I need to connect a 100mm Dryer flexible Vent Hose to a 102mm pipe

[IanB]

You are right about the drawbacks of what I suggested:
- possible condensation on the radiator
- more cumbersome design
- more complicated mechanical construction

However, we do not need to worry about the temperature of the cold air (unless you want to sit directly in the draft), and we do not need to worry about heat from the room warming up the radiator. That is what is meant to happen, otherwise the radiator would not be working.

It all comes down to an energy balance. The ice and/or cold packs inside the ice chest represent a reservoir of "cold" (= lack of heat). "Cold" can be measured in kJ, just like heat. We can do an energy balance on "cold" flow just like we can do an energy balance on heat flow. This is just the same as considering conventional current flow from positive to negative, or electron flow from negative to positive. Both perspectives lead to the same conclusions.

Our goal is to release the store of "cold" into the room at a controlled rate, measured in kJ/h. It doesn't matter if heat from the room warms up the radiator or exposed piping, because this is just "cold" escaping into the room, which is wanted anyway. The outcome is that the rate of cooling of the room is slightly higher than otherwise.

Having lower air temperature is not a measure of performance in and of itself. A small flow of colder air has the same cooling effect on a room as a larger flow of less cold air. But a larger flow of colder air will just exhaust the cold reservoir faster, which is fine if it's what you want, but not fine otherwise. It's all about energy balances and rates of change.

Where outside heat matters is, as you suggest, any heat dissipated by the fan or the water pump motors. These are indeed bad sources of heat that will reduce the efficiency of the whole setup. So ultimately, you want the most efficient motors possible, which are probably DC brushless motors.

And just a note: if there is a box fan, there is no blower motor, and no flexible ducting. That all disappears.

But in the end, you should go ahead with the design you have planned if you are already along that path. It has videos where people have built it, and presumably it works as demonstrated.

[IanB]

Just for reference, here is a sketch of the concept I have in mind:

[YouCanDoIt]

However, we do not need to worry about the temperature of the cold air (unless you want to sit directly in the draft), and we do not need to worry about heat from the room warming up the radiator. That is what is meant to happen, otherwise the radiator would not be working.

It all comes down to an energy balance. The ice and/or cold packs inside the ice chest represent a reservoir of "cold" (= lack of heat). "Cold" can be measured in kJ, just like heat. We can do an energy balance on "cold" flow just like we can do an energy balance on heat flow. This is just the same as considering conventional current flow from positive to negative, or electron flow from negative to positive. Both perspectives lead to the same conclusions.

You demonstrated that what seems to be obvious to the eyes doesn't necessarily match the laws of physics and the energy balance seems to be one of such cases.

Please correct me if my following assumptions are still correct:

• The colder we can make our cold source (the bottom ice chest) the better is. Besides a well designed insulation, the coldness of the output airflow depends mainly from the cold source and the colder is the airflow, the faster are cooling times and this allows better energy balance along with power saving (when target temp is reached we can slow down the process and enter in some sort of "maintenance mode"). The process of making the cold reservoir as cold as possible does not involve special high power consumption devices, but is just a matter of finding the right design. For this I have opted for the gel ice packs since they last longer. I'll only use 1000ml bags (the biggest I could find), but it appears that there are different variants. This discussion talks about the different characteristics like freezing point, melting point, latent heat, heat capacity etc. This is again physics territory and at some point, instead of losing time in reading lot of papers, I ended up to save in my purchase list some of this cheap gel ice packs for which there aren't even specs (I need my cooler now, not after the summer). You know very well physics, so I would really appreciate if you want to suggest me some particular variant which allows to make the reservoir coldest and long lasting as possible.
• You said that I do not need to worry about the temperature of the cold air as long as there is the right energy balance and I understand the concept. However, supposing two scenarios, both with the same good energy balance design. The first one has 5 Celsius output airflow while the second one 15 Celsius. The first one will perform better anyway due to the various reasons described in the previous point.

