PDA

View Full Version : Gravity Recirculated PHE



gwapa
24-12-2006, 02:52 AM
Dear Friends
A costumer bought a Gravity Recirculated PHE to cool wort from 32°C to 0°C .
The ammonia separator tank was designed to be at 2 m high. Now they want to locate the tank at 10m hight in a second floor

What could be the performance of the PHE with the tank located a 10 m?
thanks for your advices

Josip
24-12-2006, 11:31 AM
Hi, Gwapa :)


Dear Friends
A costumer bought a Gravity Recirculated PHE to cool wort from 32°C to 0°C .
The ammonia separator tank was designed to be at 2 m high. Now they want to locate the tank at 10m hight in a second floor

What could be the performance of the PHE with the tank located a 10 m?
thanks for your advices

I think the performance will be the same but you can use smaller feeding valves (see for Danfoss ICM-EXP - motor operated expansion valve)

Best regards, Josip :)

gwapa
26-12-2006, 12:12 AM
Thanks Josip
I am worry due to the fact that with the separator located at 10 m hight there will be a hidrostatic presure in the riser (bigger) than make me lower the main sucction presure in the tank. We are cooling to close to the freezing point so if we lower the succion pressure in an unstable time we run the risk of froze the Plate Heat exchanger
What do you think about?

Josip
26-12-2006, 01:15 PM
Hi, Gwapa :)


Thanks Josip
I am worry due to the fact that with the separator located at 10 m hight there will be a hidrostatic presure in the riser (bigger) than make me lower the main sucction presure in the tank. We are cooling to close to the freezing point so if we lower the succion pressure in an unstable time we run the risk of froze the Plate Heat exchanger
What do you think about?

I do not follow this....

Can you tell us what are you cooling exactly (what kind of wort) - direct or indirect cooling- maybe beer;)

You have only hydrostatic pressure in the liquid feeding line in return line over PHE there should be a wet gas. Regulation of liquid inlet into PHE you must regulate by regulating valve. (There are some other devices for regulation..)

You must have some protection devices (flow switch and digital thermometer) to avoid the risk of freezing.

Can you upload some scheme of your plant with some more details, please?

Best regards, Josip :)

Rio69
26-12-2006, 03:05 PM
Hi Gwapa,
you are right. Static pressure will cause evaporating pressure to increase in the PHE, and therefore performance of your PHE will be decreased. According to my experience you should avoid that solution, because there are also other issue that can be affected, f.i. recirculation rate; Nucl. Boiling Point etc.
I would recommend you to ask manufacturer of PHE.
That is what we do when we are in doubt.

BR, Rio

gwapa
27-12-2006, 02:33 AM
Jisip /BS Rio

I attached a flow drawing of the PHE
Yes it is a beer wort cooler for a local brewery
The brewmaster do not want to located the separator tank in the beer area,so he wants to rise it at 10m hight outside the beer area

I agree that we have to lower the pressure in the separator and so the succion of compressors. I am afraid that in unsteady conditions the PHE could frost

Renato RR
27-12-2006, 11:10 AM
Well I was thinking that only two companies in Croatia do the Amonia are you guys compettition or from same company?

Renato

Josip
27-12-2006, 11:56 AM
Hi, Gwapa :)


Jisip /BS Rio

I attached a flow drawing of the PHE
Yes it is a beer wort cooler for a local brewery
The brewmaster do not want to located the separator tank in the beer area,so he wants to rise it at 10m hight outside the beer area

I agree that we have to lower the pressure in the separator and so the succion of compressors. I am afraid that in unsteady conditions the PHE could frost

One question: What about cleaning and washing of that PHE (CIP)

To avoid frost there are some other solutions...

