Sabroe SMC100 broken valve plates

[RANGER1]

We lok after a site with multiple Sabroe SMC108L MK2+3 compressors .
All on NH3 and 1440 rpm
-13 SST + 35 SC water cooled

mobil SHC226 oil


We have to service them every year without failure obviously due to speed .

Suction valve plates usually have big chunks broken out destroying cylinder liners .

Any suggestions about why or is it just metal fatigue ?

PS can't slow them down .

[AUcooler]

Cant compress liquid. I'd look into liquid flood back.

[Magoo]

Heads overheating, super clean head cooling circuits and more water through heads, check more often than yearly, as water galleries really calcify. I would rip the water heads of and send to a acid cleaning guy , the same people that clean engine blocks.

[Grizzly]

Yes I would go along with Magoo, although I suspect someone of your calibre has already thought of that!
I have noticed a trend in recent years.
For the Head Cooling to be supplied via a circulating pump, off of the evaporative condenser sump.
This has it's own problems where the phosphates within the water treatment.
Come out of suspension and silt up the heads.
When they are opened up they are full of this grey sludge. Which when left to dry goes white!
Good luck yours is certainly an unusual problem, not a normal occurrence for smc's.
Particularly as you service them every year!
Can't wait to hear if you resolve this one Ranger!
Grizzly

[mad fridgie]

I agree with both magoo and grizzly, but for a slightly different approach how about a closed loop (stop any further fouling) then using a side stream from NH3 or water HX (depending upon available temps).
It may seem to reduce efficiency (true on the short term) but over the year may improve due no/limited damage

[RANGER1]

Thanks for replies but I don't think its a water cooling problem or flood back .

In my opinion their running under severe duty .

1/ speed
2/ can run unloaded which Sabroes don't like at -13/+35 .
Sabroe programme would suggest only unload to 50% or discharge temp will exceed maximum ( 150 deg C ).

Its running on probably all maximum recommended limits so they have a reduced life than say 960 rpm machine .
Valve plates are very thin a lot of flex , so I think metal fatique .

Would anyone have any similar duty machines with better results .

This is accepted at this plant as its done the same for 20 years .

I have an open mind about it and are using genuine parts only .

Thanks RANGER

[AUcooler]

I once had a 12cyl Sabroe that suffered some valve problems, which turned out to be caused by a crack in an oil perculator. The liquid line within the perculator had a crack in it allowing refrigerant to accumulate in the perculator on the off cycle resulting in liquid hammer on startup. R22 system though so perhaps unrelated. It was hard to nail down though, other than I noticed an oil return line sweating a little on the off cycle.
Are oil failure faults with this plant common? or unusual?

[RANGER1]

Never had a real problem to my knowledge . Tested annually set at 3.5 bar .

[Grizzly]

Thanks for replies but I don't think its a water cooling problem or flood back .

In my opinion their running under severe duty .

1/ speed
2/ can run unloaded which Sabroes don't like at -13/+35 .
Sabroe programme would suggest only unload to 50% or discharge temp will exceed maximum ( 150 deg C ).

Its running on probably all maximum recommended limits so they have a reduced life than say 960 rpm machine .
Valve plates are very thin a lot of flex , so I think metal fatique .

Would anyone have any similar duty machines with better results .

This is accepted at this plant as its done the same for 20 years .

I have an open mind about it and are using genuine parts only .

Thanks RANGER

I can't fault your explanation Ranger, nor have I an answer.
Due to the drastically increased load would site allow you to service the machines more often?

Are these MK1 or Mk2 machines?
Cheers Grizzly

[RANGER1]

Grizzly they are MK 2&3 and overhauled every 8-9000 hrs and costs a fortune .
Would prefer a screw to take more load and maybe keep a few recips slowed down to 960rpm ( direct drive ) .

[Grizzly]

Grizzly they are MK 2&3 and overhauled every 8-9000 hrs and costs a fortune .
Would prefer a screw to take more load and maybe keep a few recips slowed down to 960rpm ( direct drive ) .

Are they on Unisab Control ( just wondered if the control range could be smoothed out a bit?)
Is this application petrochemical by any chance?

Surely your above option makes sense?
Or is this one of those where the reactive budget is flexible.
When the capital project one is not?
Grizzly

[mad fridgie]

You are running at 1400RPM and then indicate unloading to 50%, why not speed drive, sorts out both problems

[RANGER1]

System has its own PLC and has to unload at varios times , as its a food production plant ,
Production lines start / stop at anytime .
As usual money is a problem when you want to modify plant .