Our goal is to release the store of "cold" into the room at a controlled rate, measured in kJ/h. It doesn't matter if heat from the room warms up the radiator or exposed piping, because this is just "cold" escaping into the room, which is wanted anyway. The outcome is that the rate of cooling of the room is slightly higher than otherwise.

I totally agree on the controlled rate concept and with this is in mind I have on my purchase list a controller board which slow down the fan speed (and the water pump flow) based on the detected temperature. You may want to give a look at the thread I opened on that matter. The one I found is very limited in capabilities, but is the better I could find with the needed amperage and that works with 2 wire fans. I opened a second thread because I found another much better board, but it only works with 4 wires fans and no one here seems to know how can be adapted to work with my 2 wire fan, so I ended up with the limited model that I know to work (I have not yet done the purchase, so, if you know how I can adapt the 4 wire controller board with my fan I would really appreciate).

Thanks for the tip about the two exposed pipes you see in my diagram. Those are the tubes that comes with the radiator and I would say they are already well insulated (16mm external diameter, 9 mm internal diameter). I was going to do the mistake to buy some pipe insulation foam for further insulation!! 
At this point I'm wondering if I should replace such tubes with smaller and less insulated ones. I have not been able to show such detail in my diagram, but the two holes on the ice chest box lid are not 16mm as you may expect, but 10mm. This because I want to reduce any heat penetration inside the ice chest box. Basically, between the lid and the external pipe there is one "10mm to 16mm reducer adaptor". This means that tube inside the ice chest (and the hole itself) is 10mm which expand to 16mm when outside. Based on what you said I'm afraid this is another flaw and I should just use 10mm tubes everywhere except may be for the left tube (the one that pushes the hot water in the reservoir, we want this well isolated and avoid it to push further heat inside the reservoir).

Having lower air temperature is not a measure of performance in and of itself. A small flow of colder air has the same cooling effect on a room as a larger flow of less cold air. But a larger flow of colder air will just exhaust the cold reservoir faster, which is fine if it's what you want, but not fine otherwise. It's all about energy balances and rates of change.

I totally agree, the air temperature alone is not a measure of performance. There are many variables that comes into the play like the flow size/speed etc. However, at the same flow size, speed etc, a lower temperature still provide better performance since it allows better control on energy saving, cold preservation etc (in combo with a temperature controller I plan to buy, for example).
Whether this target should be achieved by optimizing the cold source, the working layer or both, I'm not sure (for this reasons I'm focused on both).
The hardest part is the working layer:
one air intake? Two air intake? Container dimensions should be reduced? The right gasket size to sucks the most possible largest area from the radiator? Etc etc
I think the best I can do is buying a couple of styrofoam boxes and experiment with the different configurations (fortunately they are cheap).

Where outside heat matters is, as you suggest, any heat dissipated by the fan or the water pump motors. These are indeed bad sources of heat that will reduce the efficiency of the whole setup. So ultimately, you want the most efficient motors possible, which are probably DC brushless motors.

Yes, heat dissipated by the fan plays a role in reducing efficiency:


Do you have any idea on what could be a workaround? 11F is a noticeable loss!

Regarding the pump I was planning to buy this one. It is commonly used in many similar DIY projects and provides inlet and outlet tube connectors for non-submersible install (which I suppose is the best one).
Do you see any heat dissipation issues? it's just a couple of dollars and I can invest extra money if you want to recommend some better alternative. This one is already DC brushless.

And just a note: if there is a box fan, there is no blower motor, and no flexible ducting. That all disappears.

But in the end, you should go ahead with the design you have planned if you are already along that path. It has videos where people have built it, and presumably it works as demonstrated.

You seem to confirm what were my suspects. Combining two different devices like a blower motor and a box fan can result in many adverse side effects during their interaction.
If I see that I'm unable to get all the cold from the radiator even with the best configuration, the only think I can do is adding a secondary blower. I can reuse the left hose and probably a cheap motor could be more than enough for what has to be considered as an helper.

