PDA

View Full Version : Liquid injection vs thermosiphon oil cooling



Josip
26-11-2006, 12:43 PM
Hi, all :)


One thing you might consider is talking to FES. They provide us with a lot of help with other people's machines as well as their own. They recently helped us pick the pieces for a recent conversion of a thermosyphon to liquid injection on a a Stahl.

Your opinion about liquid injection vs thermosiphon oil cooling system...

For me this is not a good technical solution (on big industrial screws), but maybe I am not right:confused:

Best regards, Josip :)

Andy
26-11-2006, 02:07 PM
Hi Josip:)
liquid injection is a very poor way of cooling oil:(
We had a screw pack that used this method and was reverse cycle defrost system, kept tripping. So instead of thermosyphon (condenser same level but 60m (200ft) away) I designed an aircooled oil cooler:D piped the oil up to a finned aircooler above the screw pack plantroom then down to a thermo regulating valve and finally back to the same original oil system. We kept the inject and set the settings up for back up. Result one very happy customer, no tripping and a good 10% reduction in energy usage AND a plant which pulled the chambers down much quicker, on cheap nighttime electric, which is in his case cheaper than running the plant during the day:)

I would consider the following
1st Direct oil cooling by air
2nd Thermosyphon where possible
3rd Direct oil cooling by water
4th Liquid Injection (only as a last resort0

Kind Regards Andy:)

Takeshi
26-11-2006, 03:24 PM
On a 4 compressor pack I work on the oil is air cooled, it passes through a almost condensor-like radiator. Is the oil pumped around by the compressor's itself?
And for some reason one of the comp's keeps going out on oil pressure switch..any ideas why this may be?

Andy
26-11-2006, 03:39 PM
On a 4 compressor pack I work on the oil is air cooled, it passes through a almost condensor-like radiator. Is the oil pumped around by the compressor's itself?
And for some reason one of the comp's keeps going out on oil pressure switch..any ideas why this may be?

Hi Takeshi:)

seen this before, can be due to refrigerant in the oil (in the oil separator)
Oil could be too cold (should be 10 k warmer than condensing) again refrigerant in the oil
Also could be high pressure drops in the oil system causing any refrigerant in the oil to flash out and foam the oil.
Could have the oil cooler located in the wrong position, or sitting in a cold place and starting up full of cold oil (again refrigerant in oil)

Against manufacturers advise I prefer the oil cooler above the oil separator so that the oil can drain back to the heated oil separator during standstill

Also the system may not have enough oil volume in it for start up.

Hope this helps:)

Kind Regards Andy:)

NH3LVR
26-11-2006, 05:33 PM
Andy;
Sounds like you did some good engineering on that screw pack .
A couple of thoughts on Thermsyphon.
The conversion that Josip quoted on was a additional cooling system. We were having oil problems and bottling of the condenser. The equalizing line and purger are not installed properly. Further work is planned. We are using the Thermosyphon now, but it might not be effective in the summer. Now that we have got things working properly the customer has to decide whether or not to make the recommend changes before next Summer.
We installed these two used Stals and found we could only run one at a time on the Thermosyphon. If we had designed the system from scratch we would have been responsible for the problems. Troubles just happen sometimes in old plants.
I do agree that Thermo is the best way to go from a efficiency standpoint.
More expensive than liquid injection to install however. To retrofit into an existing plant can be very difficult, particularly when you have multiple engine rooms in the basement of a three story building. One plant we are working on would require tens of thousands of dollars of work to bring Thermo to the new screw we are going to install. Sometimes it just is not worth it to the customer to make the capital outlay and suffer through the downtime.
I did have a strange experience this summer. One of our very good customers has me out every three months to do preventive maintenance on their screws. Every thing in this plant is in fine shape and the operator lets nothing go (his engine room floors are cleaner than my kitchen) all problems are addressed immediately. All back up pumps etc are serviced and it is a pleasure to go there.
I inspected the larger screw (which was not online when I arrived). I had arrived prepared to stay the night if necessary.
Every thing went well. The operator helped as I calibrated the sensors and checked the motor alignment. I was done in 1/2 a day!
We started the screw. I started to pack my tools and went back to look at the machine. The head pressure had risen by 40 lbs. (2.8 BAR) and the oil temperature was climbing rapidly.
The condenser has three motors with two prop fans running off a jackshaft for each motor. One of the motors had failed.
With the increase in head pressure there was not enough capacity to maintain oil temperature.
This was a factory built package with the components selected by them. I believe we would not have been in trouble with liquid injection.
The point I am trying to make is that Thermo is not without its problems and not always viable in older plants.

