sludge/varnish in steam turbine oils ?

[Shannow]

we've just done a centrifuge overhaul, and found a layer of dark brown "gunk". Some is solid and chips under a scraper, and some is a little thicker than molasses.

The stuff smells like coffee with chickory.

It's soluble in neither turbine oil or water.

Varnish ?

 

[Shannow]

Update.

The stuff exposed to air needed to be scraped off, while the bitumen in the oil space could be scooped out.

The semi liquid that ran down the side of the cup, and was exposed to air took about 8 hours to set, and formed a hard laquer.

 

[Shannow]

further update...

Plant was out for an unscheduled pit-stop, and I asked the maintenance people to remove a few covers and have a look at the valve gear hydraulics. All looked sweet.

Could the centrifuge be generating the stuff ?

The centrifuge has an external water supply make-up to form the water seal.

 

[bruce381]

high heat and possible moisture in a steam turbine with oil that has been run way to long will COKE and form the varnish you have seen.

Use a good quality GPII oil (chevron has a good one) and have a UOA done every few 1,000 hours to trend the TAN number max of 1.0 is cut off I think for a turbine.

Also a FTIR will be a good way to see when to change it so it will not be at a point of sludge and varnish, also check cooling system for clogs.

Oil centrifuge maybe introducing water from leaky coolant jacket and be running to hot also.

bruce

 

[bruce381]

also if gunk is from oil centrifuge it maybe what it is designed to do to clean turbine oil of solids and as such is Normal BUT still turbine oil should last years but it must be kepr clean and dry.
bruce

 

[Shannow]

bruce, the 35,000 litres in the system is group 1, and undergoes a continuous change of around (well at least) 10% per year (pommie design, and lots of leaks to the collection sumps).

Oil cooler outlet temps are pegged at 45C (113F), and bearing outlet temps are not allowed to exceed 85C (185F) at all.

Being a steam turbine, steam gland leakage is directly into the oil labarynths, meaning that the oil gets pretty wet. Most of it drops out in the tank, or is removed in the centrifuge.

Unit runs 8500 hours per year. Monthly, we test wear metals and moisture post coolers (for leaks). Six monthy RBOTs, which are always good.

My "fleet: has 4 units, two ltos of two identical units, and this one's sister unit shows nothing.

 

[bruce381]

what filtering is on line
were is/was gunk from
what are the TAN numbers on the oil
what is your idea of a good rbot?
vibration anylisis or wear metal show any bearing problems on turbine?

what does spectro of gunk show? metals, dirt,
boiler water treatment chemicals of any kind?
contamination of wrong oil added in past

bruce

 

[Shannow]

quote:

---

Originally posted by bruce381:
what filtering is on line

---

20um in line filters (no bypass), with a 5um on the centrifuge loop (tank back to tank).

quote:

---

Originally posted by bruce381:
were is/was gunk from

---

not sure, as we've only seen it in one unit, and only in the centrifuge.

quote:

---

Originally posted by bruce381:
what are the TAN numbers on the oil
what is your idea of a good rbot?

---

TAN, I don't know.
RBOTs have all been 250mins plus (see continuous oil change above)

quote:

---

Originally posted by bruce381:
vibration anylisis or wear metal show any bearing problems on turbine?

---

We had a big rub 5 years ago, and also had a wipe on a few bearings as a result. We dollied them up at the time as there was no maintenance window to fix them, and the wipes were very small.
No wear metals evident since, and bearing temps are good.
Vibration is above what I'd like, but the orbits look OK (but big), no sign of constraint, and no whirl.
When barred, the alignments look OK.

quote:

---

Originally posted by bruce381:
what does spectro of gunk show? metals, dirt,
boiler water treatment chemicals of any kind?
contamination of wrong oil added in past

bruce

---

gunk results are on the way.
boiler water treatment is purely ammonia. We had an issue with colloidal silica (and therefore steam carry over) a few months back, but both units recieved the water, only one has varnish.
Both units have a common oil supply tank, with bulk delivery. Problem should show in both.

The whole thing comes apart in April for maintenance, so we'll see what we can see.

 

[Shannow]

BTW Bruce, thanks for the questions...I need some answers, and that's the only way I'll get the right test done.

 

[bruce381]

sound like you have covered all the bases.
when I asked where gunk was from ment in filter housing, centifuge itself, lines, covers it sounds like a over heated oil deal in only that unit and the centifuge has done its job of knocking out the solids.

What is particle count?