Many thanks for the sketch, I really appreciate. Except for the box fan is very similar to mine. I placed the water pump outside the cold container to avoid any heat dissipation within it (and save it from an additional hole needed for the power wire).
You attached the tubes vertically rather than horizontally like my design. I also thought to do the same while designing, much more clean and less waste of space. However, I would have lost directionality of the airflow and detachability (for when I just need the ice chest as a picnic chest). With a more complex design I would probably be able to achieve the same with a vertical connection, but I'm not sure if it worth the hassle.

Yes, I'm already on that path, too late to come back. The show must go on and before the end of the summer because from where I'm writing this message is nearly 90F!!

[YouCanDoIt]

To reduce the heat introduced from the air intake it came to my mind the idea to use a peltier module:



For simplicity I omitted the water pump and many other irrelevant details from the diagram.
Basically I changed the position of the air intake to the vertical wall and the peltier module system just on top.
The air sucked from the outside through the air intake will have to go through the COLD heatsink.
A Peltier module does not create cold, but only transfer heat and I think this is probably what I want:
by redirecting a certain amount of heat back to the room, the radiator, the water pump and in general the whole system, will stay colder resulting in an airflow with lower temperatures. The ice from the reservoir should also last longer because the water pushed back from the radiator should have a lower temperature.
Maximum size of a peltier module is 50mm square size with 40mm being the most used one. Pretty small, but I think enough to process a good amount of heat.

Mine could be a very bad idea and in contrast with thermodynamics, but intuition combined with my sixth sense make me think that could work.

Apart my peltier module (crazy?) idea, I have a concern about my original design. The position of the air intake:



All designs I saw are based on the first diagram. The air intake is basically placed as far as possible from the radiator and I suspect the main reason is to keep the radiator and the water as colder as possible (the same reason for which I thought to the peltier module). However, the air intake placed just above the radiator like the second diagram should prevent a good amount of heat from getting into the working system.
I feel like the two options are going to produce different results and the second one does not necessarily means worse results especially with my design (which, unlike the common ones, the cold reservoir is kept on a separated and well insulated container).

I'm in doubt and I'm afraid we are again into thermodynamics which is obscure territory for me.

At this point I think I can proceed to buy all the missing parts and I can't wait to post here the results of my project!
I still have concerns about optimizations and issues described in my previous message, but there is time for this, as is it should work wonderfully.

Thanks to all again especially to IanB.

[IanB]

The colder we can make our cold source (the bottom ice chest) the better is.

But you cannot make it colder. Assuming water is the working fluid, the lowest possible temperature of water is 0°C. Which means you can never make the water going through the radiator colder than that.

Besides a well designed insulation, the coldness of the output airflow depends mainly from the cold source and the colder is the airflow, the faster are cooling times and this allows better energy balance along with power saving

Unfortunately, the temperature of the airflow has no bearing at all on efficiency of the system, other than personal comfort if you want to sit in the breeze.

To reduce the heat introduced from the air intake it came to my mind the idea to use a peltier module

Using a Peltier module can only make things worse. A Peltier module consumes electric power, and power is heat, and heat is not wanted. This is for the same reason that putting a refrigerator in the middle of a room with its door open will heat the room up.

Here is a simple calculation for cooling a room, if we assume the ice chest contains water ice. The calculation with gel packs will be similar, but I don't have the numbers for those. Also note that none of these calculations will involve the temperature of the cold air flow.

Firstly, assume a typical room size of 4 metres by 4 metres by 3 metres. We will assume the room is perfectly insulated and no heat can get in from outside (not true in reality, but gives a best case scenario).

The volume of the room is thus 4 x 4 x 3 = 48 m³

The density of air is 1.293 kg/m³.

Therefore the mass of air in the room is 48 x 1.293 = 62 kg

The specific heat capacity of air is approximately 1 kJ/kg/degC.

Therefore, suppose we want to cool the room down by 5 degrees C, then the heat removal required will be:
M cp ΔT = 62 x 1 x 5 = 310 kJ

This is the amount of heat to be removed to cool the room air down, one time only, by 5 degrees.

Next, consider the ice (the source of cooling).

The latent heat of fusion of ice is 336 kJ/kg. This is nearly the same as the heat removal calculated above.