Andy
26-11-2006, 07:00 PM
Hi NH3LVR:)

Thermo cooling has it's limit, getting very complicated in very large plantrooms with 10 or more compressors:eek:
Thermo is better suited to large packaged units or on large plant with a few compressors.
Water cooled is more suitable on multi compressor jobs, with a separate closed circuit evaporative condenser servicing the oil coolers only:)
Aternatively a separate coil in the condenser is possible but not as good:)

I have never tried it but I think a small evaporative condenser could be used very effectively to cool the oil (one for each compressor) directly, one for me to try some day:D

Kind Regards Andy:)

US Iceman
26-11-2006, 09:29 PM
Is the oil pumped around by the compressor's itself?
And for some reason one of the comp's keeps going out on oil pressure switch..any ideas why this may be?


You have to be careful when you review a screw compressor system. Some of these do not have external oil pumps (or integral either), so the oil pressure is the difference between discharge pressure and suction pressure (nibus the oil filter and piping pressure losses).

Loss of oil pressure can be caused by many things, but I would start with the easiest items first. Check the filter pressure drop first. Then other items as you see fit.

If an external oil cooler is used you have to inlcude all of the piping losses and the pressure loss of the cooler in calculating the NET oil pressure.

Adding an external REMOTE oil cooler requires more consideration than simply piping the lube oil lines to the cooler.

US Iceman
26-11-2006, 09:43 PM
Liquid injection oil cooling is a cheap way to perform oil cooling "on high stage compressors".

If the liquid is injected in the booster discharge lines, the only penalty is due to the compression of the additional mass flow in the high stage compressors.

If the liquid is injected into the screw booster, then the penalty remains in effect for the booster + the compression of the mass flow by the high stage compressor.

Some of the prevalent problems with liquid injection are:

lower oil quality
erosion of the rotor tips from the liquid ammonia expansion
Lack of control at lower liquid feed pressures
and, liquid line flashing in warm weatherThermosyphon systems can be difficult to arrange "properly" in multi-level or separated spaces, but it is possible.

Water-cooled (or glycol) cooled systems are much easier to install and offer some means of heat recovery for freezer floor warming, etc.

Air-cooled oil coolers can present problems too as you have to make sure the fins are kept clean all of the time. You also have to contend with the piping & cooler pressure losses so you do not lower the net oil pressure to the oil injection ports.

Another possibility on thermosyphon systems is to use a separate condenser for the thermosyphon system only. That allows you remove any oil logging problems, which sometimes occur in the oil coolers.

Thermosyphon systems are pretty simple. They either work or they don't. When they do not work, it is usually due to the piping installation, or a couple of other commmon problems.

