Trouble with a milk vat - what on earth is this valve?

Images: http://imgur.com/a/oMSU0

A customer of ours bought this "package" of a second-hand milk vat and new condensing unit (labelled R404a but scribbled out with a pen and R134a hand-written over the top) to be used to chill water. It was supposed to be an all inlcusive kit, so we cleaned it up and assembled it probably 12'ish months ago using nothing but what they gave us. It has mostly worked over autumn/winter/spring, but now that summer is here (southern hemisphere :D) it is very inefficient.

They are using 316L/hour of water that enters at +31°C that they would like to leave at +2°C which according to my heatload calculating (in)ability comes in at around 26,000 Btu/hour.

The condensing unit fitted is rated at 31,110 Btu/hour (desired conditions, actually closer to 45,000 Btu/hour in the current operating conditions) which means it should have sufficient capacity.

The issue I have is very low TD between SST and fluid temperature (read: 3°C) and very low superheat (read: 1.5°C system superheat). It makes me think the expansion devices are not providing sufficient pressure drop - but there seems to be no adjustment? The bolt on the bottom of the valve is seized absolutely solid.

Anyone with milk vat experience able to provide any thoughts on what the balls that valve is, and what sort of operating conditions I should expect to find?

Hi Red.
looks like a berkit ( spell check )re-injection valve. Re-in trains liquid over flow from dimple plates, back into circuit, basically a jet flow induction valve arrangement. A fine art in setting them up, sight glass will always be bubbling from memory. Get it wrong and will flood back. They worked ok on R12, do not know about 134a
Haven't seen one for years. SST/ versus suction super heat, at design conditions is best method of checking gas charge.

magoo

Tell 'em they're dreaming.

I'd be surprised if that thing could pull 316L water from +31°C to +2°C in 6 hours.

Think about typical milk vat usage - full it up with the good stuff, & it gets brought down to temperature during the day, then the tanker picks it up in the evening... Not 316 litres per hour..

Quote:

---

Originally Posted by Magoo

Hi Red.
looks like a berkit ( spell check )re-injection valve. Re-in trains liquid over flow from dimple plates, back into circuit, basically a jet flow induction valve arrangement. A fine art in setting them up, sight glass will always be bubbling from memory. Get it wrong and will flood back. They worked ok on R12, do not know about 134a
Haven't seen one for years. SST/ versus suction super heat, at design conditions is best method of checking gas charge.

magoo

---

Thanks for the reply. Is 'Berkit' the right spelling, or is that phonetic? I can't seem to find any information on that besides older posts by you on RE.com :P Tried a few spelling variations but nothing obvious besides an African wiring manufacturer.

Does this work as a fixed orifice expansion device? The trouble we're having could be exactly explained by overcharging (or an inefficient compressor) not providing sufficient flow rate to create the pressure drop needed. Maybe I need to reclaim the gas and start from scratch :-\

Quote:

---

Originally Posted by HVACRsaurus

Tell 'em they're dreaming.

I'd be surprised if that thing could pull 316L water from +31°C to +2°C in 6 hours.

Think about typical milk vat usage - full it up with the good stuff, & it gets brought down to temperature during the day, then the tanker picks it up in the evening... Not 316 litres per hour..

---

I've spent a few hours looking into milk vats so I'm totally an expert now :P ... but milk vats should bring the milk from "cow temp" of ~37C to < 7C within 3 hours of the milking starting, then to < 4C within 3 hours of the final milking being completed. The blended temp of the incoming+stored product should be < 10C the whole time.

See? Totally an expert. And also completely irrelevant knowledge to the problem at hand :(

Not familiar with those valves. There are some fixed orifice tanks here, were state of the art 30+ years ago, they have no receivers but have a suction line accumulator and are critically charged.

Is it feasible to convert your system to TEV's? Can you get any info from the manufacturers on how its meant to work?

Quote:

---

Originally Posted by monkey spanners

Not familiar with those valves. There are some fixed orifice tanks here, were state of the art 30+ years ago, they have no receivers but have a suction line accumulator and are critically charged.

Is it feasible to convert your system to TEV's? Can you get any info from the manufacturers on how its meant to work?

---

The manufacturer hasn't existed in many years, unfortunately. Plenty of their equipment still around for sale at second hand dealers but it's all the same sort of vintage.

I'd love to convert to a TX valve, but without knowing the capacity of the milk vat it means sizing based on the condensing unit capacity - which I guess will probably be fine. The other trick will be physically getting the valve on there... it's stainless everywhere. That sucks to weld :P That's our fallback position at the moment.

Also, the condensing unit has a receiver but no accumulator which seems a bit arse-backwards. I suspect it's meant to have no receiver, a big accumulator and be critically charged like your fixed orifice stuff. The company who sold the package has disappeared to so I can't even get an idea of why they felt they were a good match.

Just another second hand equipment mess.

