Mysterious Loss in TEV Performance

Every now and then someone comes across a problem the details of which are worth archiving, usually with the intention that it could or should be used for training purposes. Component manufactures will often do this, some will turn the experience into a set of generic instructions that are intended to help future component appliers avoid similar problems, others will record a description of the particular scenario as experienced in detail from beginning to end. It's the detailed scenarios that I always find most interesting.

This article is a representation of one such ,classical, detailed scenario.

Figure 1

System description

The troublesome system was apparantly a small 1.18kW R12 ice-cream maker consisting of a freezer mixing section and a storage section, the storage section is called a conserver, both sections were fed from one common air cooled hermetic condensing unit and each section had it's own thermal expansion valve (TEV). The conserver system ran continuously cycling the condensing unit on temperature control. The freezer system would then be brought into operation when serving was required and this of course was done by energizing the liquid line solenoid valve (LLSV). The component configuration was as per Figure 1.

Trouble Symptoms

While the conserver section ran continuously without a hitch, the freezer mixing section consistently showed signs of obvious performance loss and this was always toward the end of the freezing cycle. Each time the freezer mixer started there was indeed the required rapid freezing of the product, however, just as the ice-cream hardness approached serving quality there was the said sudden loss in freezer evaporator performance causing the walls to in fact begin thawing and the ice-cream to quickly re-soften rendering the product unsuitable for serving.

Initial Speculations

Some areas a technician might be expected to examine first for symptom cause could be:

1. The freezer branch of the 'T' junction situated down stream of the liquid line filter drier, could there be a restriction there?
2. The mechanical suitability or integrity of the LLSV. Is the valve pilot operated or direct acting, if pilot operated does it have sufficient end-of-cycle flow induced pressure drop to function correctly and not restrict refrigerant flow?
3. The refrigerant charge quantity, would more refrigerant be required? Or would more refrigerant merely flood the condenser during pull down causing excessive head pressures to delay pull down?
4. Is there perhaps end-of-cycle preferential feeding of liquid refrigerant to the conserver circuit?
5. Is it possible that the increased evaporator temperature difference (TD) in the conserver section toward the freezer sections cycle-end is robbing compressor capacity.
6. Would the addition of a refrigerant receiver help?
7. Someone might even have thought the ice-cream's molecular structure altered with temperature possessed by some sort of thermo-molecular properties hysteresis. Just kidding.

Diagnostics Procedure

Often, to aid system diagnostics, system modifications of a temporary or even a permanent nature have to be applied. These modifications would be to either provide for visuals such as sight glasses or indirect visuals such as pressure gauges. In this particular case the foremost question to be asked would be "At end-of-cycle, is there a solid column of liquid ahead of the freezer section's TEV and is it at the requisite pressure?". A combination of liquid line pressure and surface temperature readings at the condensing unit might indicate all is good but the better technician would insist on knowing what these readings are immediately before the offending TEV which would of course require the addition of a pressure tapping there. Otherwise, the circumstances might suggest that a sight glass be installed instead. Which option to use might be a subjective decision. Before deciding, the technician tending this problem might consider a few principles:

1. At the end of a cooling cycle there is more refrigerant in the evaporator and of course this means less refrigerant in the condenser or liquid line available to be subcooled. The freezer evaporator being bottom fed would permit some liquid trapping there during the expected reduced end-of-cycle refrigerant velocities.
2. Since too the end-of-cycle condenser TD would have diminished so would, further, the degree of available subcool.
3. If, ordinarily, without overcharge, this system's design dictates end-of-cycle liquid subcool values of say 1K to 3K then the accuracy of a technicians instruments could leave too much room for doubt. Even with the most accurate instruments available, reading a 1K or even 2K subcool, there is no guarantee of a vapour-free liquid column available ahead of the TEV. It is quite normal to witness vaporous liquid flow coinciding with subcool readings of 1K and 2K.

Figure 2

Re-considering The Initial Speculations

1. The flow pressure drop through the 'T' junction, situated down stream of the liquid line filter drier, would reduce as the square of the refrigerant velocity and with low load mass flow reductions would likely reduce more so than any coinciding subcool loss.
2. The mechanical suitability of the LLSV could be considered good, the valve being of the direct acting type. ?
3. The refrigerant charge quantity could be verified as sufficient for end-of-cycle TEV supply by the installation of a sight glass at position SG1 seen in Figure 2, assuming there are no substantial pipe frictional or liquid lift related pressure drops.
4. Preferential feeding to the conserver section tending to starve the freezer section could be determined by the installation of a sight glass in the freezer sections liquid line branch. If the sight glass is positioned after that lines LLSV at position SG2 as seen in Figure 2 then the technician would only need to decide whether the offending component was either the 'T' junction or the LLSV. Certainly, the working parts of the LLSV could be temporarily removed to eliminate it from this short list of potential causes.
5. If the conserver sections TEV superheat setting is adjusted to increase conserver superheat just as the freezer section reaches its end-of-cycle run, the conserver can potentially be eliminated as a cause of overload in respect of compressor capacity, but then the resulting increase in liquid subcool could falsely remove the symptom of true cause.
6. The addition of a refrigerant receiver shouldn't really be necessary since the excess refrigerant found on the condenser side of the system anytime the freezer section solenoid valve is shut should be comfortably stored in the now proportionately increased condenser space available since now approximately only half of the condenser would be needed to handle the reduced heat rejection being brought from the conserver load on its own.

Diagnosis

As a result of some of the above considerations it was the technicians decision to install the two sight glasses already mentioned and shown in Figure 2 namely SG1 and SG2. During the next test run the technician observed that while SG1 showed a permanence of vapour free liquid approaching the said 'T' junction, SG2 clearly showed large volumes of vapour accompanying the liquid supplying the freezer sections TEV and occurring only near the end of the freezer sections run cycle coinciding with the drop-off in freezer section evaporator performance. The sight glass was showing with out doubt that the freezer sections TEV was not being fed the required solid liquid column needed for its rated performance. The same observations were made again but this time with the LLSV's working parts removed except that it now seemed the end-of-cycle vapour proportions seen in SG2 had increased. It was at this point that the technician speculations turned toward LLSV's solenoid coil itself. Touching the solenoid coil and then valve body the technician could feel they were both unusually warm, much more so than ordinarily expected. The technician then first partially blocked the condenser air flow to raise liquid line temperatures which had the expected effect of eliminating the seen liquid line vapour. This effect was because the temperature difference between the liquid and valve body had reduced and too the liquid line velocities had increased meaning there would have been less heat flow into a greater and more subcooled refrigerant mass with reduced likelihood of any re-vaporisation. Naturally, the same improvement was affected when power was cut from the solenoid coil allowing it and the valve body to cool.

The explanation established as a result of these diagnostics was that the solenoid valve was of an excessively compact construction selected to fit the tight space constraints designed into the overall system and it was this compact design that provided for what we might consider the efficient transfer of coil heat to the liquid line and the carried refrigerant.

In summary, the heat produced by this solenoid valve, albeit excessive, was proportionately sufficient to have the problematic effects due to the very low refrigerant mass flows associated with the freezer evaporator system's slight capacity. The freezer evaporator being bottom fed would tend to trap disproportionately more liquid refrigerant at low loads as a result of the associated reduced refrigerant velocities. The effect this trapping had on reduced available liquid line subcool coupled with the normal end-of-cycle reduced liquid line mass flows provided the right conditions for this very rare phenomenon to occur.

Solution

A solenoid valve of less compact design was used to replace the offending solenoid valve and so timely delivery of quality ice-cream servings were of course finally realised.