Strange TXV setup

[Gary]

The freezer sol valve, off when comp goes off (stops liquid migration) Do not worry about pump down, as trying to set a very low differential on this application would be difficult.

Good idea... but the solenoid valve should be in the common liquid line, preventing migration to all of the plates.

[Peter_1]

http://www.refrigeration-engineer.co...+liquid&page=3 post 47

[Chef]

Start 100F 1BTU/Lb/F
Freezing Point 32F (not quite true but close enough) 144BTU/F
Final 0F
Start 100F 0.5BTU/Lb/F
so 68* 68= 4624
68*144=9792
68*32*.5=

Well the plate freezes at -20C so we also need to account for that drop in temp and the boxes also have some thermal inertia that should be accounted for and The Grande Fromage will always be adding some food in the fridges and freezer as well but not too much.

So when I add those numbers up I still get the same result for the cooling effect.

Chef

[mad fridgie]

Well the plate freezes at -20C so we also need to account for that drop in temp and the boxes also have some thermal inertia that should be accounted for and The Grande Fromage will always be adding some food in the fridges and freezer as well but not too much.

So when I add those numbers up I still get the same result for the cooling effect.

Chef

Only half freezes at -20C, the other half is -6C(?)
But we are not dealing with water, so heat capacities stated are higher than actual, and the actual mass is less than calculated, i sort of allowed for this to cover the boxes, (maybe the Grande Fromage, is killing the calculations? lol)
I am presuming these are only little boxes (3-4 cuft each) Thin metal internal sheet coving some sort of blown insulation (so low thermal mass). Slow thermal inertia, need to run again for an hour.
My figures are correct give or take abit.

[Chef]

Your compressor calculations do not account for volumetric efficiency.(which is an unknown)

OK missed the last part of your calcs post - now I see it all.

Your right about the efficiency so if that is factored in I get 2.3Kw which is getting close to your 1.9Kw so it seems we can agree somewhere around this position.

But it is useful to hammer it out and get closer to the real values as it will be needed later to spec out a new compressor if thats the way it finally goes.

Thanks MF

Chef

[mad fridgie]

OK missed the last part of your calcs post - now I see it all.

Your right about the efficiency so if that is factored in I get 2.3Kw which is getting close to your 1.9Kw so it seems we can agree somewhere around this position.

But it is useful to hammer it out and get closer to the real values as it will be needed later to spec out a new compressor if thats the way it finally goes.

Thanks MF

Chef

It would be strange if 2 engineers could come up with the same figure, when we actually have no real known facts to work with.

[Peter_1]

Thanks, doing some very BASIC calcs, I think you are running at 30% volumetric efficeincy (very low)

Almost impossible to calculate this: you start with warm plates, compressor will evaporate till its max extreme LP limit (can be +10°C), so a very high COP ...IF...the condenser can handle this high load.
As eutectic temperature drops slowly, load drops, LP drops accordingly and so the COP. So you don't have a linear process but a compressor starting at +10°C going to -15°C LP and flattening a first long time at the eutectic point of the liquid with the highest temperature.

For me, there's no problem to connect them all together - we did it several times - but client (in our cases) must be aware that once the -5°C is frozen (where we already went through the eutectic point) that we will subcool it further down to solidify the plates with a colder eutectic solution. So the -5°C plate will while cooling down for the colder plates become also -10°C. If there's then that moment goods it the -5°C box, then those goods could be frozen.

Connecting them all together has an advantage: you only need once a zone in the last plate for superheating, otherwise you will need one in each plate.

All the copper tubes in the plates we make are placed/bended in such a way that refrigerant flows down so that no oil never can be collected in it. We made 4 plates 2 weeks ago in stainless steel from 2 x 1 m x 6 cm thick, connected to a 3/4 HP on R507


All the applications we have run only 6 to 8 hours overnight while holding then the whole day their cold.
The first day when they charge those up from ambient temperature, a charge of 20 to 24 hours is needed.
We're speaking then of a whole vending van.


We have applications with vans starting at -25°C and ending the day at -20°C (in the van)
The size of the compressor is only important if you need to cool it down within some time margin.

