PDA

View Full Version : Advise on a plate HE in a chiller







Peter_1
20-11-2006, 08:31 AM
I have a mind jerking problem with a machine and I need you thoughts about this.

We have a Rhoss chiller at a client of +/- 90 kW, consisting of 4 compressors.
The compressors are connected in tandem by 2, so 2 circuits of 2 compressors.
Theres a plate HE with 2 woven circuits in it, so each compressor has his own independent cooling circuit and the water-flow is split in the HE in 2 directions, one for each compressor.
On the main water outlet of the condenser is the flow-switch (FS) shutting down the 4 compressors.

The compressors are controlled by a digital step controller which steps down the compressors the more the temperature setpoint is reached. The DT and delay time between each step can be programmed freely. So 0.2K before reaching the setpoint, only 1 compressor will be running and as soon the DT between actual and set temperature increases to 0.8 K, all are running.

Suppose no the following scenario.We need a water temperature of 0C (32F) and the water is protected with antifreeze 6K lower. (-6C/21F)
One of the 2 water-circuits becomes clogged with dirt or debris for lets say 50%.
The FS will see a reduced flow of 25% and will still not trigger a shut down of the unit.
But in the partially blocked circuit, evaporating pressure will go down very quick due to the reduced mass flow.
The controller doesnt see this fast descending evaporating temperature because it controls the compressors in relation with the temperature setpoint.
So, the blocked circuit will evaporate for example at -20C(-4F), so far below the protected temperature level of the flowing glycol.
The evaporating pressure is then also still above the safety cut-out pressure.

I think that theres then a change that the blocked circuit will freeze up: its evaporating too low, the FS doesnt see the reduced flow and theres not enough glycol in the mix to protect it from freezing up.
Or will the flow of the other circuit prevent the blocked one from freezing up?

Whats your opinion?
I think that NoNickName can give also valuable information on this subject

NoNickName
20-11-2006, 10:09 AM
This is a very specific situation caused by a convergence of multiple causes.
The easiest way to protect the cooling circuit is to add glycol up to protect the minimum evap temp corresponding to the LP switch cut out.
If the circuit is likely to get clogged by debris, then the use of a PHE is to be avoided and a S&T evap is preferred, but if the chiller is existing nothing can be done is these regards.
To further protect the cooling circuit I would install a differential pressure switch together with the FS. Whether the DP is higher or lower than a predetermined range, the DPS will cut out.
I would also monitor the superheating: it is a good marker of heat exchange.
Finally I would also monitor the difference in evap pressure between two circuits of the same PHE or even across PHEs. When the difference increases over a threshold, that may be an indication of clogging.

Josip
20-11-2006, 10:29 AM
Hi, Peter_1


Suppose no the following scenario.We need a water temperature of 0C (32F) and the water is protected with antifreeze 6K lower. (-6C/21F)

What is your evaporating temp?

Best regards, Josip :)

Peter_1
20-11-2006, 12:42 PM
Thanks already NoNickName and Josip.

1st circuit is evaporating -11C, 2nd at -27C.
I'm sure it's clogged or the plates have been crushed in such a way that some channels are restricted where the flow is now via the way of the least resistance.

What do you think about the theory if one circuit becomes clogged that it then can freeze up?

NoNickName
20-11-2006, 01:21 PM
Yes, Peter, it's possible that it can freeze up, but only if the heat load of the plant is very little. Otherwise the returning temperature is positive and in that case it is unlikely the glycol-water will reach freezing point.
It is also possible that the PHE is already frozen and it blocks the water circulation on one side. Stop the compressor for 24 hours and see if the evap temp improves.

Josip
20-11-2006, 02:54 PM
Hi, Peter_1 :)


1st circuit is evaporating -11C, 2nd at -27C.

...and your glycol is antifreeze protected down to -6C (is that maybe 6 K lower of -11C what is then -17C :confused:

or, maybe I get all thing wrong???

For me glycol must be protected (5-10K) below evaporating temperature.


It is also possible that the PHE is already frozen and it blocks the water circulation on one side.

This is possible because you can have micro crystallization just on wall of plate what reduces heat exchange and then reducing the flow of glycol...

We had similar problem with direct cooling of young beer (ammonia/beer) in S&T HE. Was not possible to use PHE due to cleaning (CIP).

