Overpumping emitters.

Harvey Ramer

If you want acruate results from the UHF you have to correct for density. Makes a difference.

Paul48

It's not more efficient to run a tight DT. You lose more efficiency with every degree under designed-for DT.

Gordy

True Harvey, but the ratios remain the same. 485ish at 180* .

bob_46

Gordy, not saying more efficient . More flow equals more turbulence and more heat transfer. Not desirable to exceed about 4.5 FT/SEC but it would transfer more btus.

Gordy

In the engine cooling of the bmw it's a one way street. Get the heat off the motor, and it dumps into the air. A very high delta in that scenerio of a cylinder head at operating temperature. flow is variable with engine rpm if there was a water pump involved.

Is the superheated micro bubbles of steam inherent to the bmw's cooling jackets design? Or is this something that happens in most liquid cooled engines that bmw decided to capitalize on as marketing hype?

Paul48

The higher DT at lower flow says differently. I'll go along with more btus presented, though.

Gordy

Okay in a closed loop system. If one molecule of water is sent out at 1gpm it spends X time at HX, and X time at emitter. If I double the flow to 2 gpm it spends 1/2 the time at the HX, and 1/2 the time at the emitter yet makes a lap twice as fast so it's a wash?

hot_rod

Lets just look at one common emitter, the fin tube. Jumping from radiant loops to fin tube to forced convectors may be complicating the conversation.

Each type of emitter would have it's own specific model. this is super simple to model with the HDS software or others, but if you like long hand rathe here goes.

Figure 2-5 shows where most baseboard specs run out to 4 gpm, beyond that flow the output probably isn't worth chasing, but the graph runs out to 8 gpm, in 3/4 tube that is getting to the fps where the tube will start talking to you.

bob_46

Don't know Gordy.

hot_rod

Here is what it looks like in a simulation program.

Keep in mind after you balance the entire system with all zones calling, when any zone or zones turn off the flow does not stay at the previous balance in the remaining "open" zones. The only way to "nail" a consistent flow in each zone would be to use a dynamic balance valve, like a PI pressure independent valve. That type of valve moves a cartridge inside in response to the varying ∆P in the system.

A ∆P circ gets you closer, a ∆P and PI better yet.

\

Gordy

The graphs show what I was trying to get at. The point of diminishing returns of flow rates once a certain point is reached. If more flow has better heat transfer then the graphs should not flatten out. Should keep climbing. Unless something drastic changes far beyond the flow rates illustrated. By then the pipes are talking, and friction becomes its own energy source. Of course that would be for that given water temp, and delta.

Harvey Ramer

Turbulent flow definitely has better heat transfer. That has never been a question. Turbulence happens with increasing flow rate. But laminar conditions can exist at higher flow rates than expected. The best analogy is that it's like balancing a ball on top of a peak. The higher the velocity, the higher and steeper the peak. Ball drops off due to a vibration or disturbance, flow goes turbulent.

In order for a piece of fintube to have diminished output at higher flow rates, there has to be a problem with heat transfer from the fluid to the tube.

If the flow was laminar it would show a narrow delta. By slowing the flow to a 20 degree delta it could possibly turn the laminar flow somewhat chaotic due to convective thermal currents and get more heat in contact with the tube. Chaotic flow is different than turbulent flow.

Bubbles in the solution would also show reduced output with a narrow delta.

Water does not need contact time to transfer energy. Energy is transferred on the atomic level from agitated atoms crashing into each other.

hot_rod

Gordy said:

The graphs show what I was trying to get at. The point of diminishing returns of flow rates once a certain point is reached. If more flow has better heat transfer then the graphs should not flatten out. Should keep climbing. Unless something drastic changes far beyond the flow rates illustrated. By then the pipes are talking, and friction becomes its own energy source. Of course that would be for that given water temp, and delta.

I suppose that is possible also. Maybe at 15, 20, 50 gpm the curve, for this particular model starts dropping off. Belimo seemed to indicate that in the coil performance curve they modeled.