As far as this information could help you further (haven't read all the posts) read also once my link I posted some boxes higher here

[Chef]

Only half freezes at -20C, the other half is -6C(?)
But we are not dealing with water, so heat capacities stated are higher than actual, and the actual mass is less than calculated, i sort of allowed for this to cover the boxes, (maybe the Grande Fromage, is killing the calculations? lol)
I am presuming these are only little boxes (3-4 cuft each) Thin metal internal sheet coving some sort of blown insulation (so low thermal mass). Slow thermal inertia, need to run again for an hour.
My figures are correct give or take abit.

The boxes are about 4 cu ft and they are mainly GRP (fibreglass) with some blown foam insulation. But a pull down from 30C is quite a lot of work and as you say the Grande Fromage always messes up the calculations, sometimes she even puts frozen food in the freezer before it is down to temp - what to do eh!

Chef

[mad fridgie]

The boxes are about 4 cu ft and they are mainly GRP (fibreglass) with some blown foam insulation. But a pull down from 30C is quite a lot of work and as you say the Grande Fromage always messes up the calculations, sometimes she even puts frozen food in the freezer before it is down to temp - what to do eh!

Chef

How dare SHE!!!!

So we need around 2KW, now you have choose what SST. If you choose -20C I believe you will be well on the top side.
Actual data for copeland is available on R22
"Select 7" free down load

[Chef]

For me, there's no problem to connect them all together - we did it several times - but client (in our cases) must be aware that once the -5°C is frozen (where we already went through the eutectic point) that we will subcool it further down to solidify the plates with a colder eutectic solution. So the -5°C plate will while cooling down for the colder plates become also -10°C. If there's then that moment goods it the -5°C box, then those goods could be frozen.


As far as this information could help you further (haven't read all the posts) read also once my link I posted some boxes higher here

Peter the point about the connected plates going below the set temp is part of the thread earlier and if we did indeed connect them and correct the TXV situation we would as you say get frozen stuff in a chiller. Thats really the crux of the problem, how to keep these boxes at the right temp. Seems dual TXV's is possible solution as well as leaving it as it is and just storing a bunch of compressors under a seat somewhere. Just not that elegant.

I read the post and and am not convinced either way about pumpdown - some good points from both sides of the fence really. Do I want to do it? sounds a bit scary and so maybe not.

As for the coils in the plates - pretty sure they start at the bottom and wind their way up the plate in an S fashion but with many oil trap loops.

Chef

[Peter_1]

In the chiller section, the temperatures decreases indeed to the same temperature as the coldest plate but as soon the compressor is shut off, the chiller plate rises very fast again to its eutectic point. So when charging during night, a small plate of PU is placed before the plate sometimes.

Or you could fill both up to -15°C and place a small isolated plate (we then use forex for this, this a foamed white plastic plate of 3 to 15 mm thick) before the eutectic plate in the chiller with on top of it a small fan (rectangle 120 x 120 as used in PC's) controlled by a lowcost thermostat and a space on the underside of the forex. The cold is then drawn from behind the eutectic plate only when it's needed and you can control temperature. We've done this several times.

[Chef]

In the chiller section, the temperatures decreases indeed to the same temperature as the coldest plate but as soon the compressor is shut off, the chiller plate rises very fast again to its eutectic point. So when charging during night, a small plate of PU is placed before the plate sometimes.

Now this is a very neat idea - just insulate the main face of the plate against direct contact with the goods and the additional cold past the eutectic temp is lost very quickly, probably in just recooling everything in the fridge/chiller down. It would be quite simple to have a little variable door of PU in front and just change this over a small trial period till it is correct.

I can also teach the Grande Fromage to close the sliding door when she has lettuce and open it if full of Castle Beer.

Not much surgery to the system - good news, controllable with small experimentation - also good news.

Thanks for this.

Chef

[Gary]

Now this is a very neat idea - just insulate the main face of the plate against direct contact with the goods and the additional cold past the eutectic temp is lost very quickly, probably in just recooling everything in the fridge/chiller down. It would be quite simple to have a little variable door of PU in front and just change this over a small trial period till it is correct.