Peter_1 can you post a scheme to see that more clear;)

Best regards, Josip:)

Andy
20-11-2006, 10:38 PM
Hi Peter:)

My bet is on a damaged plate due to an earlier problem:(

We are a Rhoss agent so if you need ny manuals ect I should be able to get them:)

Is it an energy 400 controller or one of the later carel PCO.
From memory the energy controller is LP protected, usually set to cycle three times on LP switch then lock out the controller on an E05 fault

Damaged plate is my best guess:)

I'm sure you have re-weighed the charge in and checked the TEV valve and drier operation

Kind Regards Andy:)

Peter_1
20-11-2006, 11:32 PM
Andy, It's indeed an Energy 400 controller(=Eliwell)

The unit is indeed LP via a mechanical cut-out. It was set at a pressure which cut-out the circuit just above atmospheric pressure and the DP was the smallest possible, so swithcing off at -30C and swicthing on again something around -27C.
We re-programmed the cycles now to 6/hour (it was originally set on 0) so taht they can keep using the machine.

I see this set-up more as a design fault: some dirt enters a circuit, blocks it partially, waterflow goes therefore even more to the other circuit (FS doesn't detect the fault) and the evaporating pressure in the blocked circuit goes down because the Energy 400 switches on all the compressor. Temperature goes down well below the freezing point.

NoNicKName has a point if he says that the protecting temperature must be determined via a DP which is calculated via the cut-out pressure/temperature of the LP cut-out switch.

Or an analog LP signal to the energy 400 which switches off compressors as soon pressure drops under a certain safety level so that pressure can't go that low.

In the past we had at normal conditions entering at -3C and leaving at -8C , so bot under the freezing point and protected at -15C.


Josip, I was just giving an example to explain my theory. I measured today the protection level and it was -12C.
I'm with you that you must protect 5K below evaporating temeprature but in the scenario I described - which is perhaps wrong - it can go lower then intended. Like we have now:-27c for -5C glycol.

Rhoss asked me today if the used glycol was mono-ethylene glycol or propylene-glycol? Does that make any difference?

In this whole unit is not one valve in the refrigeration circuit. Very easy if you have to work on it :(

Andy
20-11-2006, 11:40 PM
Hi Peter:)

cool pack gives mono ethylene 35% safe to -15 deg c and freezing at -20 deg c

mono propylene 35% safe to -12.5 deg c and freezing at -17 deg c

On the valves, as all Italian chillers valves are a small extra price on list:D

Kind Regards Andy:)

US Iceman
20-11-2006, 11:48 PM
Rhoss asked me today if the used glycol was mono-ethylene glycol or propylene-glycol? Does that make any difference?


Peter, the biggest difference would be between the thermal conductivity and viscosity of the two fluids are far as heat transfer is concerned.

From memory, propylene glycol (PG) has a higher viscosity than ethylene glycol (EG) at the same concentration and temperature, which can reduce the heat transfer coefficient. I also seem to remember PG has a lower thermal conductivity, which also reduces the film coefficient.

That might be why they are asking you that question.

TXiceman
21-11-2006, 03:35 AM
PG and EG have fairly different thermal properties as does the concentration effect properties. In order to properly evaluate your HX, you will need to provide the supplier, the type glycol as well as the concentration. If you can not provide the weight concentration, you will need to get an accurate temperature reading and a specific gravity with a lad calibrated hydrometer. If not hydrometer, get an accurately measures volume sample and weigh it. Based on the temperature and weight per unit volume you can get the density and a concentration.

Ken

Peter_1
22-11-2006, 07:59 AM
Two new separte PHE ordered at AIA Sweden and they will be replaced Saturday.

Andy
22-11-2006, 07:03 PM
Two new separte PHE ordered at AIA Sweden and they will be replaced Saturday.


Hi Peter:)

AIA and PHE, now thats something I never thought of, never knew they did plates:o

Kind Regards Andy:)

TXiceman
23-11-2006, 03:04 AM
We have an application that we are doing the service and maintenance. It is cooling a gasoline stream with a Flat Plate brazed plate HX. For the second time in about 2 years the HX plugged up and stopped flow. The operators did not catch the problem in time and keep trying to restart the compressor (tripping out on low pressure) and managed to wreck the compressor.

When I went to the site to check with the tech, first thing I found was that the gasoline stream did not have a strainer to protect the BPHX. Strainers on the gasoline processing skid required regular cleaning due to contaminates. Had them install dual 40 mesh strainers and block valves on the unit. The sister unit will need to come down soon and have strainers added.

Be sure ot keep at least a 20 mesh strainer in front of the BPHX. Thes things are great strainers by themselves.

Ken