[Grizzly]

System has its own PLC and has to unload at varios times , as its a food production plant ,
Production lines start / stop at anytime .
As usual money is a problem when you want to modify plant .

Sympathies I have several sites like that myself!
Grizzly

[josef]

Service life of discharge and suction
valves
In order to ensure that the compressor always
works perfectly, it is advisable - at suitable intervals
- to replace the suction and discharge valve
ring plates.
It is difficult to give altogether precise times for
such replacements, as the durability of the valve
ring plates depends on the following factors:
• If the compressor is exposed to liquid stroke
or moist refrigerant gas, the service life is reduced.
• Speed of the compressor:
At 900 rpm, the service life of the valve ring
plates is considerably longer than at 1500
rpm.
• The compressor ratio at which the compressor
operates:
At high compression ratios, the load on
valve ring plates and springs is considerably
larger than at low compression ratios. When
the valve ring plates are changed, the valve
springs should also be replaced.

[RANGER1]

Thanks everyone , valves and springs are changed each year .

It seems its the duty plant is under .

I have seen 1 SMC106L at 960rpm with 45000 hrs and not touched .
( not that I would expect any industrial recip to achieve this on a regular basis ) .

In Europe / UK is it a common thing to run a sabroe at this speed ?

[josef]

I quoted from the manual SMC. Yes, widely used in Europe RPM 970 and the 1460th Service plate low stage 10 000 hours, high stage 5 000 hours. I work on the compressor SMC from 1970 + Atlas-ice machines 12 hours a day, 5 days a week and 22 years old plate work well. I think the iron is not very good.

[Grizzly]

Yep I agree with Josef.
Lots of very reliable Sabroe recips about.
Similar to Josef. I have several 20yr plus Smc's on Finsam Ice plants.
Plus one big mother (SMC 8 180 MK1-1974) 35yrs old!
Which I am in the process of rebuilding at the moment.
With a lot of help from my friend and colleague Paul. We have just removed the Crankshaft for a re-grind (220kgs approx).
Anyway as we agree yours is a peculiar one Ranger!
Cheers Grizzly

[josef]

Grizzly, I have the impression that they are very brittle (hard)

[mbc]

Dear ranger
1- SMC 100 has two different type of valve (suction and discharge) for NH3 and CFC So Have a good look.
2- You should cool oil (( top and side cooling OR injection liquid of NH3 to crankcase )) as much as you can.
3- You should use Heater during off time.
4- Watch to return back liquid to compressor
5- Do not tight too mach screw (hexin) of discharge valve plate

[mad fridgie]

One option is to get your components cryogenically treated (some times known as cryogenic tempering). Plates, valves, springs and liners. This process gives greater metal stability.
Never tried on refrig compressors, but unbelievable results on just about all car parts (engine and transmission) There are couple of places in Aus that undertake this process. Pricing is not to bad either (pay by weight)

[sevilou]

How are the retaining plate ? I saw some problems when the plate comes in contact with a defective retaining plate.

[RANGER1]

It's a common problem on all the comps on this site .

Have another plant with mycom WA recips and does 2 yrs no problem .

[Andy]

Thanks everyone , valves and springs are changed each year .

It seems its the duty plant is under .

I have seen 1 SMC106L at 960rpm with 45000 hrs and not touched .
( not that I would expect any industrial recip to achieve this on a regular basis ) .

In Europe / UK is it a common thing to run a sabroe at this speed ?

Hello Ranger

speed is the problem or rather the combination of speed and the long stoke, just too much gas going through the valves.
Direct drive machines? Change the motors to 960 rpm.

Another possibility would be changing the discharge valve assemblies to the R22 or if they are low stage machines the HP holders, this would help with the increased volume, but only on the discharge valves.

Kind Regards Andy

[Magoo]

OK, back to basics, what is the actual suction vapour temperature entering the compressor?. And what is the actual discharge gas temperature leaving the compressor?.
magoo

[RANGER1]

Liquid recirc plant with 1-3 s/heat on suction .
Dischardge temp varies due to unloading 120 deg C fully loaded up to 150 deg C when unloaded 50% ( SMC 108L) .
Above also changes to lower temps in winter , as these are summer readings .
Oil temps 45-55 deg C MOBIL SHC226

Andy's comment appears to be relevant in this case .