Which brings us full circle, back to a spillover system.

As you can see, there are a variety of ways to do this... and all can be made to work.

[Gary]

Then a desuper was added to stop the discharge line melting stuff.

Tell us more about your desuperheater, location, water temps, etc. Possibly we can do something with this.

By your description, I am assuming this is a water cooled discharge desuperheater. It may be far more useful as a suction desuperheater.

[Chef]

Which brings us full circle, back to a spillover system.

Well not exactly - the original spillover did not have fridge or chiller plates installed and this option of a PU board in front of the plate is just to protect the goods in close proximity. This is based on putting the TXV in the chiller so it works like it is supposed to.

The desuperheater is a copper tube in tube (1" inner and 1 1/4" outer) unit mounted 2 meters from the compressor discharge and its output goes directly to the condenser which is a coil inside a PVC housing. About 15 turns in a spiral 4" in diameter. Pipe is 3/8" cupro nickel.
All water cooled and temp is 29C on a cold day and 31C on a warm day.
Then to reservoir, filter/dryer, sightglass, diaphragm valve and 4m of 3/8" pipe to the liquid - suction exchanger. Most of it is lagged to stop the liquid getting any hotter en route.

Chef

[Gary]

Well not exactly - the original spillover did not have fridge or chiller plates installed and this option of a PU board in front of the plate is just to protect the goods in close proximity. This is based on putting the TXV in the chiller so it works like it is supposed to.

The desuperheater is a copper tube in tube (1" inner and 1 1/4" outer) unit mounted 2 meters from the compressor discharge and its output goes directly to the condenser which is a coil inside a PVC housing. About 15 turns in a spiral 4" in diameter. Pipe is 3/8" cupro nickel.
All water cooled and temp is 29C on a cold day and 31C on a warm day.
Then to reservoir, filter/dryer, sightglass, diaphragm valve and 4m of 3/8" pipe to the liquid - suction exchanger. Most of it is lagged to stop the liquid getting any hotter en route.

Chef

The water entering the desuperheater is 29-31C?

[Chef]

Absolutely 29-31C - measured with a very accurate thermometer. The water flow rate is very high so it will not increase in temp by more than .1 or .2C through the condenser and then the desuper.

Chef

[Gary]

Absolutely 29-31C - measured with a very accurate thermometer. The water flow rate is very high so it will not increase in temp by more than .1 or .2C through the condenser and then the desuper.

Chef

Not much we can do with water that warm. It was just a thought.

[RLK]

Chef., My hundred plus engine driven holding plate system designs all using the SD 508 compressor have provided knowledge of what works and what does not. If your system is not in balance, performance will be poor and compressor’s service life will be at risk. It is true that compressor’s temperature will be too hot to touch during the first few minutes on first pull down of warm plates. If condenser’s size and its cooling medium are adequate high pressure at compressor will not exceed 135 psi using 134a refrigerant. As plate liquid drops one degree per Btu pound and reaches freeze point ice starts to form on coils lowering load on compressor. When holding plate’s evaporator coils reach a temperature 20 degrees below solution’s eutectic point plate is frozen solid and high pressure may drop as low as 95 psi. The important thing to remember is a holding plate is not a conventional evaporator and it is never at a stable thermo dynamic state, as ice forms on coils first reducing heat transfer. Twenty five years ago it was believed that plates in series did not affect daily compressor run times today most of these systems have more than one TXV.

Something you should think about is would you select a 25 ton condensing unit for 4 one ton air handlers in series, this is what you are doing by running this engine driven compressor at 1800 rpm. Just because you might rate a holding plats heat absorption when frozen at 4000 btu does not mean it will except that rate of heat transfer during a short freezing cycle. The 25 ton will remind you of the design error by frequent cycling and tripping temperature restart overload, your out of balance design will not protect compressor.

[Gary]

Perhaps a combined strategy would work best here.

Dual TXV's with insulative shield/spillover to protect product from freezing during pulldown, thus eliminating the need for the EPR. This should take care of the evaporator end of things.