Your PHE is connected to one U shape pipe one side connected to liquid phase and another one to gas phase in separator. Thus we should have the same liquid level in U pipe and the same pressure. I do not think that static liquid pressure could affect evaporation. That can be a problem if your return pipe is in very cold area without insulation (like for oil coolers) thus you have again condensation of gas and return of that liquid but this can be avoided with some kind of constant level control to have flooded only PHE. After this valve-level control with flooded PHE we have only suction pressure in that area and complete static pressure in liquid feeding pipe down to PHE we use only to feed PHE in the moment when we need.

Only what you need is to obtain good flow of gas bubbles (purging) in return pipe i.e. constant rising of return pipe.

This is only my opinion. I do not have experience with such case and maybe I am not right...;)

But, here we have other guys too...

I have some other idea also but I will come back later...

Best regards, Josip :)

Josip
27-12-2006, 12:02 PM
Hi, Renato :)


Well I was thinking that only two companies in Croatia do the Amonia are you guys compettition or from same company?

Renato

Which two?

I'm freelance RE.

I do not know who is Mr. Rio69 but he can tell us something more about himself;)

Best regards, Josip :)

Renato RR
28-12-2006, 09:49 AM
Terma HGK,Tehno frigo?

Best regards,
Renato

Rio69
28-12-2006, 12:13 PM
Frigoterm Zgb.
You?

Rio

Renato RR
28-12-2006, 06:40 PM
Elektrolux

Best regards,
Renato:cool:

NH3LVR
28-12-2006, 07:41 PM
Gwapa;
I looked your drawing over a bit. It seems to be the standard set up.
Throwing the 10 M height into it may cause a problem similar to a condition called "Brining", when a Pump Recirculated Evaporator is overfed. The ideal way to adjust a Pump Recirculated Evap is to increase the feed until the TD across the Evap drops of a bit, then back off on the feed rate a little bit. In reality this is not often done.
I believe this may well cause the same problem.But I have never located a vessel that far above a Evap so this is only an opinion.

Peter_1
28-12-2006, 07:41 PM
Thanks Josip
I am worry due to the fact that with the separator located at 10 m hight there will be a hidrostatic presure in the riser (bigger) than make me lower the main sucction presure in the tank. We are cooling to close to the freezing point so if we lower the succion pressure in an unstable time we run the risk of froze the Plate Heat exchanger
What do you think about?

Perhaps I'm wrong but will the temeprature in the PHE not increase instead of decrease? So less danger of frosting up the PHE?

What about suction temperature of -6°C and T NH3 at -2.5 °C? This is confusing for me.

US Iceman
29-12-2006, 02:31 AM
The higher vessel height will increase the static pressure of the liquid refrigerant. Where the liquid drop leg enters the PHE the total pressure will be equal to the evaporating pressure (in the PHE) plus the static pressure of the liquid column 10 meters above the PHE.

The increase in total pressure will also increase the boiling point of the liquid. Determine the saturation temperature of the evaporating pressure and subtract it from the saturation temperature of the total pressure entering the PHE.

You will find there is a higher equivalent temperature at the liquid inlet to the PHE. This increases the evaporating pressure in the PHE and also decreases the overall LMTD. In effect this causes the PHE capacity to decrease.

In order to get the "design" evaporating temperature, the pressure at the inlet pressure regulator will have to be decreased. The decrease in pressure would be equal to the static pressure of the liquid.

I have seen several PHE's and flooded air-cooling coils where this has been done. There always seem to be some problems.

30 meters is quite high. The normal height is to have the vessel just above the PHE.

Andy P
29-12-2006, 03:56 PM
Maybe I am not as optimistic as the rest of the guys here - I would expect trouble if a gravity fed PHE was changed from 2m head to 10m head. Part of the trouble will come from the effort needed to persuade the gas bubbles to rise up through 10m of liquid in the wet return, and the consequent reduction of overfeed rate that will result from this. Assuming that the system with 2m head worked well (which is a big "if" in some of these systems!) you can reckon on the evaporating temperature being lower (due to reduced overfeed rate). As PHEs do not provide uniform flow on the beer side, and as wort will have a freezing point not much below 0oC, I suspect that your heat exchanger will freeze up on the beer side.

