpm to 1350 the vibrations went up to 9,5 mm/s and then shortly suddenly went to 12,5 mm/s.

Currently operating at nitrogen for testing purpose, differential pressure at 400 mbar, 1000 rpm, vib RMS reading is around 4 mm/s.

Historical bearing RTD trends shows that there are sudden temperature spikes at bearings on the gear end. Temperature jumps from 80 deg C to 115 deg C very sudden.

The blower is balanced to g2.5 as per ISO 1940.

Aligned with proper laser tools.

There are NO vibrations on the foundation, or after the blower inlet/outlet expansion joints. No vibration levels means below 2 mm/s.

It seems like the machine is responding super bad to increasing the rpm above 1100 rpm. >1100 up until our max of around 1450 rpm works super bad with vibration levels at min 7.5 mm/s.

Problem is that the mechanical seals are leaking at high vibration levels. Need to maintain around 4 mm/s to have those seals leak free...

Any ideas?

 

[georgeverghese]

Gearbox or flex coupling on this machine ? Any problems here?

 

[hirschaplin]

Flexible 500 mm spacer coupling between motor and blower.

I think we might be on to something. We stopped the test cycle last night to perform hot alignment and once the bolts holding the blowers 4 foots was loosened, the blower was moved sideways by itself by the expansion joint.

Inlet and outlet expansion joint has a length of 430 mm. About 120 mm is a bellow and about 310 mm is solid pipe and flange meaning it is not 430 mm of bellow...

On the inlet side the available space is 410 and on the outlet side the available space is 450. Hence is seems like the expansion joints put unwanted strain on the nozzles when they have been "forced in position"...

Another thing is that these expansion joints are without studs, hence they can move freely which means we may experience the phenomenon "pressure thrust" when the gauge pressure is pushing on every mm2 of the effective area of the expansion joint. Can someone elaborate or talk about pressure thrust?

I think to replace the expansion joints to get better fit and also to add reinforcing studs to restrain and control the amount of travel that we allow the expansion joint to push against the blower having the pipe takibg up most of that load instead.

Obviously I am really hoping this is super influencial but I am also thinking that this may be grasping for something that is not there...

 

[hirschaplin]

Continue to run now with Hydrogen (88% H2 and 12% N2). Expansion joints are tweaked a bit, seems like no difference...

Here are some logs from tonight playing with rpms and pressure settings:

Unfortunately the system doesn't have a proper recycle line, hence it is a bit difficult to set the pressure to any desired value.

Here is a poor video showing the spectrum live on NDE above the driven rotor:

https://drive.google.com/file/d/11N1qTVP3flXKjCAxZ4q5CNRsDOgCHC5X/view?usp=sharing

I can play more with the spectrum tomorrow and on different positions if you have any ideas we can try to determine what is going on.

 

[hirschaplin]

We have continued to measure the spectrum at different positions on and around the blower.

It is clear that all clear peaks are att 88,5 Hz (4 X RPM) which is the pocket passing frequency (rpm*4/60=1327.5*4/60=88,5 Hz). Meaning pressure pulsations.

The readings at 4 X RPM occur most clear on the seal house around the drive shaft.

Upstream and downstream the blower, 4 X RPM is not clearly identifiable which I take as that the silencers/pulsation dampeners are effectively eliminating pulsations created by the blower.

What to do? It seems like the blower is not rigid enough by itself to carry the load from its own pressure pulsations or how to explain it?

 

[hirschaplin]

Can the pulsations create a pressure thrust motion in the unrestrained expansion joints that acts on the blower flanges?

 

[georgeverghese]

"Unfortunately the system doesn't have a proper recycle line, hence it is a bit difficult to set the pressure to any desired value."

There ought to be an auto operated recycle line to maintain desired suction pressure, or discharge pressure.

A roots blower would operate much like a pd compressor. Can we see a process control scheme which you may elaborate on ?

 

[pierreick]

Hi,
Take a look,

https://ramequipment.com/standard-v...ls must,thru 1.0 are considered satisfactory.

Pierre

 

[hirschaplin]

@georgeverghese there is only a recycle line with a FCV to control the flow to downstream operations. But this line is not fully sized intentionally to handle the full flow, i.e. 100% recycle. Seems like they forgot about testing purposes.

@pierreick That reference is a joke and it originates form Dresser Roots vibration guidelines. Similar values can also be found in VDI 3836. But 0,45 in/s is not an acceptable level for any blower. Especially if you expect continuous operations for years between overhaul and ever more so if you have mechanical seals. You will never find a sophisticated mechanical seal that is rated for more than 4 mm/s... above that the seal OEM will just tell you to fix your vibrations before you complain about leakage...

The best vibration guideline you will find for a process blower is API 619 if you can meet those targets you have a machine that performs well.

The blower is back in operation and runs around 3 to 4 mm/s at the rated operating point. Above that however and especially at swift speed changes we are still experiencing problems. All vibrations peaks at 4X rpm when measuring on the blower casing. Upstream/downstream equipment, foundation etc. experience super low vibration levels.

 

[Tuckabag]

Does the sudden temp increase/vibration correspond with the recycle line running @ 100%?
Could it be rotor slip?
I'd imagine that a 762mm L rotor would have a fair bit of flex when slipping and generate some good pulsations.
No signs of rubbing anywhere?