Thanks

[Andy]

Liquid recirc plant with 1-3 s/heat on suction .
Dischardge temp varies due to unloading 120 deg C fully loaded up to 150 deg C when unloaded 50% ( SMC 108L) .
Above also changes to lower temps in winter , as these are summer readings .
Oil temps 45-55 deg C MOBIL SHC226

Andy's comment appears to be relevant in this case .

Thanks

Hello Ranger

seems they are running a bit warm

I have seen water chiller and rink plants using chilled water to cool the heads and the oil. If you use water about 12 deg c it makes the valves last a lot longer.

Try a packaged water chiller on one compressor (or a plate heat ex with HT pumped liquid), this will cure your problem.

Alternatively Sabroe have a thermo pump system, expensive to retrofit, but cheap to run.

Kind Regards Andy

[RANGER1]

Andy ,
Thanks for ideas as sometimes you accept how things are . Sabroe programme says its OK but obviously pushing it to limit .
Would a Danfoss TEAT liquid injection valve be OK for this situation , if piped directly into suction gas stream of each compressor suction .
Or is this a bit primitive possibly causing other problems .

[Magoo]

Ranger1.
the compressors are revving their nuts off, and particually for an L series,.
The sabroe selection will state real fine conditions for application, discharge and suction conditions, the discharge gas temps are way too high for SHC226 synthtetic, part load /unload will only agrivate problem.
Definitly go for a suction saturation devise to get some sort of saturation and reduce discharge gas temps.
Look at staging compressors rather than unloading. May not be practical with the installation

[Andy]

Andy ,
Thanks for ideas as sometimes you accept how things are . Sabroe programme says its OK but obviously pushing it to limit .
Would a Danfoss TEAT liquid injection valve be OK for this situation , if piped directly into suction gas stream of each compressor suction .
Or is this a bit primitive possibly causing other problems .

Hello Ranger

Liquid injection into the cylinders would just expand, reducing the pumped volume.
Sabroe have a thing called thermo-pump, which uses flooded liquid to cool what was the water jacket on the heads, returning the gas to the discharge side of the compressor (I think). If you can get a drawing of a thermo pump system it might be possible to copy it.

I think chilled water to the heads is the best option, it does work very well.

Kind Regards Andy

[2400martin]

I know this sond stupide. But we have the expirience in following area. Is it possible that your oilseparators expirience a airfolw during operation?.
Heat them up and isolate them. your problen is over. the material is not the probel. your lubrication fil is driven away. like blistering by shaft seals.

[RANGER1]

2400martin ,
don't know quite what you mean and think lubrication is ok .
Its same as race car engine does'nt last as long as normal road car .
More RPM , limits are pushed and things break .
If I had my way several screws would replace most of them .
The ones left would run at 960rpm .

[Magoo]

Hi Ranger 1.
any more news on the high rev machines. The discharge temps would worry me, are the discharge lines after compressor big enough for flow rate.

[KFD111]

Do the compressors stop/start many times pr day????

do they use a lot of oil?

Have you tested the disc. valves / bypass valve??? for the bypass of disc. gas

kasper

[RANGER1]

Hi Ranger 1.
any more news on the high rev machines. The discharge temps would worry me, are the discharge lines after compressor big enough for flow rate.

Thanks Magoo , no news as I think its excepted by client and everyone else it seems .
There are multiple comps all the same.

Was looking through some Sabroe litarature today that a friend gave me .
It was from a Coca Cola plant in Denmark which they indicate changing valve plates every 5000 hrs as a standard practice .
I guess that would solve problem and Sabroe and us get richer .

On Sabroe programme all temps add up with conditions .

I have worked on a similar plant with -10/+35 NH3 and thermopump and same problem but worse in other ways . Has no pressure fed gudgeons so pistons pull in half .

[Frank Day]

Hi Ranger 1,

We look after an ammomia plant with 3 x Sabroe SMC 112 L compressors and like you we have to change out valve gear on an annual basis as a preventative measure.

Looking at your operating conditions of -13 SST and 35 SC on NH3, one comment I would make is that most makes of industrial compressors do not like going down below -10 deg C suction , simply because the compression ratio increases and so does the discharge temperature. Any particular reason why you have to control the suction at -13 deg ?

It is clear that the compressors are operating right on their upper limits and anything that can be done to limit elevating the discharge temperature should be done.