It might be possible to pipe the liquid feed from the expansion valve straight to the plate inlet, and control it to give superheat in the riser (making it a dry suction), but you might lose some cooling capacity by doing this.

cheers

Andy P

US Iceman
29-12-2006, 05:51 PM
This is getting into an area I find really fascinating, i.e., how does the actual recirculation rate change due to applied heat flux and the resulting hydraulic balance in the system (PHE & piping)?

Heat flux is relatively simple to model, but the big question is how to model the hydraulics in the riser for two-phase flow and pressure drop. This is identical to the same problems faced with pipe sizing for thermosipon oil cooling systems.

All of the single-phase gas or liquid is simple. The two-phase area raises a lot of questions on methodology used and the equations for the model.

I have seen some cases where the vessel was mounted about 10 meters higher than the heat exchanger and nothing worked properly. Insufficient capacity, failure to acheive stable evaporating pressure, etc.

One line of an argument that is sometimes used is the resulting increase in static pressure causes the evaporating pressure to rise, which makes some sense. Total pressure = vapor pressure + static pressure. The saturation temperature of the total pressure data will be higher.

In this case, someone suggested the use of a hand expansion valve to decrease the total pressure down to the desired evaporating pressure. In other words, their recommendation was based on the fact that by throttling the liquid pressure down to the desired evaporating pressure the system should work.

In effect, the hand expansion valve was supposed to decrease the effect of the static pressure.

I have always felt this methodology has some pitfalls as it would tend to decrease the actual recirculation rate of the coil. To me this appears to change the dynamics to something that more closely approximates a DX operation.

And, from Andy_P's comments, it appears this is the case.

Most of the time, the resulting higher evaporating pressure (due to the increase in total pressure) is compensated for by a reduction in the actual suction pressure.

I commonly find these installations with the suction presssure operating lower than the design in an attempt to reach the "design" capacity.

If anyone has some ideas or comments for discussion I would like to pursue this with them.

Andy P
31-12-2006, 08:34 PM
All of the single-phase gas or liquid is simple. The two-phase area raises a lot of questions on methodology used and the equations for the model.

...and made more complex by the fact that the gas will be rising faster than the liquid (bubbles going up through the liquid flow) even when there is a liquid overfeed.

We made some mods to a plant that had gravity fed vessels at 5m above the PHEs - performance was much better once they were reduced to 2m above!

If the original 2m design in the brewery has high pressure drop plates then you could reduce the adverse effect by changing them for low pressure drop plates, but if you are going to all that expense it would be cheaper to relocate the plates as well as the drum!

cheers

Andy P

Josip
01-01-2007, 12:04 PM
This is something what I was thinking about before. Due to language problems it is much better to make a drawing;) at least for me:D

Best regards, Josip :)

Andy P
02-01-2007, 10:11 PM
Dear Josip,

Thanks for the sketch - I'm not sure that the brewmaster would be happy to have all the controls still in the lower level even though the main drum was up high. That's assuming that his objection was based on the need for maintenance and/or the risk of leakage from these components. You would need to be very careful in the piping of the ammonia outlet from the plates to make sure that the liquid and gas went where you wanted them to - if the float chamber gas-locked then the system would still have a 10m wet riser and could flood the compressor

cheers

Andy P (still pessimistic!)

US Iceman
02-01-2007, 11:00 PM
Hi Josip,

I printed out your drawing and wanted to think about it.

I think your idea of using the smaller surge drum with the liquid feed controls eliminates the problems due to excess static head. In effect, the liquid feed valve train decouples the static head from the PHE.

In this instance, the low pressure liquid in the surge drum would be at the evaporating pressure + the riser pressure loss the way I see it.

The small amount of flash gas generated in the small surge drum would be vented to the riser at least from a visual review.