Hi Red,
Burkert Fluid Control Systems , may be a starter. www.burkert.com I thing they are a German company.

Quote:

---

Originally Posted by The Red Krawler

The issue I have is very low TD between SST and fluid temperature (read: 3°C) and very low superheat (read: 1.5°C system superheat).

---

I would slowly remove refrigerant until the compressor inlet superheat is at least 5-8K. Don't be surprised if the suction line temperature drops before it rises as you remove refrigerant. That's what happens when a fixed orifice system is grossly overcharged and is pumping liquid.

I wouldn't worry about welding to stainless. You can braze quite easily with the right silver solder flux rods. With no need for bottled flux and the mess that comes with it, transition joints become simple. If you find the right TXVs, I recommend using the following silver solder.

http://www.silfos.com/products/catal...-p-1-c-52.html

Hi Red.
very good advise from Gary. These systems were designed to cope with a dairy farmer starting vat empty before milking started, that is how from memory we set them up, an empty vat. Check refrigerant charge that way and suction super heat differential at around 10' C, minimum, ignore what the liquid site glass is doing.
What size vat is it, the level gauge on the side will have a full volume figure stamped at the top. The bigger the better, your process load at 316 litres per hour is nominal 10.6 kw.

Quote:

---

Originally Posted by Gary

I would slowly remove refrigerant until the compressor inlet superheat is at least 5-8K. Don't be surprised if the suction line temperature drops before it rises as you remove refrigerant. That's what happens when a fixed orifice system is grossly overcharged and is pumping liquid.

---

I've organised for it to be emptied next week so we can start with an empty vat. I'm going to start with 5KG of refrigerant and see what we get. I'm also going to check the frost pattern at the same time and make sure we've got good contact between evaporator and stainless skin.

I'll update the thread with the results in case it helps someone else in the future

The following data would provide a much more accurate picture:

Evaporator air/water in temperature = ?
Evaporator air/water out temperature = ?
Low side saturation temperature = ?
Suction line temperature = ?

Condenser air/water in temperature = ?
Condenser air/water out temperature = ?
High side saturation temperature = ?
Liquid line temperature = ?

A quick update...

After getting a hold of a bloke in another state who has worked on plenty of these, we got some great information. Specifically, that the vertical "pipe" as seen in http://imgur.com/a/oMSU0#2 which I'd assumed was just a piece of framework is actually a pipe used for oil return. A some-what common fault in these fixed orifice designs is that the orifice wears to the point where the "jet" function of the valve no longer works and instead the liquid just bypasses the valve and ends up travelling down the oil return line... ie: you perform no work because a large portion of the liquid is entering the evaporator almost right at the end! This has the rather obvious downside of risking liquid floodback too.

Our solution was the chop the valve off completely using my friend the hacksaw. The rusted mild steel pipework has been cut off flush with the vertical pipe, and then a piece of 5/8" copper slipped up through the middle to make the connection to the ~7/8" stainless evaporator inlet. The copper was then blue tipped to the stainless, and the gap around the copper on the vertical ~1 3/4" pipe was bluetipped as well.

Then we fitted a pair of externally equalised TX valves sized at 50% of the condensing unit capacity with the bulbs strapped to the suction line at the same point. There's a risk the valves may fight each other, but thus far we achieved +4K superheat with only minor fluctuations.

We also added an oil seperator and a fan cycle control to keep the head pressure around 45C and ensure plenty of velocity through the evaporator to return any oil missed by the seperator.

Total system charge to a clear sight glass was 11.5KG which is probably more than it needs but it has a huge receiver and condenser so the more the merrier I say.

The system was a rocketship from the +27C the water temp was at when we finished down to around +25C by the time I'd finished paperwork and packing up etc (not bad for something like 1500L of water, with the customer using 316L/hour on top of that!) but no idea what the minimum temp will be.

The big negative of all this is that I'm pretty sure we've lost a valve on the compressor. It pumps down to a vaccum and holds off just fine, but the pumping down process is quite slow which suggests it's not pumping effectively. There may be a compressor change yet which will be a shame, but after a full production run in the middle of summer the water temp was +11C the other day so they may just run with it until another valve goes heh

Hi Red
thanks for up-date, sounds like a good result finally.

Hi again Red.
thinking more about the changes you have made, all good. But once the system comes closer to design, without the re-gen thingo orifice the whole thing could go very pear shaped and flood back. If you have not already installed a large volume suction accumulator on each circuit recommend you do so. The dimple plates on vat are not the best for circuit flow rates and flood over real easy and rely on the re-circulation factor. Can go from normal to disaster in a flash, in effect you have removed a factor of the system regeneration cycle so will end up with a seriously full evaporator coil full of liquid. I do not believe an externally equalized TXV will react quick enough. If it does react, it will be too late anyway. Upping the SDT will aggravate the whole situation.
Let us all know how you get on..,
magoo