At the compressor end, proper RPM can be established and steps taken to ensure compressor inlet superheat is not excessive.

[Chef]

RLK - thanks for the thoughts. I suppose the main reason it runs at 1800RPM is the genset is USA style ie 60Hz and so runs at that speed and the manual says to drive the compressor with a 5" pulley, one must assume the original installer followed the guidelines.

The manual also shows multiple boxes with serial piping of a single TXV or where the boxes are more apart it shows 2 TXV's with individual solenoids to allow either box to be used independently or together

We pretty much understand the eutectic solution and how it freezes and that's why we set 2 of the boxes with lower concentrations to control the freeze temps and so the box temps, this works fine.

Also the condenser seems to sized OK as it runs now but that may not be the case when the system is changed to one or two TXV's.

The thing we are trying to stop is the excessive Superheat whilst still controlling the box temps within reason.

Thanks for your input

Chef

[Chef]

Dual TXV's with insulative shield/spillover to protect product from freezing during pulldown, thus eliminating the need for the EPR. This should take care of the evaporator end of things.

This does seem to be the best choice at the moment.

But running one loop only at a time or both loops together seems to put a very wide operating envelope on the whole design. Trying to balance it with any one of 3 possible combinations seems tricky. Maybe its better to run both loops at all times?

I also get the impression a bigger receiver is going to be needed.

Chef

[Gary]

This does seem to be the best choice at the moment.

But running one loop only at a time or both loops together seems to put a very wide operating envelope on the whole design. Trying to balance it with any one of 3 possible combinations seems tricky. Maybe its better to run both loops at all times?

I also get the impression a bigger receiver is going to be needed.

Chef

I assume the 3 combinations you are referring to are:

1. Freezer alone
2. Freezer + fridge
3. Fridge alone

Would there be a circumstance where #3 would be needed?

[Peter_1]

Just read your comment on RLK's post about the high SH:
you normally select the orifice of your TEV for standard running conditions which will be in your case evaporating at -10°C to -15°C. But at startup, this orifice will be far to small so as you know, you will have high SH at startup and this for a long time, especially with plates linked together in serial.

Advantage of connecting the plates in serial is that you only sacrifice one area in one plate for creating SH.

What we have done already done to improve system performance is selecting the orifice for normal running conditions (in your case -15°C) and installing a suction, accumulator with the bulb of the TEV after the suction accumulator)

When we have more than 20 to 25 m length of copper in the plates, we also install a a TEV with external equalization.

[Chef]

Just read your comment on RLK's post about the high SH:


Advantage of connecting the plates in serial is that you only sacrifice one area in one plate for creating SH.

To save you the time of searching back through this post I will precise the situation.

The TXV is in the freezer and bulb is in the chiller so under normal operation the diaphragm on the TXV is controlling and maybe right at the last minute the bulb may take over but this is not known.
So the TXV is acting like an orifice (or a pressure regulated valve) and the last 2 plates then add too much superheat as the system has to be critically charged to work. This is the stumbling block really.

Is it normal to put the bulb after the accumulator?

I think the high superheat during pulldown from hot is acceptable but we have it each cycle even when cold and thats what is bad.

Chef

Chef

[Chef]

I assume the 3 combinations you are referring to are:

1. Freezer alone
2. Freezer + fridge
3. Fridge alone

Would there be a circumstance where #3 would be needed?

Maybe not but its human behavior if the option is there to use it someone probably will -I think I will just stick with option 2 and not have any problems of "well why can't we just run the fridges?"

[Peter_1]

......
Is it normal to put the bulb after the accumulator?
I think the high superheat during pulldown from hot is acceptable but we have it each cycle even when cold and thats what is bad.
...

OK, that's clear for me now.
Suction accumulator: we do it this way to have smallest allowable SH in the plates itself at startup and we need the SA only when system is on regime when orifice is then too big.

[Peter_1]

Why not simply placing your TEV - perhaps already proposed in earlier posts - in the chiller room?
And install the bulb of the TEV after the HX to increase system performance and reduce a little bit SH in the plates?
You know what you're doing, so I don't see any problem doing it this way.