However, I think the dominant problem that remains is that all of the controls are still in the brewery and the more important item (at least to me) is the need to accurately size the riser for all load conditions.

If the PHE load is reduced very much (or if the riser is sized incorrectly), the flow regime in the riser could revert into a solid column of liquid with some bubbles. This would also increase the evaporating pressure in the PHE.

If the load variation on the PHE is large it might suggest the need for dual suction risers, which compounds the riser selection process.

In either case, if the brewmaster wants all of the valves out of the area we still have the same problem.

Josip
03-01-2007, 12:07 AM
Hi, Andy P :)


Dear Josip,

Thanks for the sketch - I'm not sure that the brewmaster would be happy to have all the controls still in the lower level even though the main drum was up high. That's assuming that his objection was based on the need for maintenance and/or the risk of leakage from these components. You would need to be very careful in the piping of the ammonia outlet from the plates to make sure that the liquid and gas went where you wanted them to - if the float chamber gas-locked then the system would still have a 10m wet riser and could flood the compressor

cheers

Andy P (still pessimistic!)

Speaking about brewmaster:

Gwapa is not in very nice position because he has to obey his customer's wishes or he has to quit the job, see attachment ;)

On another hand we never ever try to tech them how to brew beer but they always try to teach us our job :eek:

I am not sure about maintenance (maybe to save some space) but anyhow in case of leak liquid ammonia will drop down (now from 10m height), or

What do you mean with:
if the float chamber gas-locked then

Regarding return line we can make a very low level of ammonia in top separator (just at the bottom - in that case we do not have enough ammonia to flood the compressor), but we still have 10m column. Low level float chamber is there just to reduce high static pressure (it has a role of low level separator) and regulator valve (not expansion valve) is to regulate inlet of liquid ammonia into PHE.

In case of lock of float chamber we can fill liquid in both pipes and in top separator only up to low level.

Of course if everything went wrong Murphy has been right again ;)


Best regards, Josip :)

Josip
03-01-2007, 12:35 AM
Hi, US Iceman :)


Hi Josip,

I printed out your drawing and wanted to think about it.

I think your idea of using the smaller surge drum with the liquid feed controls eliminates the problems due to excess static head. In effect, the liquid feed valve train decouples the static head from the PHE.

In this instance, the low pressure liquid in the surge drum would be at the evaporating pressure + the riser pressure loss the way I see it.

The small amount of flash gas generated in the small surge drum would be vented to the riser at least from a visual review.

However, I think the dominant problem that remains is that all of the controls are still in the brewery and the more important item (at least to me) is the need to accurately size the riser for all load conditions.

If the PHE load is reduced very much (or if the riser is sized incorrectly), the flow regime in the riser could revert into a solid column of liquid with some bubbles. This would also increase the evaporating pressure in the PHE.

If the load variation on the PHE is large it might suggest the need for dual suction risers, which compounds the riser selection process.

In either case, if the brewmaster wants all of the valves out of the area we still have the same problem.

My idea was to eliminate high static liquid pressure in liquid feed line and easy regulating ammonia level within PHE via regulating valve to avoid generation of liquid column in suction pipe.

Regarding unequal load, from my brewery experience;) I can say they have the same cooling load for very long time, until they empty one silo what is many 000 of liters. Than they have to stop to make CIP (there is not replay from Gwapa about, but that is extremely important).

To put all valves out of the area is possible to request but I will never obey.

If they want ammonia free brewery they have to install indirect cooling system, no other way, but that is another story.

Best regards, Josip :)

US Iceman
03-01-2007, 12:45 AM
Hi Josip,

As you say, the customer or Mr. Murphy always win, and they are NEVER wrong.

My only experience with beer (besides getting a hangover) is very little. If the loads are constant until CIP, then the use of dual suction risers would not be necassary then.

