View Full Version : Top Entry/ Bottom Entry concept for plate freezers

11-06-2008, 12:41 PM
There had been lots of discussions on this aspect earlier. But plate freezer manufacturers somehow have a standard design for the headers. ( suction as well as Liquid). What should be the techincally correct connections to each plate in each of this concept.?

US Iceman
11-06-2008, 01:22 PM
Well, I believe this goes back into the same discussion as air-cooling evaporators. For years (very many in fact) the general recommendations are to use bottom-feed for applications where you need to defrost the coil. For air-defrost you could use top-fed coils.

And as you might guess, there will be a lot of opinions on the only correct way to do this. And secondly, this type of logic has been perpetuated for so long I think everyone repeats the same logic verbatim and accepts it without any further consideration.

There has been some discussion here before about plate freezers. I'll be the first to admit I have no experience with them. However, from a conceptual approach evaporators typically work by using a phase change process (liquid boiling). While on the plate freezers I am to understand they might use a different approach of just pumping liquid only through the plates (sensible only the way I see it).

From my perspective pumping 100% liquid through the plate would require more liquid as this (to me) would appear to be a sensible heat transfer mechanism. If this were the case then I don't think it matters which is top or bottom; it's all liquid.

This should be an interesting discussion.;)

12-06-2008, 04:26 AM
I have seen no of plate freezer catalogues made in India & do not find any specific consideration for this concept. But at the same time it is rather important to get this issue sorted out in proper technical perspective. ( Like Orifices in the plate freezer plates)

13-06-2008, 09:24 AM
With liquid overfeed i believe circulation rate is about 16:1 usually with manual expansion valve set to this requirement.Its always nice if suction riser is kept to absolute minimum for better performance.

US Iceman
13-06-2008, 01:38 PM
With liquid overfeed i believe circulation rate is about 16:1...

What I cannot fathom is why the circulation rate is so high. Any thoughts on this Ranger1?

13-06-2008, 10:49 PM
I will try to find out theory of it.

US Iceman
14-06-2008, 12:18 AM

The only thing I can think of why the circulation rate would be so high is because this must be sensible cooling only such as; Q=m*Cp*∆T instead of latent cooling as Q=m*∆h.

If that's not the answer then I'm all ears!:D

14-06-2008, 10:59 AM
There had been lots of discussions on this aspect earlier. But plate freezer manufacturers somehow have a standard design for the headers. ( suction as well as Liquid). What should be the technically correct connections to each plate in each of this concept.?
Hi Guys.
Follow this link below.
Posted by Winfred.Dela not long ago.
It has a brilliant Alfa Laval Manual packed full of plate heat exchanger info.

It's huge mind!
Cheers Grizzly

15-06-2008, 07:07 AM
No Doubt that information is very exhaustive & to the point. But still it misses out on the plate freezers though.

15-06-2008, 08:11 AM
Smpsmp45 and everyone else.

My sincere appologies as the posted link is NOT the one I intended.
As you point out S-45 its irrelivent to you enquiry.

US Iceman
15-06-2008, 04:57 PM
Grizzly, you still get points for trying to help.;)

15-06-2008, 10:34 PM
Grizzly, you still get points for trying to help.;)

You are very kind Sir!!!
I have found the details I wanted to post but as the file is so large.
I have forwarded them to smpsmp45 direct.
Hopefully it will reach him?
Cheers Grizzly:D

US Iceman
16-06-2008, 12:01 AM
Which files are you sending Grizzly? The one from Alfa-Laval or others?

16-06-2008, 05:57 AM
Plate freezers are a mine feild, some have orifices and some don't, some have headers with individual feeders to plates / and a series of plates. Ask supplier for install requirements, main thing allow suitable wet suction line back to surge/accumulator. Pressure drop is the killer, you cannot do all refrig. work in accumulator, plus make accumulator horizontal for surface area, good luck. A word for wise is make sure the client does not use corrigated cardboard cartons if the product is pre-cartoned, that miniscrual air gap will extend freeze time expotentially. Learnt the hard way.

16-06-2008, 06:13 AM
Hi Magoo.

Seems to be practical person. Can you pl. share your experiences with the main topic points?

16-06-2008, 06:15 AM
I have the file which Grizzly has sent. I shall try & post it here through you sendit.

16-06-2008, 06:15 AM
Which files are you sending Grizzly? The one from Alfa-Laval or others?


16-06-2008, 07:00 AM
you have answered your own question, ask plate surpliers what is there criteria for feed rate [ 16: 1 is over the top in my opinion. ] The critical thing is wet suction pipe diametor between plates and accumulator, make PD absolute minimum, and as well defrosting plates will throw systems into destruct mode at compressor end. Loading of compressor should be slow to the point of being painful, but the surging in accumulator is low. Economiser control is also critical, do not start economizer below 75% compressor load, otherwise the econo; will flood and kill oil pressure.
I will say again , talk to plate suppliers for refrig; criteria. Horizontal/vertical plates. Blackstone/sabroe, what ever brand they all have critical supply stuff. Get it wrong and they will bite you. All plates are galleried for distribution and feed rates, like I said earlier some have individual orifices, some have multi group orifices, some don't have anything and rely of actual feed rates. Beeeee careful.

16-06-2008, 09:46 AM
I could get the data from one of the manufacturers to know how they achieve Bottom feed.