[RLK]

Each engine driven holding plate refrigeration system is a one of a kind design. Chef, you referred to a manual who was the author? It sounds like the information for series plates and pulley size came from one of my earlier books on do it yourself boat refrigeration.

Small TXV orifices and low compressor rpm are important for compressor longevity. In my earlier post I recommended TXV orifices should be sized to math thermo conductivity of holding plate evaporator coil and not plate size this is important to reduce liquid fluid back on stat up. In normal operation after the first days run liquid refrigerant and oil migrates into cold plates. On the next compressor start TXV’s are at full orifice opening allowing liquid to be pushed and pulled back to compressor. Typically on cold start up accumulator will frost over and suction fitting on compressor will frost up for a short time. The risk of liquid reaching compressor is going to be greater because of your generator's fixed high rpm. There are hundreds of these systems in charter boats around the world. Standard instructions for operating engine driven refrigeration are to engage compressor at low engine rpm.

Balancing superheat is approached differently on engine driven holding plate systems because of the ever decreasing evaporator coil temperature. To cover plate’s evaporator coil's full temperature range the TXV selected must have a full temperature range of -40F to +40F. A freezer plate’s TXV is going to regulate SH for 80 degree F solution temperature down to -20 degree F, so at what point is the SH adjusted? Best to leave manufacturers valve setting alone and monitor suction pressure and SH until plates are frozen solid.

My own sailboat has three holding plates with each having its own TXV and two thermostats to activate two solenoids in liquid lines in order to control boxes and plates separately. A separate small 12 volt condensing unit also supplies refrigerant to extra coils in the same plates using two additional TXVs, if alternative energy is available.

[mad fridgie]

I think we need to step back from the tech areas for a short while, we need to look at the process.
1: do we need to ensure that pull down is achieved with 2.5hrs.
2: Which cabinet is used the most.
3: Does Chef, want/can make possible major changes. (what can be completed without destroying the cabinets)
4; What level of controllabilty is required.

From my point of view, the system is presently running short of refrigerant, causing high superheat, and then the SH is made even worse by the Suction/liquid line heat exchanger.

[mad fridgie]

If major changes to the cabinets is not an option,
I would remove the suction.lquid line heat exchanger.
Install suction accumlator
Increase reciever size
Install liquid sol valve (stop migration)
Slow compressor speed
Install insulating strips on fridge evap plates (reduce effect heat transfer area)
Leave TXV bulb where is.

[Chef]

Why not simply placing your TEV - perhaps already proposed in earlier posts - in the chiller room?
And install the bulb of the TEV after the HX to increase system performance and reduce a little bit SH in the plates?
You know what you're doing, so I don't see any problem doing it this way.

This was discussed earlier to move TXV to chiller but it meant subcooling the chiller down to -15C or thereabout so freezing the foodstuff, but your idea of the PU plate resolves this so it is a very nice option. Bulb after the HX is noted and an SA to protect the compressor. Seems the easiest and best option.

Thanks Peter

Chef

[Chef]

RLK - The author is probably C F Horton.

I understand your comment about reducing the reserve capacity of the TXV to help stop floodback and it will help in providing good control at the lowest flowrates.
As most of the liquid will reside in the freezer plates I would think that the higher temp fridge and chiller plates will reduce the liquid floodback when starting up from cold.

The TF2 is from -40F to +50F so that will be fine and I suppose I will just have to wait and see what SH i get the system is running again, but on the TF it is easy to try new SH settings as the control is on the outside.

So what boat do you have - a Jongert? With a shore assist on the plates it sounds as if you have a good redundancy there and if I could, I would fit one to mine but alas no chance now.

Whats your book called?

Chef

[Chef]

I think we need to step back from the tech areas for a short while, we need to look at the process.
1: do we need to ensure that pull down is achieved with 2.5hrs.
2: Which cabinet is used the most.
3: Does Chef, want/can make possible major changes. (what can be completed without destroying the cabinets)
4; What level of controllabilty is required.

From my point of view, the system is presently running short of refrigerant, causing high superheat, and then the SH is made even worse by the Suction/liquid line heat exchanger.

