Heat Exchangers - Serial or Parallel?

A location would be good.

The inlet water temperature/some guess of the potential temperature difference of the air being cooled/the volume of air being cooled. Assume volume of water is unlimited?

I'm sticking with my original answer: parallel

You'll have more airflow in parallel, but the air will not be cooled as much as in series.

In series the smaller amount of air will be twice as cool (or have half the heat i suppose, If i want to be correct), but again the airflow will be half.

I'm not in HVAC, so my knowledge is very basic. Coulden't say which would work better for your particular scenario. I believe in order to run the right calculation you will need to know such things as the outside ambient air Temp, the Desired internal ambient air tem, the water temp, the water flowrate, the air flowrate, and probably a bunch more.

HVAC guys?

"I'm designing a system to cool a residence by passing chilled water through several fin tube heat exchangers. If we assume that the air flow through each exchanger is equal, and each exchanger is identical, should the exchangers be connected in series or in parallel re water to maximize the cooling of the air? Assume that in the parallel case the water flow would be equally divided amongst the exchangers. Is it possible to give an answer with this information, or is more info needed, and, if so, what info and why?"

NOTE: the exchangers are not in series or parallel for air flow; each exchanger will have it's own air flow, regardless of how the water is routed.

OK. The note clears things up a bit.

If you put them in series, then the first one will have the initial chilled water running through it. It will transfer heat from the surrounding air into the water that goes to the second exchanger. That means that the second exchanger will not cool the surrounding air as much. If this is for a split level residence, that might be beneficial, as the first unit could be on the top (hottest) level; and the second on the ground (coolest) level.

If you put them in parallel, then they will both cool the surrounding air by the same amount. But you will need twice the flow with this arrangement. But this will work better for rooms/area that are on the same level.

I use two water to air heat exchangers stacked together under my furnace for my hot water heat system in my house.

I have the hot water coming in to the one that the air goes out of. And the cooler water leaves from the one the air goes into. Its the most efficient heat transfer method for getting the maximum temperature change from inputs to outputs.

I believe its called a stacked reverse parallel exchanger.

GA. Indeed. Reverse stacked heat exchangers will give you the lower air temperature possible, as you want, with the limitation of restricting air flow. If the air flow through one heat exchanger passage sized is all you want, then you have your answer.

If they were in parallel, you would have the maximum air flow chilled and, because the air would be in a higher temperature for both heta exchangers, this would give you more heat exchange, but through the use of a higher air flow and higher temperature.

I'd go Parallel.

The short answer is no it is not possible to give an answer with this information. There are advantages and disadvantages of each depending on the situation.

In an earlier answer you stated that the airflow was not parallel or in series. Do you mean the fin tube heat exchangers are independent from each other with respect to airflow (cooling different areas of the residence)?

Kilowatt0 made a good point and a good example of a situation where a series connection would be the best choice, but for most applications where you would be cooling areas independent from each other parallel would be preferred.

If efficiency is your main concern both air and water should be in series as a primary/secondary configuration with the water supply on the air leaving side. One example of this would be the "reverse stacked" configuration that bhrescobar mentioned.

One other consideration to keep in mind is the condensate drain from each heat exchanger. Typically all the cooling coils would be in one location to easily facilitate the condensate piping.

If you can give us more detail on the layout of the residence, if there are separate zones to cool, if forced air duct exists, and just generally more detail maybe we can give a better answer.

Troy, I don't think the items you mention are pertinent to my question. I'm not looking for advice about how to build a specific system. I'm asking if, in the general case I described, would parallel or series water connection yield the most cooling of the air. As to air flow, assume that the exchangers are independent and each has the same amount of air flow. That is, the total air mass divides equally to the two exchangers and merges after passing thru them.

I'm an EE and out of my specialty here, so I can't be sure if factors such as water temp, air temp, flow rates or whatever would change the answer as to which connection is better. If more info is needed to give the answer, maybe someone can say what and why.

In all the years I've worked commercial HVAC, systems that use "water"- hot (heating) or chilled (cooling) for HVAC, all the distribution discharge coils were connected in parallel. Reasons (1) efficiency, (2) zoning control

For each option, you should determine the air temperature in and out of each exchanger, and the air flow; also the water temperature in and out of each exchanger, and the water flow. Evaluate the heat exchanged via LMTD (logarithmic mean temperature difference) in each case, and then compare the results. As Troy says, it does matter what you are trying to do. Sketches of each option would help.

In serial, the heat recover from the chiller water will be optimum. But the limitation could be, differential temperature between 'air and water (precisely, the inlets)'. Referring the coolers in serial, the second cooler (second stage) receives the chilled water at a higher temperature than that the first. So is third and so on and so forth. At a stage the water will not have enough chillness to cool the air further.

How many stages is optimum? Depends up on the 'differential temperature – delta T, the quantity (mass) and the specific heats of water and air. So, unless the masses (air and water) and delta Ts are known, it may not be precise, even to suggest a ballpark figure on number of stages. Since you already have the coolers (presuming), the area of 'heat transfer area' also has to be considered.

Several coolers…Imagine a combination arrangement, two (or more) coolers in serial (a set) and several sets in parallel is sure to give an optimum solution.

Parallel is best.

Why?

The airflow slows down due to being divided between two coolers, so that the air is cooled longer. More air is exposed to a lower temperature.

If you need even cooler air, then slow it down some more....

If series, the flow will be more restricted and as cooling/warming is also a question of temp difference, in the 2nd rad the difference will be less, so the cooling effect will be leass....

