Evaporator Flash over

[Chef]

New problem? We are using a 2KW compressor feeding 4 holdover plates controlled by a JT valve.
Sometimes the liquid seems to bypass the 1st plate and sometimes he 2nd plate and start to evap in the 3rd and 4th. This causes the last 2 plates to freeze but 1 & 2 are at liquid temp? The compressor is liable to get slugged so we stop the system. Later it works OK and then later again flashes past a couple of plates. Dx is 125psi Sx is 3psi and SH is 14F. Any ideas?

[Brian_UK]

HI and welcome to the forum.

It sounds as though the control valve (JT?) is not controlling properly.

The pressures quoted do not help a great deal as you forgot to tell us which refrigerant it is running ;)

[Chef]

Thanks for the welcome: I am pleased to be here.

The gas is 134a and we have tested the JT valve. A Danfoss T2 unit. We put the bulb in a freezer and then put 150psi air on the inlet and monitored the pressure on a closed off outlet. As different settings are set on the adjusting screw so the oulet pressure varies. Readings from 2 to 20C SH were obtained so this appears to be working fine and gives us a nice cal point for setting the SH.

The system is dry - the sight glass is bright green.

Each eutectic plate has about 5 metres of coiled 5/8 copper pipe and then 1 - 2 metres of connecting pipe so the whole 4 plates have some 25 metres of pipe. To measure the temperatue at the valve outlet and 3/4 down the evaporator we put 2 pressure tappings and get (at the mid cool down point) about 10 psi across plates 1 to 3 and 6 psi across plate 4 and the suction/liquid line exchanger. Is this a normal pressure drop?

It certainly makes calculating the SH more complicated. As we cool down the flow rate slows and so the pressure drop falls and SH changes!

So back to problem why does the JT valve loose control and some other part of the evaporator become the effective restiction device.

Bestest - Chef

[Electrocoolman]

Hi Chef,
The original description of the valve as a 'JT' probably threw us...however Danfoss T2 tells us its a TEV (Thermostatic Expansion Valve).
Was there another letter on the label? TN2 or TEN2?

The control phial needs to be mounted correctly to the pipes...securely clamped to a horizontal pipe and Insulated for it to control properly.

See http://ra.danfoss.com/TechnicalInfo/literature/manuals/01/PF000G102_chapter_01.pdf

Don't forget that any adjustment on the valve should be done in small steps and then wait to see what the result is of the adjustment (ie for system to stabilise)

[Chef]

Thanks Electrocoolman - this url is exceptionaly usefull, especially as it mentions evaporator hunting. This is the Danfoss tech term for what we must be calling flash over. There is much to consume in this paper and we shall try the recomendations tomorrow and see if it helps and check out the installation is correct.

Sorry for the JT bit but we always used to call the valve a Joule Thompsom Effect valve after the guy who discovered that expanding a gas (liquid?) produced a drop in temperature. It is (on further thought) not really correct as almost all of the cooling is done in the evaporation phase way past the valve.

So, our TEV is a TN2

Later - Chef

[Electrocoolman]

TN2 valve is for 134a....just wanted confirmation gas and valve matched.!

It is however only an internally equalised valve. I would have thought that an Externally equalised valve would have been fitted....with a pressure pipe from the far end of the evaporaters back to the valve. This improves the control of the valve.
i.e. TEN2

[Chef]

We understand the idea behind a TEN2 but being at sea and away from a fridge parts shop poses some problems and so we must try to fix what we have for the time being. If the pressure drop in our evaporator causes the problem will a TEN fix that?

[NH3LVR]

Chef;
Welcome to the forum!
This is a bit out of my area, however I won't let that stop me:).

I have worked on a couple of these systems and have built one.
They do operate with a good bit of pressure drop, especially when warm.
I am assuming that this worked properly at one time. It is definitely a situation calling for a externally equalized valve. If it did work not work properly that could well be your answer.
Does it actually flood back to the compressor? Only at the end of the cycle?
These type of units have a VERY long pulldown. 12 to 14 Hours from warm is not unusual.

Dole has a little technical information in their catalog.
See http://www.doleref.com/Assets/Sales%20Catalogs/Catalog%20CE-%20Holdover%20Truck%20Plates.pdf and http://www.doleref.com/Assets/Sales%20Catalogs/Kold-Hold%20Catalog.pdf

[Chef]

Hi All
Thanks to the Danfoss manual from Electrocoolman and the NH3LVR url about truck holding plates we may have found a work around for the moment.

The truck cool company mention that a large resevoir will allow the refrigeratant to collect in the plates and then if this is too large may slug the compressor, so they recomend a minimun charge and low reservoir size. We are using a car aircon compressor so get some fairly high flow rates and cooldown takes about 1 hour.

So we tested the temp of the plates and find -10C in the freezer and -4C and 0C in the other 2 compartments (using those neat little inside/outside thermometers for houses!!). So the liquid gathers in the freezer plates - as we start up this flows and then flashes off in the 2 warmer plates and so cools them quickly but the liquid seems to just flow through the plates 1&2. Now as we have a TN2 the pressure drop in the last 2 plates is sufficient to keep the fluid flow going and the valve stays open. These plates just seem to act as a subcooler.

Eventually (20mins) the frost line starts towards the compressor and the 1st plate starts to warm up caused by the liquid temp being about 15C and we soon have to shut down to stop compressor damage.

So the system is very critical to amount of charge and reducing the charge by 10% and fitting a small reservoir stops this flash over at the moment.

We dont know if we run at lower temps, say -15C, it will start again or stay stable. But we do know that it is not working at its max potental due to lower charge.

This is just one example where BFI (brute force and ignorance) just does not cut it.

Thanks guys
Chef