Takeshi
04-12-2006, 07:44 PM
Thanks for the help guys. I will check this ASAP:)

Flimsical
20-07-2007, 03:44 PM
We use thermosyphon here at our plant, and even on the 100 degree plus days it runs nicely. Our system has 7 Frick screws, 1 500 horse, 3 250's, 2 100's and a 30 Hp. At least 5 of them run 24/7. Havn't seen a temp alarm in the 2 years i've been here.

gwapa
22-07-2007, 06:04 PM
A greeting for all
Just a question regarding Thermosiphon vessel

A thermosiphon vessel requiere a vortex breaker in the outlet going to the oil coolers?
Thanks for your very sharp post

mohamed khamis
23-07-2007, 10:20 AM
The exit oil pressure from the screw compressor is at the discharge pressure and the oil is usually returned to the compressor suction pressure. So without oil pump the oil can easily injected back to the compressor as most of screw compressors. The only benefit of using oil pump is to ensure continuous running of oil and the proper function of the oil to separate the two flutes during the compression process by an oil-film plus it overcomes the pressure drop in oil filter and external cooler if it exists. Therefore the oil pump is selected to generate oil pressure by only 200 kPa to 300 kPa over the discharge pressure to involve all the sources of the pressure drop (filter, cooler,ect..). IF the filter has been blocked partially the oil amount is decreased.Thermosyphon cooling, water-cooled and air cooled oil coolers cause an additional pressure drop in oil path but it is not significant as long as there is an installation of oil pump which will offsets that. In my view, The external cooling systems are more efficient than thermosyphon system in terms of energy saving. The oil is supposed to carry over the heat of compression burdens on the condenser again. Therefore in fact the condenser itself carries over the heat of compression instead of the oil. If we make a separate condenser it does make any modifications the separate condenser will need another coolant. The only benefit of thermosyphon is the simple construction and less cost for installation. In contrary at low ambient conditions the thermosyphon can cause acute load on the condenser due to the increase in latent load of the refrigerant to condense. However, in air-cooled oil cooler the fan can be cycled off or the speed can be modulated.

Cheers:)

US Iceman
23-07-2007, 03:52 PM
A thermosiphon vessel require a vortex Breaker (http://www.refrigeration-engineer.com/forums/glossary.php?do=viewglossary&term=127) in the outlet going to the oil coolers?


Hi gwapa, I recommend a vortex breaker for the reason you are asking. Most vessel manufacturers will put this in the liquid connection if you ask for it. But as a regular practice, I think they do not install them.

Vessel manufactuers are not system designers, and a vortex breaker relates to system design and operation.

US Iceman
23-07-2007, 03:59 PM
...The external cooling systems are more efficient than thermosyphon system in terms of energy saving.


I do not agree with this. A thermosiphon is an external oil cooling system and has the lowest operating cost of the other methods.

Andy
23-07-2007, 11:30 PM
Hi gwapa, I recommend a vortex breaker for the reason you are asking. Most vessel manufacturers will put this in the liquid connection if you ask for it. But as a regular practice, I think they do not install them.

Vessel manufactuers are not system designers, and a vortex breaker relates to system design and operation.

Yep you learn something everyday;)

Never specified a vortex breaker on a thermosyphon job, but I do know the pipe sizing is important, maybe larger pipework negates the need for a breaker, less velocity.

What about the return line, what way should this be selected:confused:

Kind Regards Andy:)

US Iceman
23-07-2007, 11:36 PM
Hi Andy,

I hope you are drying out. How has the work been? Busy?

I have always felt the vortex breaker is important to reduce the effects of the induced flow through the liquid connection from the vessel. We use these all the time in refrigerant pump suctions, but hardly ever in thermosiphon vessels. It's a mystery to me why more are not used.

Are you talking about the liquid return line from the oil coolers, or the vent line back to the condenser?

samiam
24-07-2007, 12:27 AM
Gents,

For industrail applications - my preference in TSOC, than liquid injection - thats what my superior taught me & it works.

Pipes need to be sized correctly (current vs future) and the physics does the rest.

Pipe sizing guideline differe from manufacturer to manufacturer.

gwapa
24-07-2007, 03:12 AM
Hello all
The thermosiphon tank is a very import part of the oil cooler system.
Most of the dising I have ever seen the outlet conection is in the bottom of the vessel .I personaly thing thant this point is better place to vortex formation.
When the tank is vertical I prefer take the out let to the oil cooler at one side of the shell (bottom)
If the tank is horizontal I install a vortex breaker.