I see what you are trying to do now...;)

gwapa
03-01-2007, 02:37 PM
Dear gentlemen
First of all I wish all you a very cool new year

Josip is right regarding to the operation of the PHE . We divided the operation in steps

Step 1 The PHE runnig in normal way. That is the flow and temperatures of beer most constant
Step 2 The beer inside the PHE is pull out with water. A viscosity sensor detect the water
Step 3 Start the CIP . Sometime with hot solutions which push the ammonia inside the PHE towards the separator. In this step the setting of the BPR Valve is changed to higher setting. Also we take advantage of the pressure a drain the oil
Step 4 the water inside the PHE is pullout with beer and start again Step 1

As you could observe the separator tank is very import due to the fact it is also a holding tank

It is impossible to eliminate the ammonia valves and controls inside the room . The only way is change to a secondary refrigerant

The brewmaster does not want to have the tank inside the processing area, that’s all.

Andy P , wrote something interesting. He think that as long we increase the ammonia inlet pressure we will have an higher available pressure drop across the PHE. That could means that the PHE could handle more flow of beer.

The Josip’s idea should be analyzed . In Step 4 we must fill the PHE rapidly otherwise a big batch of beer will leave at a higher temperature
gwapa

Andy P
05-01-2007, 12:49 AM
Andy P , wrote something interesting. He think that as long we increase the ammonia inlet pressure we will have an higher available pressure drop across the PHE. That could means that the PHE could handle more flow of beer.

gwapa

Hi Gwapa,

I think you may have misunderstood me at some point - sorry I wasn't clearer. The effect of increasing the height will be to reduce the flowrate of ammonia through the PHE, so even if you have a lower ammonia temperature you will not necessarily have a higher beer cooling capacity because the heat transfer coefficient is a function of TD and flow pattern and low flow is not good. Also as I said if the evap temp drops much below freezing point there will be a tendency for ice to form in the stagnant areas on the beer side (places where the beer velocity is low, such as around the edges of the plate, below the inlet port etc etc). This ice may not completely block the plate but it will reduce the useable surface area - again the beer cooling capability will be reduced.

cheers

Andy P

PS - Thanks Josip: many laughs out loud in your posts!!

Mark Sanchez
05-01-2007, 02:05 AM
We discussed similiar situations a few times at the local RETA (Refrigeration Engineers &Technicians Association).

A problem realized with similiar instances is the static head increases the realized saturated pressure at the evaporator/Heat exchanger.
The flow rate is also increased due to the elevated static pressure, which could cause brining.

A few jobs have been fixed by installing a hand expansion valve on the liquid inlet to the evap/hx.

This decreases the pressure and reduces flow.

I may not be fully up to speed with the changes the plant is making, but noticed that a smaller surge drum may have been applied.
I would recommend that someone calculate the dropout for the reduced vessel. this may have a larger impct during CIP.

Just a few thoughts
Mark

Mr. Ammonia
14-02-2007, 01:50 AM
I have had some interesting experiences in related problems which could be used in this case:

The tests we made where for knowing how a system is going to react when a VERY large vertical suction separator (also used as surge drum) is changed from "nearly empty" to "nearly full" condition and therefore the liquid column level is going to change nearly two meters.

What we discovered is that the PHE always tries to do "the same": If you enter the same flow of water at the same conditions, the ammonia evaporates at the same pressure/temperature. The main difference is that the ammonia flow is going to increase and consequently the outlet quality (the % of liquid in the flow after the PHE) is increased.

How much the ammonia flow is increased? The required in order to compensate the extra ammonia column static pressure with the extra gas pressure drop (higher velocity).

The problem comes when the PHE has only one outlet connection, which can cause maldristribution problems and you can finally even loss performance in the PHE!!!

The most important conclusion is that as the evaporator is working at nearly the same conditions (evaporating pressure) but the pressure drop of gas is bigger (equal the liquid column static pressure), the compressor will work at LOWER pressure. Therefore less cooling production and worse COP.