Would it be possible to:
1 Up the suction to -10 deg C?
2 Lower the head pressure.
3 Instal VSD's keeping compressors fully loaded rather than running on part load for long periods.
4 Drop temperature of head and sump cooling water.
5 Fit oil coolers.
6 Carry out a top end inspection every 6 months as a preventative measure.

Trust this will help

Rgs

Frank Day

[RANGER1]

Hi Ranger 1,

We look after an ammomia plant with 3 x Sabroe SMC 112 L compressors and like you we have to change out valve gear on an annual basis as a preventative measure.

Looking at your operating conditions of -13 SST and 35 SC on NH3, one comment I would make is that most makes of industrial compressors do not like going down below -10 deg C suction , simply because the compression ratio increases and so does the discharge temperature. Any particular reason why you have to control the suction at -13 deg ?

It is clear that the compressors are operating right on their upper limits and anything that can be done to limit elevating the discharge temperature should be done.

Would it be possible to:
1 Up the suction to -10 deg C?
2 Lower the head pressure.
3 Instal VSD's keeping compressors fully loaded rather than running on part load for long periods.
4 Drop temperature of head and sump cooling water.
5 Fit oil coolers.
6 Carry out a top end inspection every 6 months as a preventative measure.

Trust this will help

Rgs

Frank Day

Frank ,
Thanks for suggestions

1. System is common with chillers and 2 ice machines , that is why system runs at -13 C . Ice machines can be a big part of total load .

2. Head pressure runs as low as possible , but summer would be 35 C SCT .

3. VSD may be possible but expensive on 7 machines or part there of .

4. Cooling water could only be lowered by refrigerating it and would only give marginal results .

5. Comps have standard oil cooling with water jackets on sump covers

6. Client thinks he spends to much money now with annual service .

In a nutshell I do not think client wants to spend anymore money than he has to .

[Magoo]

Ranger 1
are saying they do not have water cooled heads, OK dumb question.

[RANGER1]

Frank remarked on fitting oil cooller , maybe he's suggesting an extra oil cooler ?
Yes Magoo compressor has standard water cooled heads and sidecovers , all clean with good water flow .

[Andy]

Hi Ranger

have you tried chilled water. The results I've seen are not marginal.

I would also recommend changing the valves every 5000 hours
In my experience this does the trick.

Kind Regards Andy D

[Magoo]

Hi Ranger 1
years ago had similar problem with a Hall 3 5/8 x 2 5/8 vee block, again revving it nuts off and top end re-builds annually. I figured I was not getting full story with replacement parts and the rate they deteriorated. So took to super smart eng., company and had new and old analysed. Stating all and sundry and application.
This old scrogger did a Brunell test on new and old ( Brunell, the ship and bridge building bloke , way back ). He said they were not up to the application, and we ordered a new set, made by him. Expensive but did last for, I think four years with annual checks and flipping over. Cheap by comparisson.
For the life of me, I cannot remember his company name. But they made aircraft landing gear parts for the Australian Air Force manufacturers. Those short take off twin engine transport things. Smart guy and he knew metal charactorisics and applications. Probably dead by now.
magoo

[RANGER1]

Magoo , Mad Fridgie suggested sort of similar idea with cryogenic treatment .
There were a lot of old engineers with exceptional all round knowledge of metalurgy etc but like you say probably most are deceased .

Sabroe must be on another planet suggesting change valve plates every 5000 hrs for > 1200 rpm .
Mycom K valve plates same rpm last for 20000 hrs and none broken or damaged .
Sabroe valve plates are very pliable which could be a dis-advantage
as they would flex continually .
They must be shot peened as they have a distinct appearance .
All other machines that I know of have ridgid valves .I could do some R&D and send the bill to JCI .

[Magoo]

Hi RANGER 1.
Can only suggest that you scratch around and find a metalurgy lab and do a comparison of a sabroe and mycom valve plate, it all comes down to heat treatment and hardening specs, I beleive.
The Brunell check is basicaly dropping a diamond weighted thingo on metal and measuring the indentation, which relates to a hardness factor, thats where I get lost.
Hope this helps

[RANGER1]

Magoo I think its proper terminology is Rockwell hardness .

Also Sabroe probably make their comps and parts in places like Mexico or somewhere so
quality may have dropped also .

After saying that they must be due for another oil filter modification kit by now .

[Andy]

There are a lot of Indian manufacturers making valve plates for Sabroes. Some of them may be worth a try.

I'm not sure where the Sabore parts are made, but most parts are made in Mexico (or Brazil not sure which) and shipped over in crates.

Kind Regards Andy