The question is what would it take to pump 50 gpm through 20 feet of 3/4 fin tube? At that point the heat from the motor on the pump, and the friction might replace any required boiler

Jumper suggested 3 times minimum flow in a coil is considered over pumping? If that is a rule of thumb, 3- 4 gpm would be over pumped in this fin tube example? The curve is still rising as far out as 8 gpm. I'll see if the sim software will let me flow out to those crazy high numbers , at some point it errors
out.

It is an interesting discussion, maybe over pumping is best defined by the eye or hand of the beholder.

Harvey Ramer

Gordy
The reason that curve looks the way it does is because they are only looking at changing the pumping speed with a constant Supply temp. After a certain point of flow volume, the emitter starts reaching saturation and the curve starts flattening out. The convection on the air side does not increase proportional to the increase in flow because it can only absorb a given amount of heat at the supplied water temp and is also restricted from flow within the emitter itself. The only way to make an emitter's performance match the linear line of a buildings heat load is to vary the water flow and water temp simultaneously.

Paul48

DT circ and a mod/con on ODR. Some folks seem to have a problem with forcing a DT. Perhaps a circ running on ODR?

Gordy

Some don't have a problem Paul.

Paul48

It's no more trickery, than running a buffer tank. I think the very definition of an "over-pumped emitter" is one that sends most of it's btus past the emitter and back to the boiler. You select a primary circ based on DT, why not force the DT on the secondary side and keep the boiler from bouncing off HL? Avoiding the "Chicken and Egg" debate(gobbledy-gook), what is the problem with doing that?

hot_rod

Harvey Ramer said:

Gordy

The reason that curve looks the way it does is because they are only looking at changing the pumping speed with a constant Supply temp. After a certain point of flow volume, the emitter starts reaching saturation and the curve starts flattening out. The convection on the air side does not increase proportional to the increase in flow because it can only absorb a given amount of heat at the supplied water temp and is also restricted from flow within the emitter itself. The only way to make an emitter's performance match the linear line of a buildings heat load is to vary the water flow and water temp simultaneously.

Ahhhh grasshopper, now you see what ∆T pumping with flow only adjusting has it's limits, and locks on top and bottom end. If the math keeps working, why would you limit a ∆T circ at 5-50?

Because the math breaks down at the extremes, demonstrated by the curve I've been showing on the high and low end, and the math here. Show me a system transferring heat energy at a .01 gpm flow and a 2000 ∆. The math states it's possible.







Adjust flow and adjust temperature, now you got it.

Back to a more ideal method of VS pumping working with the boiler modulation, not against it. ∆P while not perfect either could allow a mod con to ramp temperature while the pump matches varying flow adjustment.

hot_rod

Paul48 said:

It's no more trickery, than running a buffer tank. I think the very definition of an "over-pumped emitter" is one that sends most of it's btus past the emitter and back to the boiler. You select a primary circ based on DT, why not force the DT on the secondary side and keep the boiler from bouncing off HL? Avoiding the "Chicken and Egg" debate(gobbledy-gook), what is the problem with doing that?

There is no problem with adding a buffer to help keep the boiler from short cycling. if you have a fixed output, single speed boiler and a zoned, especially micro zoned system it is an excellent idea. A boiler would never cycle if it's output matched the load, exactly and continually. In reality the systems we design are for that one point in time. It may never be there Or it could be there often. Boilers cycle when the load starts going away, not because of the emitter sending excess molecules back.

Here again is the concern with a fixed ∆.

Let say 4 gpm in the fin tube example shown above is the definition of over pumping, even though the fin tube manufacturers show it as an option. That is the Belimo approach, prevent a low ∆ that just isn't worth the effort. More like a limit switch, not a variable. Or as a fix as they stated for "faulty assumptions at design"

Lets look at the heat output at a more reasonable flow of 1- 2 gpm.

