Heat Exchanger Question - For the engineering folks

BC_5

That sums it up nicely.

geoboy

Heat Exchanger performance

Yesterday I was told something about a heat exchanger that we just purchased which frankly, puzzles me. I thought I'd check it out with you experts.

Consider a shell-and-tube heat exchanger with an extremely large internal exchange surface. The shell and tubes both contain water piped counterflow. The entering temperatures of the hot and cold fluids run about 50 - 60 degrees F apart.

I have always been under the impression that, under the best of conditions, the exit temperatures of the two fluids would be equal, and under less than ideal conditions the hot fluid would exit the exchanger at a temperature somewhat higher than the exit temperature of the cold fluid.

Yesterday I was told that in the exchanger we just purchased, the exit temperature of the hot (incoming) fluid would actually be COOLER than the exit temperature of the cold (incoming) fluid. Somehow, this doesn't seem right to my limited brain cells. What do you think?

BC_5

Ideal would be exit of the hot side equal to INLET of the cold side if you had 100% efficiency and infinite area. Obviously this will never happen, but you could be between the inlet and exit temp of the cold side pretty easily in a counterflow design. With counterflow, the exiting hot side water exchanges with the incoming cold side water, so the hot side water can approach the cold side water temp. Someone smarter than me can probably come up with a good way to visualize this.

geoboy

Yes, the visualization part is difficult. So what your saying is (if I understand correctly) that it IS possible to have the following relationship among the temps in the exchanger:

T-COLD-IN < T-HOT-OUT < T-COLD-OUT < T-HOT-IN.

Thanks for your input. I'm simply trying to comprehend what's happening.

While on the subject, is the above relationship valid only in a shell-and-tube design, or could it occur in a plate-and-frame system?

EBEBRATT-Ed

That just dosent seem possible. The temperature difference of the two fluids will be closer togeather under low flow conditions with equal flow. As the flow increases (either stream) the heat exchanger becomes less efficient and the temperature difference will increase.

geoboy

Believe it or not, both shell and tub flows are being driven by the same pump, albeit through different paths. This exchanger (actually a pair of them in parallel) are components of a fairly complex off-peak electric hot water system. I'm really not in a position to guage the flows within the shells vs those in the tubes. I can say thay all circuit setters are set identically, and I strongly believe that the flow rates are well within the exchanger's design parameters.

The question, though is whether the hot fluid can emerge from the exchanger COLDER than the cold fluid emerges. It seems logical that slowing down the flows will allow more time for BTU exchange and therefore better efficiency, but common sense (not necessarily applicable here) seems to indicate that the two sides would approach EQUAL temps.

Any heat exchanger experts out here who might be able to shed som further light? Thanks in advance.

Andrew Hagen_2

Heat Transfer

Conductive heat transfer is driven by delta-T. It goes one way, from hot to cold. The heated fluid cannot be hotter than the fluid doing the heating. It would be analogous to water flowing uphill.

geoboy

That's what I thought. It appears that we are being sold a bill of goods!

The custom-built assembly in which these 2 exchangers are mounted is not performing properly. After examining the way it is piped, I thoght I had found a piping error, based on my assumption that the output of the exchanger's cold fluid side would always be cooler than (or at best equal to) the output of the hot fluid side. The designer told me otherwise, which would substantiate the existing piping arrangement. The solution he is proposing might work, but seems unecessarily complex.

I'm so confused!

BC_5

yes, the average temp cannot be lower...

but geoboy was talking about the EXITING hot side being colder than the EXITING cold side, which can happen under certain circumstances. GEA has an excellent sizing program, I'm attaching an example showing how this can be possible.

In this case the flow on the hot side is much less than the flow on the cold side to get the high delta-T. You could also increase the delta-T on the cold side to get the same heat transfer at more equal flows.

Jim Franklin

Impossible without the addition of external energy like a heat pump. I think something got lost between Engineering and Marketing.

I once saw an ad for a Gore-Tex jacket that had a picture of the breathable lining converting sweat molecules into Hydrogen and Oxygen. Burn all you want, we'll sweat more!

jim

ALH_4

Temps

I was confused about which temperatures geoboy was talking about. There is still no violation of hot to cold, but depending on the parameters of the system, either of the two exiting temperatures could be warmer. I am posting a sketch because describing this in words is confusing.

geoboy

Thanks and another question..

Thanks BC and Andrew. Your answers appear to support the designer's statement. Guess I learned something new!

Let me bother you with another question. On each of my 2 exchangers, there is a ball valve for isolation at the inlet and a circuit setter with positive shutoff at the outlet. This is true of both the shells and the tubes.

While discussing a possible flow problem with the designer, he indicated that we might perhaps need to install another circuit setter on the exchanger INLETs. This seemed odd to me as I thought that restricting the exchanger flow at EITHER the inlet or outlet would have identical effect on the flow, and therefore the performance.

Does it ever matter which end you restrict the flow at? (Assuming of course, that you are not throttling the input side of a circulator!)

Thanks for your input. This site rocks!

geoboy

Bump

Can anyone shed some light on the throttling issue?

BC_5

I can't imagine any scenario where it would make a difference - is it just a set and forget valve or is there some sort of temperature compensation?

hot_rod

I think

you want to be sure the throttle, or flowsetter is at the discharge side of the circ. Restricting flow at the intake to the circ could cause some cavitation issues.

I can't imagine you would need valves onn both sides, other then for isolation for replacement.

It would be best to get the circ sized as close as possible and not do too much throttling, as it just wastes away energy and drives the circ efficiency way down.

There are plenty of circ choices these days and the 3 speed versions really help spread the range.

hr

geoboy

The throttling valves are indeed on the circ discharge.

This is a really funky application. Basically, the circulator drives water through a pair of unequally sized tube bundles (in parallel), in order to harvest BTUs from a tank of superheated water (a off-peak water heater). Some of the water is also driven around the tube bundles. The water then proceeds to the custom-built assembly consisting of the 2 shell-and-tube exchangers mentioned in the obove posts, as well as 2 three-way thermostatic mixing valves and various pipes and balancing valves. The purpose of this external assembly is to functionally replace a similar device that was originally part of the off-peak water heater but was subsequently damaged. When the original unit worked properly, the off-peak unit was able to heat the domestic water at an even pace. The new unit isn't quite there yet.

Don't know if this info is of any use other than general interest!

hot_rod

It sounds

like you potentially have a lot of pressure drop through those combos of HX and mix valves. It would take some math to figure out pressure drop based on flow rates, and HX data.

Could it be you are short on circ head to handle all that?

hr

geoboy

Actually, we are capable of generating gobs of hot water now, especially when the off-peak unit is fully charged. Our problem is that this new assembly is not doing a good job of controlling the rate at which we harvest BTUs, so the temps vary excessively throughout the day. Examining the internals of the device, it appeared that we might have a piping error; hence my question about exchanger exit temps. Anyway, the designer thinks are flow rates are adequate.