Therefore, it will take about 1 kg of ice to cool the room down by 5 degrees, once, assuming the room has perfect insulation.

In reality of course, the room is not insulated, and it will keep warming up again. The rate of heating depends so much on the room, but let's assume it happens in less than an hour.

How much ice can you put in the ice chest? Maybe 10 kg? It depends on how big it is.

So I think the device could achieve a little bit of cooling, for a short while.

Probably it is better not to try to cool a room down, but rather to direct the cooled air directly at you, and just enjoy the cool air flow while it lasts.

[IanB]

A follow up note. A small window air conditioner might be able to remove about 5000 kJ of heat from the room air per hour. This clearly is vastly more cooling than the little ice box thingy can achieve. So I would not expect too much from it in terms of actually cooling a room down.

[YouCanDoIt]

The colder we can make our cold source (the bottom ice chest) the better is.

But you cannot make it colder. Assuming water is the working fluid, the lowest possible temperature of water is 0°C. Which means you can never make the water going through the radiator colder than that.

Good point! Well, in this case we want to be as close as possible to 0°C. But are we sure that water will not go below 0°C when there are several gel ice packs submerged?
Yes, water, it's the only fluid I saw in such projects and I can see why: the cold source is basically free from costs. Are there better alternatives?

Besides a well designed insulation, the coldness of the output airflow depends mainly from the cold source and the colder is the airflow, the faster are cooling times and this allows better energy balance along with power saving

Unfortunately, the temperature of the airflow has no bearing at all on efficiency of the system, other than personal comfort if you want to sit in the breeze.

I'm afraid we're again in thermodynamics territory which is in contrast with what seems obvious to expect. At least there is the advantage to get immediately some comfortable breeze without waiting for the time needed to cool down the room.

However, if I have understood well your previous statement, you seem to confirm what was my belief: getting the cold reservoir as close as possible to 0°C should be one of the target to reach.
The thermal camera screenshots I posted shows 67.9F (20°C) coming out from the reservoir and I'm not sure why such high temperature. The other project (the one with smaller motor and one air intake) shows a temperature of 1.5°C.
Given that both the configurations are pretty much the same along with the parts being used, I suspect that is the secondary air intake that ruins the internal temperature. It seems that for maximum efficiency, I need to identify the correct number of air intake and their dimensions. And such choice is strictly related to the blower motor size, its power, radiator type along with many other details of the cooling system. I think this will be the hardest part

To reduce the heat introduced from the air intake it came to my mind the idea to use a peltier module

Using a Peltier module can only make things worse. A Peltier module consumes electric power, and power is heat, and heat is not wanted. This is for the same reason that putting a refrigerator in the middle of a room with its door open will heat the room up.

Power consumption is something I must care about since low energy consumption is one of my main targets. The water pump is only 400ma, but the blower motor is 6A. I saw so many people talking about perltier modules power consumption and although I think to only need a 400mm block, the energy efficiency seems to depend from how the system is designed.

Apart from consumption, I suppose you refer to the heat side (the hot heatsink) of the peltier module system showed in my diagram, right?
What if we redirect the generated heat outside of the room just like does regular air conditioners?
The modification of an existing window is a very cumbersome operation and not feasible. However, it should be pretty easy to get through a door with a small exit pipe with little or any modification at all. We could do the same even for whole working system although I believe it would be useless if we use a peltier module (which reduces the heat entering from the air intake through the cold heatsink).

Here is a simple calculation for cooling a room, if we assume the ice chest contains water ice. The calculation with gel packs will be similar, but I don't have the numbers for those. Also note that none of these calculations will involve the temperature of the cold air flow.

Firstly, assume a typical room size of 4 metres by 4 metres by 3 metres. We will assume the room is perfectly insulated and no heat can get in from outside (not true in reality, but gives a best case scenario).

The volume of the room is thus 4 x 4 x 3 = 48 m³

The density of air is 1.293 kg/m³.

Therefore the mass of air in the room is 48 x 1.293 = 62 kg

The specific heat capacity of air is approximately 1 kJ/kg/degC.