At 1 gpm Iget about 8200 btu/hr transfer.

At 2 gpm I get maybe 9300 btu/hr transferred. Agreed so far?

Now look at the red line showing output as ∆ changes. IF you were to lock in at say 20∆ you the circ will attempt to modulate speed to stay at that ∆, agreed?

So you have locked in to a set output .08 gpm, 5000 but/hr. So what happens when the load changes?

Now IF you could modulate temperature at the locked ∆ then output would increased via higher emitter temperature, same gpm, that has some value and with a buffer allow that to happen.

There is no math manipulation here, no marketing spin. These output models and fin tube data are directly from the IBR testing and certification, it published and well accepted data.

The notion that any heat emitter “wants to,” or even can, remain at a fixed temperature drop as the flow rate changes is not supported by these results.

Rich_49

I offer this definition of overpumping ;

1; circulating more fluid than necessary to do the work efficiently .
2; circulating fluid at a rate that causes the source to reach high limit and shut off for a period .
3; the point at which BTU output begins to level off relative to energy used to circulate it with diminishing effect .


circulating fluid at any rate that shuts off the source because heat energy is being returned to the source therefore forcing it off for a period cannot guarantee a higher AWT through any circuit regardless of what type emitter is being employed .

Most all source equipment lists a Delta at which the equipment was tested at . Without P/S to circulate fluids at different but equal rates of heat transfer is an exercise in futility and should never qualify as " Best Practice" .

Can a source , emitter combination ever reach maximum system efficiency if the source is being shut off and the temp within it is declining for any period ?
Can you guarantee a higher AWT for an hour (60 minutes) if this happens ?

For fin tube I suggest this number lies between 1 gpm and 4 gpm since these numbers are not a mystery as they are both used in most manufacturers sizing data . I have forever noticed a warning also that states if one cannot guarantee 4 gpm the 1 gpm numbers should be used .
I opine that nobody can guarantee the 4 gpm and an AWT in a system that the source cannot run continuously .

Moving 4 gpm and getting 5.7% increase in out put over moving 1 gpm is a waste of resources and I do not believe that 5.7% increase can be guaranteed throughout an entire hour without overly complex equipment and controls . It cannot be achieved even at design with a perfectly sized boiler , emitters .
If we can determine the exact point between 1 and 4 gpm where the output turns right as opposed to upward on a chart I think we can than determine a Reynolds number that will be over 4000 and a point at which heat energy transfer does not take place as efficiently .



All this negation about the importance of or ignoring the laminar regime in favor of turbulent flows when our biggest challenges are seen at large percentages of the heating season which are well above design conditions is fascinating .

Rich_49

hot rod said:

Gordy said:

The graphs show what I was trying to get at. The point of diminishing returns of flow rates once a certain point is reached. If more flow has better heat transfer then the graphs should not flatten out. Should keep climbing. Unless something drastic changes far beyond the flow rates illustrated. By then the pipes are talking, and friction becomes its own energy source. Of course that would be for that given water temp, and delta.

I suppose that is possible also. Maybe at 15, 20, 50 gpm the curve, for this particular model starts dropping off. Belimo seemed to indicate that in the coil performance curve they modeled.

The question is what would it take to pump 50 gpm through 20 feet of 3/4 fin tube? At that point the heat from the motor on the pump, and the friction might replace any required boiler

Jumper suggested 3 times minimum flow in a coil is considered over pumping? If that is a rule of thumb, 3- 4 gpm would be over pumped in this fin tube example? The curve is still rising as far out as 8 gpm. I'll see if the sim software will let me flow out to those crazy high numbers , at some point it errors
out.

It is an interesting discussion, maybe over pumping is best defined by the eye or hand of the beholder.

There is one problem with your first sentence in this post Bob . The Belimo presentation modeled nothing , that was a REAL dataset from an active coil in a REAL building under REAL TIME conditions .