The ammonia in the return pipe from the oil cooler is a two fase flow . The calculation of this kind of state only God can figure out. I select it at lest two diameter bigger than a normal sucction pipe
Best regards

US Iceman
24-07-2007, 03:54 AM
The calculation of this kind of state only God can figure out.


Not quite, but almost. ;)

mohamed khamis
24-07-2007, 05:08 AM
I do not agree with this. A thermosiphon is an external oil cooling system and has the lowest operating cost of the other methods.

Well...could u tell me the amount of circulated refrigerant which is responsible to cool the oil in the cooler is an additional flow rate over the needed for refrigeration purpose or it is a part of it. Viz, if the design flow rate of refrigerant is 0.8 kg/s which produces 3 tons refrigeration, in the thermosiphon cooling u will put extra mass to do the cooling or this mass is part of 0.8 kg/s. If u say it is an extra mass, ok this will burden on the compressor power to compress this extra flow rate (the power is larger than the fan power in case of air-cooled oil cooler). If u say no it is a partial amount of the design flow rate this reduces inevitably the cooling capacity and i think the first guess is the correct to avoid cooling capacity reduction. In addition, the thermosiphon cooling effect is influenced strongly by the condenser performance i.e. if there is a liquid back to the condenser the condensing effective area will be decreased and the condensing pressure will be boosted up and thus make the circulated refrigerant temperature is increased and affects badly on the oil cooler performance. Furthermore in low ambient (night) conditions and the cold room is highly loaded (high imposed cooling load) the condenser pressure is also increased responding to the high cooling load increment and on the same time the circulated refrigerant temperature will also increased affecting badly on the oil cooler performance unless the oil cooler is not insulated. On the other hand, the air-cooled oil cooler can exploit the low ambient temperature by cycling off the fan or by making the compression process is performed on almost isothermal compression and reduces significantly the compression power. I think the thermosiphon cooling is popular because requires virtually no maintenance and less in capital cost than the other cooling sources.

Cheers:)

US Iceman
24-07-2007, 06:11 AM
A first comment if you please... Can you use more paragraphs to separate your thoughts please?

I have trouble reading your posts when they are all in one big, long, single paragraph. Thanks in advance.

Now on to the comments...

On a thermosiphon oil cooling system, the mass flow resulting from the oil cooling load is additive to the compressor discharge mass flow (in the condenser). There is no effect on compressor power as the compressor does NOT compress the mass flow resulting from the oil cooling load (on a thermosiphon).



...if there is a liquid back to the condenser the condensing effective area will be decreased and the condensing pressure will be boosted up and thus make the circulated refrigerant temperature is increased and affects badly on the oil cooler performance.


That is very true and illustrates why it is important to ensure the condensers are properly piped for gravity drainage. This is a different problem.



Furthermore in low ambient (night) conditions and the cold room is highly loaded (high imposed cooling load) the condenser pressure is also increased responding to the high cooling load increment and on the same time the circulated refrigerant temperature will also increased affecting badly on the oil cooler performance unless the oil cooler is not insulated.


If the cooling load increases during the night, the discharge pressure may increase, but at the same time, the ambient wet bulb temperature usually decreases during the evening too. If the condensing liquid temperature does not exceed the design conditions there should be no effect on the oil cooling.



On the other hand, the air-cooled oil cooler can exploit the low ambient temperature by cycling off the fan or by making the compression process is performed on almost isothermal compression and reduces significantly the compression power.


The ONLY time the oil cooling load affects the compressor power is when liquid injection is used directly into the compression process. That is the only case where the power increases.

Liquid injection is typically the lowest installed cost of any oil cooling. All of the others have higher capital costs.

mohamed khamis
24-07-2007, 07:17 AM
A first comment if you please... Can you use more paragraphs to separate your thoughts please?
I have trouble reading your posts when they are all in one big, long, single paragraph. Thanks in advance.