1 - Pulldown can be 3 to 4 hours if necessary but it means running out of spec at high SH for longer.
2 - Usually all 3 the same - full up thats why I think best to keep it as one series system and split usage. If I need just a freezer can always use one of the other 3 systems available, similarly for fridges.
3 - The TXV can be moved easily, the chiller piping can be changed and the HX can be removed. Bulb position can be changed and a SA is easily added.
4 - It runs on a manual clockwork timer - that will not be changed. Just need it to control it's own SH properly.

MF said
If major changes to the cabinets is not an option,
I would remove the suction.liquid line heat exchanger.
Install suction accumulator
Increase receiver size
Install liquid sol valve (stop migration)
Slow compressor speed
Install insulating strips on fridge evap plates (reduce effect heat transfer area)
Leave TXV bulb where is.

Apart from putting the TXV in the chiller this is pretty much where we are right now.

I am very much indebted to you, Gary, Peter and RLK for bashing through this putting wrong to right.

All I have to do now is source it and fit it

Thanks

Chef

[Gary]

Placing 4 coils in series is like placing 4 resisters in series. Parallel paths (dual TXV's) greatly reduces the resistance.

[Chef]

Gary - I know you are keen on the twin TXV setup and I see its merits and also its problems, it would be quite simple to add the spurs necessary to change over to 2 TXV's at a later date whilst the whole repiping is going on. I have the TXV in stock as a spare but not different orifices.

Chef

[Peter_1]

Placing 4 coils in series is like placing 4 resisters in series. Parallel paths (dual TXV's) greatly reduces the resistance.

Gary, these are very small plates with probably 4 m/12 ft max 1/2" coil inside each plate. So we have max 12 m/36Ft coil, so that much resistance will this not give. You even still can use a TEV with int. eq.

[Gary]

Gary, these are very small plates with probably 4 m/12 ft max 1/2" coil inside each plate. So we have max 12 m/36Ft coil, so that much resistance will this not give. You even still can use a TEV with int. eq.

I am not opposed to keeping the plates in series, just pointing out the downside.

The important points here are to:

1. Prevent the fridge products from freezing.
2. Reduce the compressor inlet superheat.

[RLK]

Chef, I do not understand the high super heat this compressor has been used with one TXV on systems with four plates in series and one system with two plates containing 37 gallons of eutectic solution. Two to three hours running the first day with warm boxes yes but there after one hour or less will freeze plates solid again. For plates to hold over 24 hours good insulation is needed and freezer plates needs one gallon of solution for each cubic ft of freezer and ½ gallon for each cubic ft of refrigerator. I believe that compressor suction and high pressure on second days one hour run along with frosted on lines profiles a holding plates system performance. Again a holding plate is not an evaporator it is a energy storage unit. I would like to see the second day pressures and frost areas recorded every 10 minutes for one hour. Then after a couple of hours at rest record parameters again with valve sense bulb hanging lose from line in box. When doing wide open orifice test monitor compressor and stop compressor if frost forms more than one inch down rear plate of compressor. Here is a two hour sample of readings from a balanced system although your reading will not be the same it is the profile changes that defines refrigerant flow. Typical High pressure every 10 min, 135,125,115,100,90,90,90,90.90,90.90,90. Low suction pressure,22,17,12,8,5.3,3,3,1,1,0.-1. Line areas frosted, TXV outlet line, plate connecting tubes and return line to compressor, 20 min line frosted between plates 1 and 2 and frost on return line to compressor, after 20 min no frost on compressor return line.

The SeaFrost people are well respected in the boat refrigeration industry. I bought a new Watkins 32 sailboat in 1983 and installed a used 1978 Sankyo SD508 compressor and it is still ticking today. I can’t post my web site here but Google my name and you will find my site.
Richard Kollmann.

[Mike NZ]

There have been several recent threads that have made me think about this setup in a new light.