Why is it that this industry and some of it's participants refuse to recognize new findings that have now been proven and employed for quite some time ?

Rich_49

Hatterasguy said:

hot rod said:

So you have locked in to a set output .08 gpm, 5000 but/hr. So what happens when the load changes?

Caveats:

A single zone of fin tube.
A single point in time, therefore a single SWT.


The load does not change.



Vary the SWT and the load will change.............and the circulator will change speed to maintain the ΔT.

This last statement is Factually incorrect .

Varying supply water temp has ZERO IMPACT on load . Load is determined by the Delta between out door and indoor temps . the colder it gets outside the higher the load will be . At which point ODR will raise SWT and a Delta T circ will read the Delta between EWT and LWT and maintain it throughout a huge area between the high end of the curve and low end of the curve .

To address JBs choice of language which you always pick apart and state that Taco is lying about .
Based on the fact that the only thing that changes the load of a room is differences in out door temp or changing a setting on a thermostat , both of which increase or decrease delta , JB stating that the pump will speed up or slow down is in all actuality correct although possibly the language could have been chosen more carefully so cynical folks like yourself could not use a technicality against Taco and confuse the issue .

Rich_49

hot rod said:

Paul48 said:

It's no more trickery, than running a buffer tank. I think the very definition of an "over-pumped emitter" is one that sends most of it's btus past the emitter and back to the boiler. You select a primary circ based on DT, why not force the DT on the secondary side and keep the boiler from bouncing off HL? Avoiding the "Chicken and Egg" debate(gobbledy-gook), what is the problem with doing that?

There is no problem with adding a buffer to help keep the boiler from short cycling. if you have a fixed output, single speed boiler and a zoned, especially micro zoned system it is an excellent idea. A boiler would never cycle if it's output matched the load, exactly and continually. In reality the systems we design are for that one point in time. It may never be there Or it could be there often. Boilers cycle when the load starts going away, not because of the emitter sending excess molecules back.

Here again is the concern with a fixed ∆.

Let say 4 gpm in the fin tube example shown above is the definition of over pumping, even though the fin tube manufacturers show it as an option. That is the Belimo approach, prevent a low ∆ that just isn't worth the effort. More like a limit switch, not a variable. Or as a fix as they stated for "faulty assumptions at design"

Lets look at the heat output at a more reasonable flow of 1- 2 gpm.

At 1 gpm Iget about 8200 btu/hr transfer.

At 2 gpm I get maybe 9300 btu/hr transferred. Agreed so far?

Now look at the red line showing output as ∆ changes. IF you were to lock in at say 20∆ you the circ will attempt to modulate speed to stay at that ∆, agreed?

ANSWER ; Correct

So you have locked in to a set output .08 gpm, 5000 but/hr. So what happens when the load changes?

ANSWER ; Incorrect . You have not locked into a set flow . when the load increases so will the SWT by way of ODR , The sensors will track S/R delta and vary flow accordingly . In this instance they would probably see a widening Delta and speed up flow . Your next statement is what most of us do which you admit has value .

Now IF you could modulate temperature at the locked ∆ then output would increased via higher emitter temperature, same gpm, that has some value and with a buffer allow that to happen.

RESPONSE ; Ah Grasshopper !

There is no math manipulation here, no marketing spin. These output models and fin tube data are directly from the IBR testing and certification, it published and well accepted data.

RESPONSE; These publications which are well respected and accepted need updating based on testing of new products . They are the source of many of the flawed assumptions we make and allow gentlemen such as yourself to post them as proof that certain new technologies have little to no value . Thusly I would state that manufacturers also should do some new testing . It would go a long way toward insuring that new up and comers learn better ways .



The notion that any heat emitter “wants to,” or even can, remain at a fixed temperature drop as the flow rate changes is not supported by these results.

