View Full Version : supermarket refrigeration

09-01-2016, 12:14 AM
hi all , i found my self with a problem the other day , regards to a few cases on a lt pack struggling with temp, first of all i thought the akv is not opening correct , on two cases, then i tried to pump down the cases , but found gas still pissing out of vaive , even though nothing in suction , after so much messing around i decided to shut suction put case on defrost , to see if there was a blockage, no way liquid shut off was passing on two cases , then after a while suction flew up to 120 psi , and i opened it up and a way the gas went , , so i thought oil is stuck in cases, does anyone no best way to get oil out of the cases , as this took me so long , didnt want to go through all that messing around again , did think about turning pack off for a while , to increase, back pressure , but it was xmass, and tradding was important to the store , and to be fair didnt want to cause more issues at this point ???

Glenn Moore
12-01-2016, 01:12 PM
Hi Billy
What gas is in the system and whose controller is driving the Akv . Normally on LT cases putting the case on defrost a couple of times to warm the oil up a bit helps to remove it from the case. Also if the Akv is set up correctly to run with a low superheat that should also wash the oil out of the cooler

14-01-2016, 02:41 PM
hi mate . it is 404a, akv superheat is petty low at around 4 diff ish , its hard to say it goes up and down a bit average around 4 , its on a pack they do flood a bit I noted back to main suction , sure I cleared it mate , but what a nightmare, I took armature out , and I pushed a little needle through the orifice while still in the valve , and gas burst out several times , but nothing on my gauge at suction outlet of evap , strange , like blockage, thanks for repy.so evap around -27. suction -23, when working ,

14-01-2016, 02:43 PM
to add , its a microcool , Tesco,

Glenn Moore
18-01-2016, 02:28 PM
Hi Billy
The Microcool controller was made for Woodley for use in Tesco stores. That controller uses 2 PT1K sensors for the superheat control. More often than not the S1sensor was fitted in the wrong position on the evaporator. This caused poor control and low flows through the evaporator, this in turn cause oil to be entrained in the evaporator especially on LT packs.
To check this do the following . Check the temperature of the S1 sensor with the handset or on the Front end panel screen .This temperature should be the same as the evaporating temperature on the suction gauge on the LT pack. If the temperatures are more than 1 or 2 degrees different you need to move the S1 sensor to a different place on the coil. Also check the AKV opening % as this should be between 40-75% if this is lower than 10% then you have a problem.
Let me know what you find and I'll advise accordingly Glenn

21-01-2016, 01:12 AM
ok thanks , first thing I did was to check evap probe , it was ok but I did position it straight after the akv, before the distribution , do find it always to be the best place , as long as you strap suction probe secure too, I have not been back to the case as yet , its been running ok for a few weeks , yes agree with the akv opening , must admit I did not check this with hand held, thanks , I was told theses controllers , have a memory, and maybe this comes form the Einstein, iam not sure , so I do reset them with turning them off and on, I think if it doesn't get down to temp it trys , to do what it did when it was to temp , iam not sure what truth is in this , do you know ???

21-01-2016, 01:24 AM
to add , akv opening set point did you mean? ,

21-01-2016, 07:10 AM
An interesting read Billy7.
Something to keep Glenn occupied as well!:o

Thanks for posting.

Glenn Moore
21-01-2016, 06:10 PM
Hi Billy
The Microcool uses 2 sensors for the superheat control, called S1 and S2. The S1 sensor reads the evaporating temperature of the evaporator whereas the S2 sensor reads the actual suction pipe temperature. The difference between the sensors is the evaporator superheat on which the AKV is controlled.
The S2 sensor is placed on the suction pipe where you would normally fit the expansion valve bulb, to simply read the pipe temperature. The S1 sensor must be placed somewhere towards the front end of the evaporator at a position that gives the actual evaporating temperature.
This temperature for the S1 sensor is what you would see indicated on the evaporator guage .
With the system running you could use a digital thermometer to work your way through the evaporator circuit starting from the AKV outlet. The evaporating temperature is the lowest temperature you can find in the evaporator circuit. ie the point where the pressure is lowest will be where the lowest(evaporator) temperature is on the coil, and this is where the S1 sensor MUST be placed.