The system has 3 boxes each with holdover plates, the refrigerant passes through each one in turn and running temps in each box are:-
Freezer -10 to -12C (2 plates at -20C)
Refrigerator 2 to 4C (plate -6C)
Salad Chiller 4 to 8C (plate -4C)

But the TXV is inside the freezer and the bulb is inside the Chiller so once the system's down to temps one must assume that the TXV will be sensing the diaphragm temperature and not the bulb's at some instant. Maybe as the system cools down further on cycle the bulb see temps lower than than the freezer?

However, it works, but is very susceptible to charge.

I've already got the message that it may be better to get the TXV say into the chiller but why does it work as it is? Any ideas

Chef

Oh - its R134 and running Dx = 125PSI and Sx = 2PSI

from the diagram it looks like the higher temp coils are operating as saturated and superheated extensions of the freezer coil, making it possible to control the freezer temp by sensing its "extension"

[mad fridgie]

Chef, I do not understand the high super heat this compressor has been used with one TXV on systems with four plates in series and one system with two plates containing 37 gallons of eutectic solution. Two to three hours running the first day with warm boxes yes but there after one hour or less will freeze plates solid again. For plates to hold over 24 hours good insulation is needed and freezer plates needs one gallon of solution for each cubic ft of freezer and ½ gallon for each cubic ft of refrigerator. I believe that compressor suction and high pressure on second days one hour run along with frosted on lines profiles a holding plates system performance. Again a holding plate is not an evaporator it is a energy storage unit. I would like to see the second day pressures and frost areas recorded every 10 minutes for one hour. Then after a couple of hours at rest record parameters again with valve sense bulb hanging lose from line in box. When doing wide open orifice test monitor compressor and stop compressor if frost forms more than one inch down rear plate of compressor. Here is a two hour sample of readings from a balanced system although your reading will not be the same it is the profile changes that defines refrigerant flow. Typical High pressure every 10 min, 135,125,115,100,90,90,90,90.90,90.90,90. Low suction pressure,22,17,12,8,5.3,3,3,1,1,0.-1. Line areas frosted, TXV outlet line, plate connecting tubes and return line to compressor, 20 min line frosted between plates 1 and 2 and frost on return line to compressor, after 20 min no frost on compressor return line.

The SeaFrost people are well respected in the boat refrigeration industry. I bought a new Watkins 32 sailboat in 1983 and installed a used 1978 Sankyo SD508 compressor and it is still ticking today. I can’t post my web site here but Google my name and you will find my site.
Richard Kollmann.

Hi Richard, i agree with many of your comments, but if you study earlier comments, you will find that the High Superheat is caused by the system being undercharged and by adding a suction/liquid heat exchanger.
If chef correctly charged, his product within the fridges "froze", he became an expert in balancing his load requirements at the expense of the compressor.
It would be interesting to see the load profile (pull down) as he as 3 different eutectic points.

[Chef]

RLK
The figures from warm are
Dx 150,140, 138, 135, 130, 127, 125, 120, 120 PSI
Sx 8, 8, 7, 6, 4, 3, 3, 3, 2.5 PSI
Temp 30, 28, 18, 9, 5, 2, -1, -4, -9 C

The temp is the freezer box temp 2/3 the way up, all the fridge and chiller plates are frozen and the HX is frosted right to the exit of the suction gas, suction line is weeping to compressor. Times are every 15 minutes.

Cycle data is Dx=130, 120 after 15 mins and stays there
Sx=3, 2 after 15 mins and stays there
run time is 50 mins to freeze all plates.

Mad Fridgie - Almost correct, the TXV is in the freezer and so has charge migration and the only way to get it operate was to critically charge it and allow it to work as if it had a fixed orifice, this is probably why it appears to be low on gas, ie the gas is correct as an orifice system but low if it was a TXV system.

Chef

[Gary]

Mad Fridgie - Almost correct, the TXV is in the freezer and so has charge migration and the only way to get it operate was to critically charge it and allow it to work as if it had a fixed orifice, this is probably why it appears to be low on gas, ie the gas is correct as an orifice system but low if it was a TXV system.

Almost correct. If the system were charged correctly as a fixed orifice, the compressor inlet superheat would not be excessive.