RESPONSE ; Old results done with old technology me thinks . Stated above , new updated stuff should be done . There are no dinosaurs because they could not adapt .

hot_rod

Rich said:

hot rod said:

Paul48 said:

It's no more trickery, than running a buffer tank. I think the very definition of an "over-pumped emitter" is one that sends most of it's btus past the emitter and back to the boiler. You select a primary circ based on DT, why not force the DT on the secondary side and keep the boiler from bouncing off HL? Avoiding the "Chicken and Egg" debate(gobbledy-gook), what is the problem with doing that?

There is no problem with adding a buffer to help keep the boiler from short cycling. if you have a fixed output, single speed boiler and a zoned, especially micro zoned system it is an excellent idea. A boiler would never cycle if it's output matched the load, exactly and continually. In reality the systems we design are for that one point in time. It may never be there Or it could be there often. Boilers cycle when the load starts going away, not because of the emitter sending excess molecules back.

Here again is the concern with a fixed ∆.

Let say 4 gpm in the fin tube example shown above is the definition of over pumping, even though the fin tube manufacturers show it as an option. That is the Belimo approach, prevent a low ∆ that just isn't worth the effort. More like a limit switch, not a variable. Or as a fix as they stated for "faulty assumptions at design"

Lets look at the heat output at a more reasonable flow of 1- 2 gpm.

At 1 gpm Iget about 8200 btu/hr transfer.

At 2 gpm I get maybe 9300 btu/hr transferred. Agreed so far?

Now look at the red line showing output as ∆ changes. IF you were to lock in at say 20∆ you the circ will attempt to modulate speed to stay at that ∆, agreed?

ANSWER ; Correct

So you have locked in to a set output .08 gpm, 5000 but/hr. So what happens when the load changes?

ANSWER ; Incorrect . You have not locked into a set flow . when the load increases so will the SWT by way of ODR , The sensors will track S/R delta and vary flow accordingly . In this instance they would probably see a widening Delta and speed up flow . Your next statement is what most of us do which you admit has value .

MISSUNDERSTOOD
In these examples and models it is a fixed temperature boiler, no reset or ODR. Intention is to demonstrate increased output with increased flow, SPECIFICALLY without change to SWT.

Now IF you could modulate temperature at the locked ∆ then output would increased via higher emitter temperature, same gpm, that has some value and with a buffer allow that to happen.

RESPONSE ; Ah Grasshopper !

There is no math manipulation here, no marketing spin. These output models and fin tube data are directly from the IBR testing and certification, it published and well accepted data.

RESPONSE; These publications which are well respected and accepted need updating based on testing of new products . They are the source of many of the flawed assumptions we make and allow gentlemen such as yourself to post them as proof that certain new technologies have little to no value . Thusly I would state that manufacturers also should do some new testing . It would go a long way toward insuring that new up and comers learn better ways .


RESPONSE to RESPONSE:
Never claimed new technology is not possible, Caleffi leads the industry in that respect, as you may have noticed by all the "knock-offs" of our product. If we did exactly what everyone else was, we would not be growing in a a fairly stagnant industry.

The difference between what we develop is proof that it actually does what we claim. Extensive R&D and a extensive amount of lab and actual installation testing to assure it works as promised. When installed properly, of course.
We don't ask the installers and customers to do our R&D or "hype" our products and new technology.
We are the exact company you talk about when it comes to leading innovation, and teaching newer methods Some examples the Quicksetter, Two piped buffer tanks, zone and switching relays with unique features, leaders in hydraulic & magnetic separation, talking and offering solutions to water quality. I could go on..



The notion that any heat emitter “wants to,” or even can, remain at a fixed temperature drop as the flow rate changes is not supported by these results.

RESPONSE ; Old results done with old technology me thinks . Stated above , new updated stuff should be done . There are no dinosaurs because they could not adapt .

Response to Response:
show me your numbers, math, simulation, data collection, FEA models, etc to support your "me thinks"
Without it you are just guessing, or wishing.
I've laid out my opinion with all the current data, based on sound thermodynamic and fluid dynamics principles currently being taught around the planet. If you have "new age" data, adjustments to the laws of thermodynamics or theories, show us and support it with definable numbers and facts, not wishful thinking.