Therefore, suppose we want to cool the room down by 5 degrees C, then the heat removal required will be:
M cp ΔT = 62 x 1 x 5 = 310 kJ

This is the amount of heat to be removed to cool the room air down, one time only, by 5 degrees.

Next, consider the ice (the source of cooling).

The latent heat of fusion of ice is 336 kJ/kg. This is nearly the same as the heat removal calculated above.

Therefore, it will take about 1 kg of ice to cool the room down by 5 degrees, once, assuming the room has perfect insulation.

In reality of course, the room is not insulated, and it will keep warming up again. The rate of heating depends so much on the room, but let's assume it happens in less than an hour.

How much ice can you put in the ice chest? Maybe 10 kg? It depends on how big it is.

So I think the device could achieve a little bit of cooling, for a short while.

Probably it is better not to try to cool a room down, but rather to direct the cooled air directly at you, and just enjoy the cool air flow while it lasts.

My room size should be approximately the same (at maximum 30% smaller) and all the other parameters you mentioned match.
I'm afraid that trying to estimate whether it is possible to cool down a whole room to the desired temperature is a challenge.
For example the "Deluxe DIY Air Cooler!" project seems to have a similar room with an additional hallway of unknown size. It's unclear about the time needed, but he has been able to drop the whole room temperature from 27C to 25C. Whether 2C drops is the maximum achievable or not I don't know.
However, even if the desired temperature is only achievable on a limited area (let's say 2 cubic meters) it would not be bad although I designed the whole unit with in mind higher expectations.

It would be great if 1 kg of ice can cool the room down by 5 degrees.
The ice chest is 52QT, so I would say it should be able to contain around ten gel ice packs positioned vertically and probably additional 6-7 ones placed horizontally on top of them. Each gel pack is 1 liter and I can't find the weight on the specs.
Hard to estimate, but I believe it should be more than 10Kg. Problem is that the refrigerator does not have such capacity (unless I take off the frozen foods).

1 hour is the time lapse you estimate the unit will be able to keep the desired reached room temperature until the ice melt down and cooling capability ends up, is that right?
Cool air in the project I have just mentioned lasted 4 hours using 2 frozen gallon jugs and a couple of regular ice-packs with a temperature dropped from 27C to 25C. There are too many variables that makes impossible an estimate, but under the same conditions, given how my unit is designed and the parts being used, I would expect at least twice (at worse).

A follow up note. A small window air conditioner might be able to remove about 5000 kJ of heat from the room air per hour. This clearly is vastly more cooling than the little ice box thingy can achieve. So I would not expect too much from it in terms of actually cooling a room down.

Unfortunately in none of such projects I saw such data available, so I can't compare. Regular air conditioners have the advantage to transfer the heat outside the home. This is why I'm thinking to a peltier module integration. Probably my design for which I posted the diagram is not good because I suspect that the cold heatsink of the peltier system placed in front of the air intake still allows some heat penetration even if the heatsink is below 0C. There must be a way to completely redirect all the heat. Heat redirection should be probably the easiest task: a blower motor placed above the hot heatsink (rather than the black fan you see in the diagram), then a duct from the blower outlet until the room door, then a small adaptor that send the airflow into the slot located between the bottom side of the door and the floor (in this way I can even keep the door closed).

However, what I see with an 8 row radiator like mine seems to give good hopes:


Same location and conditions except that room temp dropped to 15C instead of 25C.
Unfortunately I can't do any reliable comparison with the unit running under the bigger blower motor (the one for which I posted the thermal camera screenshots) because it has been tested outdoor, but even in such conditions the guy was still able to cool an area.

So, before thinking to peltier systems and any other stuff I should focus on how to get all the available cold provided from the radiator because as is, with a 270CFM four inches blower, I suspect that a good amount of cold is lost.
If you have any suggestions, please kindly let me know.