Are we to assume that these heat exchangers are on a common plenum or on separate systems. Incoming water also has a lot to with design. Is this a new system complete or are there existing constants that will be unrealistic to modify or change. Is this to be something to pre-condition for an existing system? I realize that these are a lot of questions with no answers, I am sure others will have more.

Ah, I see many opinions, but not much engineering. Let's try again. Forget the real world! This is a question of theoretical design fundamentals. I'm looking for a rigorous mathematical or logical explanation of whether the water flow to two identical air-to-water heat exchangers should be in series or in parallel to maximize air cooling. Assume the air flow divides equally to the exchangers and merges at the outlets. Assume water flow is constant, endless, and always cooler than the air. If you need more info, tell me what info and why it is pertinent.

I have a feeling that this is Thermo 101, but I'm a EE and its been 50 years since I took that class.

In the case you describe above Parallel will provide the highest average ΔT and therefore yield the most cooling per given amount of water.

Troy, how do you come to your conclusion? Do you have any math or logic to support it?

Hi SSCpal,

I don't have any math to support my conclusion off the top of my head just several years in the HVAC trade.

As for logic, let's compare the two scenarios. In a parallel connection both exchangers will operate at equal efficiency. In a series the first coil will actually operate at a higher efficiency because the water will only be in the exchanger for 1/2 the time (therefore staying cooler and having a higher ΔT), but the second coil will operate at a decreased efficiency because of the higer temperature of the water entering and therefore lower ΔT. Common sense and experience tells me that the efficiency losses on the second coil will outweigh the gains from the first and the water leaving temperature from a series connection would be lower than from a parallel all other thing being equal.

Hi SSCpal,

What your're talking about is a standard Fan Coil Unit, we dont use them much in the UK but they are very popular in France and the rest of Europe.

Heating can be from hot water or an electric element. The hot water is pumped thro' the system, to heat the building and cold water pumped thro' to cool the building.

The temperature is measured by a simple bi-metal temperature element and fed back to a BMS to switch the water (hot or cold) for heating or cooling; with the centrifugal fan switched on or off from the BMS to boost cooling/heating.

The one's I designed were for an American company to be sold in Europe and were always in series.

Hope this helps,

merlin

Parallel is best, but flow will follow the path of least resistance. You will need some way to monitor and control flow.

Flow meters, temp gauges, and such will be needed. The slightest difference in plumbing can make a flow difference.

While working for Villanova U. I made a parallel/series heat exchanger for show and tell for the students. The results of that class are standard learning for all mechanical engineering students.

Mike! WHY is parallel the best!!?! You seem to have the answers...now we need the proof.

Like I said, the proof is Thermo 101. I didn't take the class, I was just the tech who had the skills to make the equipment. It had lots of temp and flow sending units for the students to apply their formulas to. We did lots of show and tell, it helped the students to grasp what they were learning in their text books.

I remember a welding class. I didn't worry about teaching the students to weld, I explained and showed them how to judge welding. There are ways to cheat. As engineers, they are going to be the inspectors and designers. They learn the theories and rules. They see welds in x-rays and microscopes, like no weldor ever does.

The same goes for your question here. If you want the engineering, look it up. You have many good explanations of the situation here, from folks that know what they are talking about. We don't give homework grade answers, we shoot from the hip.

You might just have hit the sore point, well put!

AS a foot note to my post; One should put down the Beer, pay attention in class and the answers could be forth coming !!

Yes, Stan's response was excellent and I will address it and reply to him shortly. There is no "homework assignment"; I ran onto this question while personally researching home cooling by lake water and remembering my grandfather's work with car radiators. I have been to Stan's link, and the series/parallel answer is not there. No "baiting", I assure you; The "lack of info" is intentional, because this is not yet a real world question and people kept getting off track wanting details about the "lake" or the "house"! All I want is just a rigorous explanation of a simple theoretical air-to-water heat transfer system. You know, textbook stuff. I think the answer can be given with the info I've provided, and I'm a little surprised that no one, except Stan, approached it that way. I've been combing the Web and haven't yet found any series/parallel material, but I continue to look. And CR4 is a place for us to help each other, n'est-ce pas?

http://trane.com/CPS/uploads/userfiles/productpdfs/unt-prc001-en_04292009.pdf

Here is literature with info. you desire.

(Trane.com/commercial/productsearch)

If you would like more info.

Oh Yea, Parallel !

These are called fan-coil units and commonly used in hotels. They are always in parallel. The reason is that if they are in series the inlet temperature of the units at the end of the series will be high and cooling will be poor since the cooling area of fins is the same for each unit.

Bioramani

Sorry, thats Trane.com/Commercial/LiteratureSearchList

http://trane.com/Commercial/LiteratureSearchList.aspx

Sasnak, Thanks very much for the links. I found lots of interesting and potentially useful information on commercial hardware. Also found a way to ask a Trane engineer my question. Will let you know what they say.

You are welcome, let us know.

Steve

Another good question might be weather to use parallel flow versus counter flow for the relation of the airfow to the water flow. That was another part of the demonstration device I made for the Engineering class. For my own education would someone that knows the preferred configuration chime in. It's thermo 101 again, but I'll learn it better if I see it in print. My guess is counter flow.

Also the rate of change plays a big factor in this equation. The efficiency of a heat exchanger is affected by how well it transfers the heat from one medium to the other.

If the rads are long, then counterflow is an excellent point, not to be missunderstood.

If the rads are made similar to say a car rad, then flow direction makes little or no difference.

Why not just get one properly sized radiator?

Damn! I gotta start thinking outside the box....