From one gentleman to another

Harvey Ramer

Well, the conversation went where I hoped it wouldn't. I guess it was inevitable given the close relationship between the 2 subjects.

I'm still curious as hell about the overpumping phenomena. And how to explain the decrease in output with increased flow. It is quite contrary to all the physics material I can get my hands on.

Paul48

What is the definition of turbulent? Is there some guidelines of what condition "must" be present for a flow to be labeled turbulent. I envision the threads of a screw, and at a lower flow the threads would be tight. And as you increase flow the threads stretch out. At some point, across an emitter it would create a "mock" laminar flow. How's that for a WAG?

Harvey Ramer

Sorry Paul. Your wag is out in left field. I have attached Chapter 8 of fluid mechanics. If you read it you will understand. It is very good reading and has a lot of illistrations.

Paul48

Damn Harvey.......Let me give you the other cheek.....hold on

Rich_49

Bob ,
I find it hard to explain why a man with your superior intellect cannot grasp data shown by industry entities as factual . I have shown so many links in discussions with you and for others to view that it boggles the mind .

Nothing I have stated is un explainable nor out of the realm of reality as many agree . The list of those who recognize Delta T maintenance in heating and cooling systems is as follows . Grundfos , Belimo , Taco , several ASHRAE fellows , AFUE testing procedures for boilers of all types . We size circs based on information which includes Delta T ( designed for , not forgotten about )

I agree with all that Caleffi does and that they are innovative and a fine company indeed . What has that to do with the fact that a new technology exists that works , is proven ?

I am one man whom works hard designing and installing systems taking more into account for every one of these designs than you yourself states is required . These systems are out there in the world performing flawlessly and providing optimal comfort to their end users while allowing them to keep more of their hard earned dollars . I cannot justify added cost for logging equipment to prove what is known , especially when the data already exists but is disregarded by many .

There are 2 ways to to control and/or vary output from any emitter . Vary the SWT , vary the flow rate . Chapter 9 MHH 3 . Are we to believe that employing both is not possible ? Do you take us all for imbeciles ?

Then there is this quality piece by your very own Jody Samuel

https://www.youtube.com/watch?v=6mNECDHgDrg&index=12&list=PLuuV0ELkYb5VE0I4evUZ30b5U78CRlRdg

Points of interest are at 21:00 , 35:48 , 48:00 and then something really illuminating for those with exceptional perception and reasoning skills starts right at 51:03 .

http://www.taylor-engineering.com/Websites/taylorengineering/articles/ASHRAE_Symposium_AC-02-6-1_Degrading_Delta-T-Taylor.pdf

https://physics.ucsd.edu/students/courses/winter2013/physics218b/Batch56.pdf

Although it may seem that Harvey's well thought out question has gotten off track I believe he has in fact pointed something huge out that cannot be ignored and has been forgotten about or intentionally ignored by smarter men than us . Again , it is an unsolved problem and has been since the beginning of fluid flow and heat exchange research 100's of years ago .

Wait till the Europeans realize that using Delta T strategies can further decrease their energy consumption and increase system efficiencies.

Harvey Ramer

This is interesting and contributes to the mystery.

Onsager reciprocal relations[edit]
Main article: Onsager reciprocal relations
In fluid systems described in terms of temperature, matter density, and pressure, it is known that temperature differences lead to heat flows from the warmer to the colder parts of the system; similarly, pressure differences will lead to matter flow from high-pressure to low-pressure regions (a "reciprocal relation"). What is remarkable is the observation that, when both pressure and temperature vary, temperature differences at constant pressure can cause matter flow (as in convection) and pressure differences at constant temperature can cause heat flow. Perhaps surprisingly, the heat flow per unit of pressure difference and the density (matter) flow per unit of temperature difference are equal.