First my apology for that and I am sorry for long paragraph



Now on to the comments...
On a thermosiphon oil cooling system, the mass flow resulting from the oil cooling load is additive to the compressor discharge mass flow (in the condenser). There is no effect on compressor power as the compressor does NOT compress the mass flow resulting from the oil cooling load (on a thermosiphon).

Ok after I made mass balance over the whole system I found out the compressor will NOT compress the the mass flow resulting from the oil cooling load (on a thermosiphon) during the steady state conditions. However, at the transient conditions IT will compress the whole flow rate involving this flow rate. So the compressor power in increased at the starting duration only.



If the cooling load increases during the night, the discharge pressure may increase, but at the same time, the ambient wet bulb temperature usually decreases during the evening too. If the condensing liquid temperature does not exceed the design conditions there should be no effect on the oil cooling.

What is relationship between the wet bulb temperature and the oil cooling:confused:?. I can NOT understand what u meant?. In addition, as I aforementioned, at low ambient conditions the air-cooled oil cooler can exploit the low ambient temperature by cycling off the fan or by making more cooling effect for the oil (not to the limit under of pour point temperature). This makes the compression process is undergone on almost isothermal compression and reduces significantly the compression power. In a simple term, the effective cooling of the oil (in low ambient condition) will decrease the oil temperature and thus the discharge temperature will be decreased and this will offset the increase in discharge pressure during the high load with low ambient condition. This is not being found in thermosiphon oil cooling system in this case.



The ONLY time the oil cooling load affects the compressor power is when liquid injection is used directly into the compression process. That is the only case where the power increases.


There is another issue and it is not only the compression process is harmed but also the compressor volume capacity due to the boiling off of the injection liquid refrigerant which increases the suction specific volume, reducing the amount of suction charge that can be drawn in. I wish now the post become clear

Regards:)

US Iceman
24-07-2007, 01:36 PM
Thank you for the paragraphs. That is a lot easier to read.;)



However, at the transient conditions IT will compress the whole flow rate involving this flow rate.


You must be using a thermosiphon in a unique manner, because I fail to see how any transient condition would result in compressing ANY mass flow from the oil cooling load(s).

If you think of the refrigerant from the compressor discharge and the vent line from the thermosiphon as two independent fluid streams that do not mix in the condenser, the mass flow from the oil coolers would never affect the compressor power.



What is relationship between the wet bulb temperature and the oil cooling:confused:?. I can NOT understand what u meant?.


I'm basing this on evaporative condensers for condensing the discharge mass flow of the compressor + the oil cooling loads.




In a simple term, the effective cooling of the oil (in low ambient condition) will decrease the oil temperature and thus the discharge temperature will be decreased and this will offset the increase in discharge pressure during the high load with low ambient condition. This is not being found in thermosiphon oil cooling system in this case.


The oil temperature has to be maintained in a range of a minimum temperature and a maximum temperature to provide uniform oil viscosity for the bearings. Additionally, if the oil injection temperature is too low you can create other problems such as increased oil dilution or high filter differential pressure.

Most industrial compressor packages use a temperature control valve for this purpose on the oil circuit. For small commercial screw compressors this may not be the case.



There is another issue and it is not only the compression process is harmed but also the compressor volume capacity due to the boiling off of the injection liquid refrigerant which increases the suction specific volume, reducing the amount of suction charge that can be drawn in.


The suction capacity may be decreased slightly due to a decrease in volumetric efficiency as the injected refrigerant increases the pressure of the thread in the rotors where it is injected. This causes a slightly higher pressure in the thread, which in turn can force some small amount of gas to "slip" past the rotor edge back to suction.

This does not account for a noticeable reduction in capacity (actual volume flow).