A little bit of theory to help find the true evaporating temperature point.
Lets start at the liquid line on a typical R404A system Condensing at 35Deg C =15Bar pressure . So the liquid enters the AKV expansion valve at 15 Bar pressure, it exits the AKV at -20 Deg C / 2 Bar pressure. It then enters the Distributor and the distributor tubes where there is a further pressure drop typically approx. 1.5Bar. So the liquid/gas mixture at 2 bar pressure enters the distributor and the distributor tubes and exits the end of the distributor tubes inside the main evaporator coil tubes at 0.5Bar / -35 Deg C . From then on there is a slight pressure drop of 0.1 bar through the evaporator coil block.
So you can see placing the S1 sensor after the AKV would tell the Microcool controller that the evaporator was evaporating at minus 20 Deg C and it would then reduce the opening degree of the AKV so that the S2 sensor would only have to read minus 15 Deg C so that it thinks its controlling the superheat at 5K ie -20 from -15 = 5K but in actual fact the evaporator is not running with 5K of superheat but actually 20K of superheat ie evaporating temperature at 0.5Bar = minus 35Deg C and suction pipe temperature is minus 15 Degree C ie -35 from -15 = 20K of superheat. Looking at the superheat on the Front end screen or the handset would say that the evaporator was running at 5 K superheat whereas its actually running at 20K superheat. This would cause the case to struggle to get down to temperature as half the evaporator would be a pure superheater , plus the evaporator gas flow would be extremely low causing oil to drop out of the gas and become entrained in the coils.
By fitting the S1 sensor to the correct spot on the evaporator would cause the Microcool controller to see that the superheat was 20K ,this would then force the AKV valve to most likely fully open to 100% to reduce the superheat down to a correct superheat of 5K . The duty of the coil would increase dramatically making the case more efficient with less frost on the coil, a total win win situation.
So the best place to fit the S1 SENSOR is where the distributor tube enters the main coil block tube, go along that coil tube approximately 300mm and at that point carefully remove the aluminium fins of the main tube. Then fit the S1 sensor with a couple of good quality tie wraps and then cover with insulation . then sit back and watch the difference. I have had to do this mod on many cabinets and also on potatoe storage systems that use 2 sensors for superheat control irrespective of what make of controller that is driving the AKV valves
The AKV opening degree in % is a measurement only parameter not a set point parameter, unless your in MANUAL control
Always read the S1 temperature via the controller handset or front end panel . The S1 temperature MUST be within a couple of degrees of the temperature as read on evaporator guage. I can guarantee that a large percentage of systems running with 2 sensors for their superheat control are running with reduced system capacity, poor system cooling effect and excessive running times wasting a fortune in energy , so try it and see the difference

23-01-2016, 03:33 PM
wow ok I do understand , yes I will try this and then move forward with the way I position the probes , thanks so much , as a service engineer , I tend to go out in the field , and just do the best I can do in the time that is given to me to repair, several different type of problems per day. if I asked the managers at work for input , they just would not know , so I find myself a bit alone , in finding right solution , and my mind , is always thinking I wonder how that works , don't get my wrong I can do the job as far as I will get it running one way or another , if its a pack or a smaller unit , I can make it do something ,and normally will not ask for assistance, but yes iam impressed , as to the information you do have , regards , bill, which of course is just an alias, lol

23-01-2016, 03:40 PM
please note glenn, supermarkets have a working head pressure around 10 to 12 bar , the lt packs run pretty low head , most are around the 150 psi mark. , suction anything from 15 to 20 psi,

Glenn Moore
23-01-2016, 06:03 PM
Hi Billy
It makes no difference what the actual conditions are but the distributor in any system should give a pressure drop of between 1 to 1.5 Bar, but the important thing with using 2 sensors for superheat is that the S1 sensor reads the same temperature as your suction gauge and the best place is as I described.
I had the same problem with some cases using RDM controllers and AKV's not long ago. The sensors were pre fitted in the cases, totally in the wrong position, but it seems few people understand , and as long as they get some sort of cold air its good enough in their mind, but I prefer to see things installed and set up correctly and not "It WILL DO " attitude. As a poorly installed system will cause all sorts of problems to the system and compressor in the short term with continuous time wasting site visits. You sound like the type of engineer I prefer to see on my training courses. Kr Glenn