Though the Supply & Return points are at the same positions, The liquid header is divided in two parts to make it bottom entry.

16-06-2008, 09:48 AM
I have also seen that all the plate freezer manufacturers have Flanged conections from headers to plates & all imported have ball socket type joints. ANy idea how that is manufactured. We are asking Indian manufacturers to provide ball & socket type joints, but are not very keen. The flanged connections are always found to be leaking over a period of time.

16-06-2008, 10:43 AM
The 16:1 circulation rate on austalian/new zealand manufactured plates has come from extensive testing .This may only apply to the above large plates due to their circuiting.
Maybe they dont want to many gas bubbles lingering in circuit reducing liquid contact.
It is really a brine and some dont like the huge amount of ammonia being pumped around for obvious safety reasons.
Top/bottom entry surely wont matter as long as header to each plate has orifice/reg valve

US Iceman
16-06-2008, 03:48 PM
Thanks for the drawing smpsmp45. Now I understand what the problem is. As shown in the drawing the extra header in the liquid header is a means to achieve a balanced flow for the evaporators. In hydronic systems this is called a reverse-return piping arrangement.

If you follow the liquid flow, it enters the bottom plate first. This plates is also the first one to enter the suction header. The top plate is the last to receive liquid, which also makes this plate the one closest to the outlet of the suction header. They are trying to balance the liquid flow to the plates by using the extra header.

If you over-pump the liquid to the plates so that all of the refrigerant stays liquid, then you would have the problem of balancing the liquid flow to each evaporator.

Top/bottom entry surely wont matter as long as header to each plate has orifice/reg valve.

Now we are getting somewhere.

This I agree with. With this method you are treating each plate as a separate evaporator. In the drawing provided by smpsmp45 the plates are treated as a group of evaporators with one liquid feed valve train. This type of approach would undoubtedly increase the issues with obtaining a reasonable feed rate into each of the evaporators (not just the group of plates).

From a purely theoretical viewpoint, I think it makes more sense to treat each plate as a single evaporator with a single feed control valve (hand expansion, etc) with a separate suction line.

From what I have just seen it would appear the high overfeed rates are based on the issue of solving a liquid balancing problem between the plates to achieve a uniform temperature/capacity.

17-06-2008, 10:34 AM
Maybe the plates are similar to shell/tube water cooled oil coolers.In its requirement for turbulent water flow for good heat transfer.

18-06-2008, 12:03 PM
I am attaching a file from Stoecker on this subject. That book also gives important issues on circulation raios too

US Iceman
18-06-2008, 04:13 PM

Thanks for the file. If you notice in figure 8.10.a this shows a similar effect of coil circuiting as in the previous drawing you posted. The refrigerant enters on the bottom. The first circuit receives liquid initially and this same circuit is also the last one in the outlet header. If you follow the inlet header to the last circuit then follow that circuit to the outlet header you will notice it is the first circuit closest to the outlet header.

This is simple hydraulics by balancing the pressure losses across the circuits, which is determined by the piping arrangement.

This is identical to that drawing of the plates you posted earlier. Just follow the flow through the circuits one circuit at a time.

Everything that is being described in that section is based on the way we do things right now. Does it work? Yes. Are there other ways to do this? Again I think so based on what I have seen and done.

Ranger1's comment is accurate I think. If you are only pumping liquid through a heat exchanger you need to keep the fluid velocity high enough to achieve turbulent flow (where the best heat transfer occurs).

Since the plate evaporators are grouped (as the circuits in the noted figure show) then you either have to try to balance the flows hydraulically (by equal total circuit lengths) or meter each circuit independently.

Using a phase change mechanism makes more sense to me than using a sensible method requiring a lot of liquid pumping.

Personally, I think there is a better way of doing this.

19-06-2008, 07:00 AM
If I recoolevt my earlier mails which were based on extensive interaction with various top people in thie field & the conclusions was that circulation ratios even in the range of 2:1 are good enough. The Important point was to have latent heat aspect than sensible hat of the liquid. Even Stocker manual shows that max ratio has been 8:1. WIth Newzeland experience of 16:1 ratio with no reference to Top / Bottom feed etc seems strange. Probably there are no enough eforts to establish the correct means of coming to conclusion on this point. When TOP/ Bottom Feed etc have been points of discussions even in Stocker, It is difficult for me to jump to a conclusion based on some practical experience. The plate freezer manufacturers too have neglected this point till date.

US Iceman
19-06-2008, 12:33 PM
I think part of the problem is that the industry continues to do the same thing all the time because of inertia. You see this all the time. Very few new developments, except various types of electronic controls, arguments about which compressor is best/more efficient, etc. You seldom see any manufacturer change methods, however it seems they continue to re-invent the same old thing based on experiences that happened 40 years ago.:confused:

From my perspective I like to break things down into their fundamental principles. If you follow these, then you can't get into trouble as long as the component is installed and operated in accordance with those basic concepts.;)

19-06-2008, 04:32 PM
I agree with you US ICeman 100%. Even we too have started working the way you have described as I noticed that few small points do make a lot of difference. Over a period of time clients do find what is the correct way of doing things & why so minor things are critical in the first place itself. I shall try & post our other experience on similar issue which looks minor on the face of it.

US Iceman
19-06-2008, 04:37 PM
It is from these small details that a majority of the problems can be solved. And...what I have found is: people learn these and it makes life much easier to work on refrigeration systems.;)