Perhaps problem is old has been with you since startup?
and you are seeing the result after the fact.
or has this unit been down before and this is new?

love to see what gunk sample shows both additives and wear meatals

any ODD history of this unit since it has a sister and both on a common lube system you may have to find what if any difference in heat, water, flow, filter change out, online time anything that is out of the ordinary

also I would like a >500 hour RBOT might look at a oil upgrade

I have had lube problems in the past like this and even tho both units are "the same in all areas" something is not, you may still find a maintence problem somewhere upon tear down but be thankful unit still on line and running OK.

email me if you like with any further info on gunk or if you have a UOA and any other info you would like to send I will keep all correspondence in confidence.

Bruce
CLS
G-C Lube

 

[Shannow]

Molakule,
that's a thought.

I was down the governor end of the machine the other day and the fumes from the ammonia tank were very storng (like I've not seen before).

Was also wondering whethere the generator hydrogen seals might have something going on (no indication of rubs or leaks, but the oil does come into intomate contact with the hydrogen there).

 

[va3ux]

I was an operator in a steam plant many years ago. We had two 25,000 HP steam turbines and four 10,000 HP turbines.

These machines were overhauled about every 4 years. I remember talk around the dept about a varnish problem being discovered in units 4 & 5 one year.

So it can happen.

Phil

 

[Shannow]

Phil,
I've heard of it happening before in my state about 20 years ago, and leading to serious control valve issues, so I'm worried.

BTW, these are 675,000 hp, and the others 940,000

 

[bruce381]

Did you get the gunk lab work done and please sepcify exactly where did the sludge come from?

Is system oil open to air?? if not ammonia fumes will do no harm even if ammonia is a problem both systems on a common lube line should suffer.

But only 1 unit has sludge are you sure have you looked in the sister system?

If only 1 unit what is the difference? in operating time and load?

from what you told I vote for contaminated or over heated oil due to clogged heat exschanger.


bruce

 

[MolaKule]

Just a theory, but could ammonia intrusion somewhere be upping the N levels and creating this sludge or varnish?

It sounds as if a high nitrogen level combined with moisture is creating your sludge.

 

[Shannow]

bruce,
the samples haven't even been looked at at present

...I can't seem to get the condition monitoring people to realise that it's possibly important.

If overheated, it must be a localised bearing issue (not unlikely), but the oil leaving each bearing passes over a thermocouple, with alarm at 85C, and trip at 90. Maybe overheating local to the centrifuge, but the water seal should stop it getting much past 95C.

While not open to air, the bearings have a simple labyrinth seal. The oil returning down the returns acts as a boundary layer pump, and drags air through the labyrinths, and pressurises the tank...stuff might get in that way.

Coolers are full flow, with an alarm at 45C, at which point the flow is switched to another.

Operating hours for both units are about 160,000 hours. Load profiles are similar.

I agree, it's going to be something simple...if I could just get the people charged with condition monitoring to get off their freckles, and do their job.

 

[Shannow]

was getting some tyres the other day, and there was a diahatsu pulled down for bearing problems...the sludge blocking the pick-up screen smelled just like my turbine sludge.

 

[bruce381]

Your sludge is oxidized oil from?? what conditions may be known only upon unit rebuild.

Loacalized over heating

water or other contaminates

filter residue or bearing material may also be a precursor to form sludge and coke

Coke also will form from oxidized oil upon splash on to a HOT metal surface.

Copper or brass components in the lube system will promote sludge

All above will show sludg and varnish just have to find out what is different from sister unit and/or what is cause

Upgrade to a GPII oil will also help over time will raise RBOT to >500 hours I would consult oil supplier to see if a like additized and compatible GII or better oil is avalible and look at a change out or at least us the better oil as makeup oil from now on.

Let us know any more thoughts or info on this
a full oil report would be my starting place with full metal spectro and TAN and Ferrography aswell as a particle count.
bruce

 

[bruce381]

One other thought can you test the oil for water?
Karl Fisher method and compare both units after the heat exchanger or cooler? you may have a leak in the unit with sludge that is not noticable and may only show in a oil sample on line compared the the none sludged sister unit.
Basicly you will have to commpare both units oil sampled from same places and see what one is different.
bruce

 

[Shannow]

bruce, KF is done regularly, and has been pretty normal.

the sludge unit had 54ppm after the cooler, and the other 41.

As to your other comments, the relay gear is all brass sleeved. the shaft metal temperatures are in the low 100sC at worst after shut down.

Still waiting for the sludge analysis