23-01-2016, 06:12 PM
AKV's and probe placement, since the first ones came out where to put the s1 is critical. It seems to have improved since Dixell and Danfoss moved to a pressure transducer. It was no wonder we used to get so much time to commission super markets, most days were spent stripping cabinets and shifting probes. Nice lesson Glenn!

23-01-2016, 06:42 PM
Haha agree about shifting probes . Damn bad though when manufacturers dont know the ideal positions for probes in their own cases , its a proper nuisance stripping out cases that were ordered with fully fitted controllers

Glenn Moore
23-01-2016, 09:14 PM
Hi Al /joe
The first generation of Electronic Expansion valves were all controlled using 2 sensors, this was mainly to do with cost, as pressure transmitters are quite expensive and can kill the price of a job comparing Thermostatic TEV;s as against an Electronic EEV. So 2 sensors were the cheaper option to get the new technology into the market.
Using 1 sensor and 1 pressure transmitter makes the EEV more expensive but removes the risk of the valve not functioning correctly. As I told Billy, the S1 sensor must read the same temperature as the suction gauge on the system. This is exactly how using the pressure transmitter works. With the latest EEV controllers that use a sensor and a pressure transmitter, the S2 sensor simply reads the suction pipe temperature, the pressure transmitter reads the suction pipe pressure. This pressure is converted to a temperature as you would find on your refrigerant gauge or on a comparator. This works fantastic providing you set the correct refrigerant in the controller parameter. Having the facility to change the controller refrigerant setting to whatever gas you prefer to use makes using EEV's very versatile, ie 1 control works with virtually any refrigerant type now and in the future, simply by' on Danfoss controllers setting the refrigerant to number 13 (USER DEFINED) and entering new data in the Danfoss only under the A1 ,A2, A3, parameter .

I've run training courses on Expansion valves and Electronic expansion valves, teaching all the do's and dont's , getting the sensor positions right and how to get the full benefits from them for 37 years at Danfoss , and in the early days spent nights sitting watching evaporators doing all sorts of weird things, because with the electronic valves you could get so much information from its controller ,whereas with a TEV you only knew there was a problem and you only had a gauge and thermometer to find what was wrong and freezing your butt off in the process.
I did a lot of on site R&D on expansion valves and evaporators and their behaviour and discovered some strange phenomena in multi evaporator systems such as "SLEEPING EVAPORATORS" and effects of large pressure drops in evaporator distributors. An invaluable amount of knowledge was gained during this time which I always try to pass on. enjoy

Glenn Moore
23-01-2016, 09:18 PM
Thanks al
all the best

23-01-2016, 09:22 PM
That's some interesting info Glenn about the placing of the probe.
Seem to recall the Danfoss paperwork stated to place the S1 probe on the first return bend but agreed, your method is more accurate.
Unfortunately I've not dabbled on a system with an AKV for a few years now. Seen a few not working because they had the wrong coil fitted, not the DC coil. Thought a coil was just a coil but obviously not. Anyone cast any light onto why?

To add about the memory from the front end I've heard a couple of versions so not sure about the truth, it's out there (a job for Mildew and Scally!), but the rough version as I believe is that you can fart about with the program at the controller but the Einstein will refresh the program on all controllers on the network with its version at a set time. Not sure if it wipes all parameters or just certain ones. Think if you change the relevant parameter at the Einstein then that secures it.


23-01-2016, 10:13 PM
I think glenn is bang on with his positioning of the s1 probe , ive had to move a lot of probes from the first return bend with some controllers as on start up with a long coil the valves just shut down, I assumed the valves were not seeing enough of a temperature drop / difference between the two probes fast enough and so shut but anyways moving the s1 closer to the first place to freeze always cured it. The return bend method might be okay on shorter coils.