[mad fridgie]

Hi Chef, I do not believe you have a migration problem with TEV. It has not yet be proven that this is the case. All that has been proven is that there is more than enough SH to keep the valve open. If the valve is open then it is a fixed orifice.
What happens to your system if you add more refrigerant? (exclude freezing of food or has this always been your limiting factor?)
If your SST is -20C, even with a SH setting of 10C, you will still get an outlet temp of -10C.
TEVs do not have a limitless SH setting!
One would expect frost at the compressor. Good!
But if the liquid line is not isolated, then liquid will migrate into the evaps, on start up, a slug of liquid will return, this will cause the TEV to close. Giving the impression of TEV migration.

[Chef]

Hi Chef, I do not believe you have a migration problem with TEV. It has not yet be proven that this is the case.

The valve is at -12C and the bulb is at +4C in the chiller before a cycle startup. I think we all agreed a long way back that the diaphragm will have the charge, this is proof beyond a reasonable doubt for me.

Gary - because it is fixed orifice if we add a little bit more the suction freezes to the compressor and if we run it for a few minutes longer it gets really cold, as there is no control on the SH the liquid will just get closer and closer to the compressor till it floods. So it is about right the we operate at the moment - wrong in many respects though as we now know thats why we want to fix it properly.

Chef

[mad fridgie]

The valve is at -12C and the bulb is at +4C in the chiller before a cycle startup. I think we all agreed a long way back that the diaphragm will have the charge, this is proof beyond a reasonable doubt for me.

Gary - because it is fixed orifice if we add a little bit more the suction freezes to the compressor and if we run it for a few minutes longer it gets really cold, as there is no control on the SH the liquid will just get closer and closer to the compressor till it floods. So it is about right the we operate at the moment - wrong in many respects though as we now know thats why we want to fix it properly.

Chef

In a normal freezer room the TXV is in the room and so is bulb, when the system goes on defrost, TXV can remain at room temperture, the bulb can warm to a temperature above freezing. These do not migrate.
many systems fit the bulb after a suction/liquid Heat exchanger where the bulb is above the room temperture, these do not migrate.
I see that you are using a no 4 orifice, quite big (and to big), but would balance with your original compressor sizing (excluding changes in voulmetric efficiency) Remember that these valves DO NOT close 100%, (generally between 10-25% of max) so at some point there could be a likely hood that the load reduces below that of the valve minimum closing point. When all fluid is frozen and poor heat transfer is occuring.

No 4 orifice around 4 kw at -20C evap 8 Bar pd,

[Gary]

Given there is a HX after the bulb, if the bulb charge has not migrated, then the TXV should modulate closed and not allow frost back to the compressor.

On the other hand, if the suction is frosted back to the compressor, then the bulb (on the other side of that HX) should be cold enough to take back its charge... in which case it should modulate and prevent the frost at the compressor.

So... why does adding a small amount of charge cause frost back to the compressor?

[Gary]

Also, if adding a small amount of charge frosts the suction line after the HX, then the fridge/chiller plates are NOT being chilled by superheated vapor. All of the plates are fully activated.

[mad fridgie]

Given there is a HX after the bulb, if the bulb charge has not migrated, then the TXV should modulate closed and not allow frost back to the compressor.

On the other hand, if the suction is frosted back to the compressor, then the bulb (on the other side of that HX) should be cold enough to take back its charge... in which case it should modulate and prevent the frost at the compressor.

So... why does adding a small amount of charge cause frost back to the compressor?

I have had a look on the danfoss site does not give superheat control limits (adjustment) but does give varying detail to open 11C SH to open to design. This then shows that you would expect frost. Also valves do not close 100%, so present valve could be giving 1KW of load at maximum shut down, which likely to be greater than the plate load, thus liquid floodback may occur.
The present orifice matches chefs equivelent compressor duty (which practically is not as high as he predicted) I am not saying it has not migrated, but neither would i presume it was.

[Gary]

Here's the good news:

All 4 plates have been running fully activated without freezing the products in the fridge/chiller.

The TXV can simply be moved to the chiller box and the system should run about the same as it has been running. No insulated shield needed if the produxts have not been freezing.