We have made some insteresting tests wherever we had this possibility of playing with the ammmonia level in the separator. As always the final results will depend in the exact plant design.

I hope this will be of some use!!!

TXiceman
14-02-2007, 04:04 AM
One very important point that needs to be considered...Submergence....

Due to the excessive height of the liquid feed column, you will now need to consider the difference in the boiling point of the liquid at the bottom of the column as opposed to the original design. It will probable require rerating the PHE.

Ken

Mr. Ammonia
16-02-2007, 01:53 PM
Dear Ken,

You are fully right.

If you are a user of Boyko-Kruzhilin formulation for the design of the Heat Exchanger, as soon as you enter in these conditions you need to use a Nucleate boiling multiplier of cero at the Exchanger inlet.

Best regards,

Mr. NH3

Andy
17-02-2007, 11:51 AM
Hi Guys:)

I have only came accross this post:o

would it be possible to change the system to liquid overfeed, either a pumped system OR a low pressure receiver:)

The low pressure receiver would be my bet with an electronically controlled evaporator pressure regulator on the plate wet return, with double suction risers back up to the LPR.

What does anybody else think?

Kind Regards Andy:)

Andy
17-02-2007, 11:56 AM
Hi Guys:)

I have only came accross this post:o

would it be possible to change the system to liquid overfeed, either a pumped system OR a low pressure receiver:)

The low pressure receiver would be my bet with an electronically controlled evaporator pressure regulator on the plate wet return, with double suction risers back up to the LPR.

What does anybody else think?

Kind Regards Andy:)

Or if cost is a problem, DX with electronic expansion valves would be the next best (keep the reference transducer up at the vessel for best flooding and use double suction risers)

Kind Regards Andy:)

Vlada
19-02-2007, 02:42 PM
Hi Gwapa,

We had experience with this problem. Rio69 is right in many points. Very complex things is happening PHE in this situations.

Recirculation rate 1:1 and complex control is not solution for flooded installations because many reasons!!

Contact manufacturer!

We lifted PHE on platform because it was only solution. You can find another place both for separator and PHE. Out of beer room.

Hi for croation neighbors!

Josip
21-02-2007, 08:47 PM
Hi, Vlada :)

Welcome to RE


Hi Gwapa,

We had experience with this problem. Rio69 is right in many points. Very complex things is happening PHE in this situations.

Recirculation rate 1:1 and complex control is not solution for flooded installations because many reasons!!

Contact manufacturer!

We lifted PHE on platform because it was only solution. You can find another place both for separator and PHE. Out of beer room.

Hi for croation neighbors!

Thanks, same to you, you can send a PM to me with your email address if you like (cannot send PM to you:confused:) maybe because you do not have 15 posts yet.

Where did you install that PHE?

Best regards, Josip :)

dolce
15-03-2007, 04:34 PM
Hi
you don't need 10m high that is enough 2 m to get a required performance
I'll send to you a pic. show you how

Josip
18-03-2007, 12:20 PM
Hi, Dolce :)


Hi
you don't need 10m high that is enough 2 m to get a required performance
I'll send to you a pic. show you how

Please, post it here;)

we all want to see how, (this is the place to learn something new for all of us:D )

Best regards, Josip :)

TXiceman
19-03-2007, 02:08 AM
The height of the thermosiphon (or gravity) tank is determined by the pressure drop through the exchanger and the pressure loss in the pipes. The static head has to be sufficient to over come the losses. Based on the flows, you can calculate the pressure losses in the lines .

The system will be self regulating and balance itself out at some point where you will be above 1 to 1 recir rate and hopefully at least 2 to 1.

Normally you design for something in the range of 6 to 10 feet of liquid height.

Ken

charlie n
04-04-2007, 04:23 AM
I'd be surprised if the cooler worked at all with the Liquid separator located more than 2 or 3 meters above the PHE. I normally build these with the separator located about 200mm above the top of the PHE. The static penalty will probably destroy the performance of this unit.