23-01-2016, 10:50 PM
thanks for the information glenn. I work a lot with danfoss controllers, and getting the correct info is liquid gold , I worked on a 550 the other day on a co2 site. they don't have evap s1 probe its a suction transducer in that line of pipe work say one transducer , per ten cases . I noticed out of the ten controller one was reading a much higher temp , I didn't get this due to the fact how can you get one reading different , one transducer one reading. one of the first things I look at is that evap temperature , well on multiple evap. and the superheat , see if theres enough gas going through the coil, but all read minus 31, yet one lone evap read minus 16, odd. the case was struggling , I have ordered new controller but was not convinced it was the controller as yet , I did pull the plug on the transducer , as I have had a few that just lock up before , any ideas, front end is a 225 danfoss, which to be fair I don't rate well .

Glenn Moore
23-01-2016, 11:20 PM
Hi joe
Yeah its to do with the start up procedure. The controller opens the AKV valve for about 1 second in every 6 seconds (Its period time) during start up. The liquid in that 1 second burst must hit the S1 sensor position. This must bring the S1 sensor down to the EVAP temperature which in turn shows a high superheat to the controller. The controller shows REG CON 1 (REGULATION CONDITION) at this point .After about 60 seconds it then changes to REG CON 2 and if everything is OK, it then goes into REG CON 3 (PID) control. IF the start up procedure doesn't get the system away due to the poor positioning of the S1 sensor then the controller can go into a WET S2 procedure which basically shuts the valve to almost 0%. ie it thinks its flooded
I worked on some I believe Costan Full Glass Door cases on an M&S site. They would not come down to temperature from day 1. When I arrived on site the cabinet manufacturers engineer told me the valve was working OK with 5 K of superheat but the cabinets would not come down in temperature. When we checked the ELM controller sure enough it said 5 K superheat, the S1 sensor was -17deg C , and the S2 sensor was -12 deg C = 5 K superheat .
But the actual evaporating temperature on the suction gauge was -28 Deg C . So the evaporator superheat was -28 from -12 = 16 Deg K superheat. After moving the sensor onto the coil block from where it was fixed on a brass plate brazed onto the 3 rd return bend of the cooler, the AKV valve suddenly burst to a fully 100% opening degree and soon the cabinets achieved their set point temperature. So the S1 sensor was reading a superheat temperature and the S2 temperature was reading a super superheated temperature, and not the saturated and superheat temperatures as required.

Glenn Moore
23-01-2016, 11:57 PM
Hi Billy
Check parameters o30, 020, & o21.
o30 is the refrigerant type
o20 is the min pressure of the transmitter (Normally minus 1 )
o21 is the Max pressure of the transmitter ( probably 49 Bar or 99 Bar )
If they are OK check the signal wire voltage on the 550

24-01-2016, 03:31 PM
thanks once again for your help . next time on site I will check , but sure I check through them on front end with others ta

24-01-2016, 09:35 PM
Very interesting posts. This thread has brought back memories from the many times i've spent first looking for the s1, s2 sensors on different manufacturers cases, then re-positioning and or changing faulty probes. I seem to remember certain manufacturers fitting duplicate probes when AKV first started to appear, such was the unreliability of the sensors then. I also remember a particular site where we had endless problems with the AKV valve body stems jamming closed or sometimes open, upon splitting open the valve body the piston part seemed to have expanded a little, causing it to become jammed inside the valve stem. At the time we used to try different wattage coil thinking maybe the stem was getting too warm? Overtime the problem seemed to go away - can't remember if we ever got to the root cause, maybe Glenn could shed some light? Now I wonder if moisture may have been a factor.

Glenn Moore
24-01-2016, 10:18 PM
Hi Conrod
No it wasn't moisture, it was simply the 2 black rubbing seals on the armature which were expanding with the refrigerant. If you took the armature tube off of the AKV body , the armature would be stuck in the tube. But after about 30 minutes the seals would shrink and the armatures would simply drop out of the tubes. Changing the seals and the dimensions sorted the problem. I spent many hours working through the night sorting that one .