Honestly I really don't know what to expect. Your calculations does not match at all what I see in similar projects, but it is better that I'm prepared even for the bad scenario you expect. And with this well in mind, I think that I should postpone the purchase of the temperature controller, am I right? If I will not be able to reach the desired temperature in the room, there is no point to regulate the blower speed based on the detected temperature because I will have to keep it running at full speed all of the time.
Edit: A controller could be helpful even in such case. I could use it to regulate fan speed to cool just the target area, no more and no less.

[IanB]

However, what I see with an 8 row radiator like mine seems to give good hopes:

The unit in the video with the desk fan and the radiator is exactly like the illustration I posted above in the thread. This is, in my opinion, the simplest and most effective way to provide cool air in a room.

Same location and conditions except that room temp dropped to 15C instead of 25C.

Not quite. The air going through the radiator may have dropped from 25°C to 15°C, but the 15°C air immediately mixes with the warmer air in the room. The room itself does not cool down to 15°C.

But generating a nice stream of cool air is a good benefit.

So, before thinking to peltier systems and any other stuff I should focus on how to get all the available cold provided from the radiator because as is, with a 270CFM four inches blower, I suspect that a good amount of cold is lost.
If you have any suggestions, please kindly let me know.

There is nothing to think about, and no cold is lost (except that which leaks outside the room through the walls, windows and doors). In the video, a big block of ice was placed in the bucket. All of the cold in that ice, every bit of it, goes into the room. Guaranteed.

The only way to lose efficiency is the power consumption of the fan motor and the water pump motor. The power going into these motors becomes heat that tries to warm the room up. But there is nothing much you can do about this, so you just have to accept it.

Honestly I really don't know what to expect. Your calculations does not match at all what I see in similar projects, but it is better that I'm prepared even for the bad scenario you expect. And with this well in mind, I think that I should postpone the purchase of the temperature controller, am I right? If I will not be able to reach the desired temperature in the room, there is no point to regulate the blower speed based on the detected temperature because I will have to keep it running at full speed all of the time.

Have you seen anyone else do a calculation? Yes, absolutely, you should not worry about temperature controllers. There is no real point.

The calculation is simple. You bring ice (or cold packs) into a room. While the ice or cold packs are melting, they are cooling the room down. When they have finished melting they are done and no more cooling is available. The only purpose of the water pump, radiator and fan is to help bring the room air into contact with the cold source and circulate the air around the room.

[YouCanDoIt]

However, what I see with an 8 row radiator like mine seems to give good hopes:

The unit in the video with the desk fan and the radiator is exactly like the illustration I posted above in the thread. This is, in my opinion, the simplest and most effective way to provide cool air in a room.

Same location and conditions except that room temp dropped to 15C instead of 25C.

Not quite. The air going through the radiator may have dropped from 25°C to 15°C, but the 15°C air immediately mixes with the warmer air in the room. The room itself does not cool down to 15°C.

But generating a nice stream of cool air is a good benefit.

Yes, I remember. At this point I'm really curious to see what is the most efficient. The guy built both, so I'm going to ask more details. What I saw by reading the discussion is that with the blower motor version he was able to cool down the whole room while with the box fan version he just said that it does a nice job of cooling a small to medium size area (which is unclear what exactly means).

Yes, you're right! 15°C is the air going through the radiator and this is a whole completely different story. He didn't mentioned any change in the room temperature, so I start to believe it only works as a stream of cool air.

One thing to take into account is that humidity matter and the reason for which the other guy, with 32% humidity, chose a blower fan which can withstand with moisture. Here we are at no less than 50%.
High humidity can cause many issues. A box fan is more suitable for dryer climate and from what I see in the comments, there are still chances of issues. It's still possible to overcome such issues even on humid climate, but requires an additional system.

So, before thinking to peltier systems and any other stuff I should focus on how to get all the available cold provided from the radiator because as is, with a 270CFM four inches blower, I suspect that a good amount of cold is lost.
If you have any suggestions, please kindly let me know.

There is nothing to think about, and no cold is lost (except that which leaks outside the room through the walls, windows and doors). In the video, a big block of ice was placed in the bucket. All of the cold in that ice, every bit of it, goes into the room. Guaranteed.

