Hello,
I am cooling 350 liter water tank by circulating water through external flat plate HEX. On the cold side is cold water aprox 14 degC 600kg/h and on the hot side is tank with the initial T of 70 degC at 600kg/h. Lenght 80 min.
Can someone suggest why the temperature stops the rapid cooldown at ~22 degC and only after 40 min cools to the temp of cold water. Similar image is for 360 and 120 kg/h.
Image attached (T_ET_in - return to tank bottom; T_CW - cold water). It is cylindrical tank with outlet and inlet at sides. Outlet close to top. Inlet close to bottom.
Do you have a drawing of your system, in particular where the temperatures are measured? It is possible that the thermal stratification in the tank, and variations in the internal convection currents are the causes of that cooling schedule.
Added a schematic (not up to size) and other temp graph. Temperatures for inlet of cooled water are measured 1 m before a flex pipe connecting to tank. Similarly 1m at top. Initally the cooling is stratified - but coming closer to 20C tank gets mixed up and thats where it sits for 30 min. I understand the slowing down of cooling but not sure how to describe that "table" from 10th to 40th min.
Was wondering if: 1) HEX could be the reason in connection with pressure drop - the expansion tank cannot provide stable pressure - 2 bar for all length of experiment, usually drops from 2 to 1 bar.
2) if I make it so that cooling is done at 20C would it give similar picture just with the table at something like 30C.
And the main problem is that when simulated by TRNSYS it does not appear - so either I find a way to remove it in exp or add in simulation - but not sure what to add.
Your tank is obviously strongly stratified after 10 minutes, and cools down from bottom to top. During this time, the cooling power of the HEX , propoortional to T_CW-out - T_CV_in and to T_ET_out-T_ET-in, is nearly constant, since it cools down hot water coming out from the tank top. When the "cold" water front reaches the tank top, the whole tank is at about 25 °C, that is slightly warmer that T_ET_in because some warm water mixes with the entering colder water. The cooling power is then reduced (smaller temperture difference) and so is the cooling speed.
The first 10 minutes is the time needed to put the cooling in steady evolution. Note that the cooling time constant, estimated from 350 kg/600 kg/h= 0.39 h that is 35 minutes.)
Best regards
If the TRNSYS model does not reproduce the measurements, it means the tthe model does not reproduce the system.So check if your system works as modelled, or if theTRNSYS model is adapted to your experiment. Maybe it does not model the mixing that ofccurs at the cooling front.
Starting at 10 minutes I see a perfectly stratified tank. At 40 minutes the tank is 'empty' i.e. most of the heat is removed. Theoretically the tank should be 'empty' after 350/600x60=35 minutes. That suggests that the tank is not cooled in the first 5 minutes. That is also showed in the graph: the inlet temperature T_ET_in is app. 65°C for the first ~7 minutes.
I think what you see in these first minutes is the transition of a standstill situation to the normal operating situation.
I totally agree with the reaction of Claude-Alain Roulet
I suggest you check things like:
- are the pumps started simultaneously (are they started at t=0?)
- is there a buffer tank in the primary cooling circuit?
- where are the temperatures T_CW measured?
Thank You Claude and Ron. Initially I could not figure out how to know the T at which the table will appear, however now it seems obvious by looking at graph that the dT = (T_ET_out-T_CW_out) = (T_ET_in - T_CW_in). And by adding external HEX in TRNSYS I do indeed get a similar result. I am surprised how did I not notice it sooner!
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Plating