TXiceman
24-07-2007, 08:23 PM
As US Iceman notes...Liquid Injection or SOC (Screw Oil Cooling) (trademarked by FES) is a cheap way of cooling the compressor.

When done correctly, it works fine,but it will cost some in capacity and power usage. There is always the oil dilution problem, but with the newer oils, this is not so much of a problem as it was.

Thermosiphon is the more efficient way to go and has many advantages. The T-siphon system is really pretty simple, but if deals installers and some contractors a lot of misery. The piping runs and elevations are critical and can be a real problem in many access restricted areas. It is also a lot more expensive to install due to the extra vessel and piping.

To condemn all liquid injection systems is to condemn all cars because one person had a problem with a tire. With anything, proper application and design are critical.

Ken

mohamed khamis
25-07-2007, 05:48 AM
Well ....It sounds the discussion here will be very interesting US Iceman



I'm basing this on evaporative condensers for condensing the discharge mass flow of the compressor + the oil cooling loads.

Ok I had a strong belief u would mention like that so my reply is:

Either evaporative condenser or air-cooled condenser or water-cooled condenser with cooling tower will be subjected to increase in condensing pressure responding to the high imposed cooling load regardless the reduction in either dry bulb or wet bulb temperature.

So the oil cooler will be less effective as a result of the increasing in the liquid refrigerant in it which may increase the discharge temperature and compressor power. While the air-cooled oil cooler at this moment is more effective as a result of reduction in dry bulb temperature.

Could u answer to me these questions:

[1] In summer conditions, the Thermosiphon cooling is effective or water-cooled oil cooler using to warm the cold room flour?. In other word, Which system consumes less power which involves "Thermosiphon or water-cooled oil cooler" ?.

[2] In winter conditions (with high cooling load), the Thermosiphon cooling is effective or air-cooled oil cooler using to warm the cold room flour?. In other word, Which system consumes less power which involves "Thermosiphon or air-cooled oil cooler" ?.

All these questions to show that the Thermosiphon cooling is not the most energy saving in the external oil cooling systems.



Additionally, if the oil injection temperature is too low you can create other problems such as increased oil dilution or high filter differential pressure.

As aforementioned in the previous post the oil will be cooled to limit above the pour point temperature to avoid the oil solidification in the tubes and severe filter differential pressure. Alternatively, normally the pour point temperature of the oil should be well below the evaporating temperature at lowest cooling load to avoid the oil to congeal on the evaporator tube surface. Thus, we have a big range to cool the oil to the superheated temperature of the refrigerant and no chance for refrigerant condensation inside the compressor and no chance for oil dilution as u mentioned.

Could u tell me which compressor consumes less power; one compresses the refrigerant on polytropic process and other on isothermal process?. I think u will answer the second one, As the compression process be accomplished with effective cooling the net work input will be smaller. On the other hand, i mentioned earlier in the previous post i will make compromise between the effcetive cooling in air cooler and the energy saving by cycling off the fans and pour point temperature of the oil.



Most industrial compressor packages use a temperature control valve for this purpose on the oil circuit. For small commercial screw compressors this may not be the case.

I fully agree with this statement but i think it is widely existed on liquid injection oil cooling.



The suction capacity may be decreased slightly due to a decrease in volumetric efficiency as the injected refrigerant increases the pressure of the thread in the rotors where it is injected. This causes a slightly higher pressure in the thread, which in turn can force some small amount of gas to "slip" past the rotor edge back to suction.This does not account for a noticeable reduction in capacity (actual volume flow).

Unfortunately i disagree with this paragraph, the liquid injection is evaporated inside the compressor and leaks
through internal compressor clearances to the suction side. This liquid will expand into the suction as vapor, with a large increase in specific volume, reducing the amount of suction charge that can be drawn in. This
causes a reduction in capacity with liquid injection, with larger effect at high compression ratios.

Cheers

lldelfin67
25-07-2007, 09:42 PM
Hi,
Liquid