I must admit that I am a bit confused. The guy that built such systems said that air not getting through the radiator can be one of the many reasons for inefficiency and given that I'm sucking air from a 250mm x 300mmm radiator through a 102mm fan, to me seems that 2/3 of the cooling capability of the radiator is lost. If I can only target a 102mm area, than I'm wondering what would be the advantages of a bigger radiator like this.



The guy also seems to consider the output temp as the main factor that affect cooling capability, the exact opposite of what you say. I'm not saying that he is right and you're wrong. I'm just trying to understand why such conflicting opinions in order to see on what I should be focused on.
High humidity is another important factor he included in the list and given I have no less than 50% I should start to think of a possible implementation to overcome this issue.

I'm afraid I'm doing the mistake to fix the issues without even know if they are issue I will have to face. The best I can do is to build the system as is and later see what can be improved.

The only way to lose efficiency is the power consumption of the fan motor and the water pump motor. The power going into these motors becomes heat that tries to warm the room up. But there is nothing much you can do about this, so you just have to accept it.

Honestly I really don't know what to expect. Your calculations does not match at all what I see in similar projects, but it is better that I'm prepared even for the bad scenario you expect. And with this well in mind, I think that I should postpone the purchase of the temperature controller, am I right? If I will not be able to reach the desired temperature in the room, there is no point to regulate the blower speed based on the detected temperature because I will have to keep it running at full speed all of the time.

Have you seen anyone else do a calculation? Yes, absolutely, you should not worry about temperature controllers. There is no real point.

The calculation is simple. You bring ice (or cold packs) into a room. While the ice or cold packs are melting, they are cooling the room down. When they have finished melting they are done and no more cooling is available. The only purpose of the water pump, radiator and fan is to help bring the room air into contact with the cold source and circulate the air around the room.

I should always keep in mind your very simple explanation on how works cooling. However I still think that many other parameters play their own role (like for example the air intake number and dimensions, the guy that built the "Deluxe" cooler explained the details in the notes).

Regarding the temperature controller, see my edit. May be I can still get some benefits if I reduce power only to the minimum necessary to cool a specific area rather than trying to cool the whole room at max speed (which would be a waste of resources if the system is unable to complete such task). Heat generated from water pump and fan should also decrease with less power. Not to mention about noise reduction.
I'm doing again the mistake to find ways to enhance functionalities without even know if there any functionalities that need to be enhanced

P.S. You didn't said what you think about my idea to redirect the peltier module generated heat out of the room. This was what you said to be the bigger issue. Pretending that my idea work, do you see any benefit from preventing the room heat from entering inside the system through the air intakes?

[IanB]

P.S. You didn't said what you think about my idea to redirect the peltier module generated heat out of the room. This was what you said to be the bigger issue. Pretending that my idea work, do you see any benefit from preventing the room heat from entering inside the system through the air intakes?

I don't think a Peltier module has enough capacity to make a difference when cooling a whole room. The golden principle in engineering is "keep it simple". Even if you could redirect the heat from the Peltier module outside the room, it would just be adding complexity to the whole arrangement. I would never even think of it.

No, there is no benefit preventing the room heat entering the system through the air intakes. No benefit at all.

You bring ice into the room. When the ice is melting, it absorbs heat from the room, and cools the room down. It really doesn't matter how the room heat gets to the ice, it cools the room down just the same.

The simplest possible design is to put a big block of ice on a tray in the middle of the room, and arrange for a room fan to blow air over the ice. This is as efficient as you can get. All the cold gets into the room, and no cold is wasted.

Every other design is just adding cost and complexity to that, with no improvement in efficiency. The only advantages of other designs are:
1. Maybe they can cool the room down faster
2. Maybe they can regulate the cooling to make it last longer
3. Maybe they are more convenient to operate

Keep in mind the definition of efficiency in this context:

Efficiency = (Total cooling delivered) / (Total cooling available)

If you bring a block of ice into a room and let it melt, the efficiency is 100%, because all the heat to melt the ice came from the room.

If you make a contraption with fans and motors and pumps, the efficiency is less than 100%, because the motors generate heat which is the opposite of cooling.