This equality was shown to be necessary by Lars Onsager using statistical mechanics as a consequence of the time reversibility of microscopic dynamics. The theory developed by Onsager is much more general than this example and capable of treating more than two thermodynamic forces at once.[10]

https://en.wikipedia.org/wiki/Transport_phenomena

The faster we pump water through an emitter, the larger the pressure drop across it. Is it possible that the fluid sets up an internal heat flow from high pressure to low pressure?

hot_rod

Rich said:

Bob ,
I find it hard to explain why a man with your superior intellect cannot grasp data shown by industry entities as factual . I have shown so many links in discussions with you and for others to view that it boggles the mind .

Nothing I have stated is un explainable nor out of the realm of reality as many agree . The list of those who recognize Delta T maintenance in heating and cooling systems is as follows . Grundfos , Belimo , Taco , several ASHRAE fellows , AFUE testing procedures for boilers of all types . We size circs based on information which includes Delta T ( designed for , not forgotten about )

I agree with all that Caleffi does and that they are innovative and a fine company indeed . What has that to do with the fact that a new technology exists that works , is proven ?

I am one man whom works hard designing and installing systems taking more into account for every one of these designs than you yourself states is required . These systems are out there in the world performing flawlessly and providing optimal comfort to their end users while allowing them to keep more of their hard earned dollars . I cannot justify added cost for logging equipment to prove what is known , especially when the data already exists but is disregarded by many .

There are 2 ways to to control and/or vary output from any emitter . Vary the SWT , vary the flow rate . Chapter 9 MHH 3 . Are we to believe that employing both is not possible ? Do you take us all for imbeciles ?

Then there is this quality piece by your very own Jody Samuel

https://www.youtube.com/watch?v=6mNECDHgDrg&index=12&list=PLuuV0ELkYb5VE0I4evUZ30b5U78CRlRdg

Points of interest are at 21:00 , 35:48 , 48:00 and then something really illuminating for those with exceptional perception and reasoning skills starts right at 51:03 .

http://www.taylor-engineering.com/Websites/taylorengineering/articles/ASHRAE_Symposium_AC-02-6-1_Degrading_Delta-T-Taylor.pdf

https://physics.ucsd.edu/students/courses/winter2013/physics218b/Batch56.pdf

Although it may seem that Harvey's well thought out question has gotten off track I believe he has in fact pointed something huge out that cannot be ignored and has been forgotten about or intentionally ignored by smarter men than us . Again , it is an unsolved problem and has been since the beginning of fluid flow and heat exchange research 100's of years ago .

Wait till the Europeans realize that using Delta T strategies can further decrease their energy consumption and increase system efficiencies.

Excellent article confirming all my opinions.
Granted this is a chilled water system which he indicates is often designed around a 10- 14∆T.
He doesn't indicate what is considered too low ∆, but refers to degrading ∆. Most of his observations as he explains are improperly designed, sized or balanced. PI valves as he suggested solve many problems, but do add pressure drop.

His graph shows the same result, higher flows, higher coil output, he does show a SWT change also which I agree and suggested with hot water system, change flow and change SWT. Variable burners to eliminate cycling, variable circulators.

If I not mistaken many (most) fixed speed chillers include buffer tanks. Specific tanks are built for chiller water use. We experienced this in our building in Milwaukee. A fixed 10 ton chiller was installed with 5 AHs connected. Cycled it's brains out. I contacted Carrier and they suggested our unit should never have been installed without a buffer and they make a buffer tank that sits right below the chiller, sheetmetal matches up. Ours was them last of the fixed speed/ buffer combos, the next version was variable output.
Ssame fix for fixed boiler cycling I would add, buffers smooth out bumps..

Most interesting was his observation the coils that should be dropped into laminar flow conditions at low gpm rates may have turbulent flow incurred by the tight loop ends on every pass of the coil. This agrees with my thoughts of adding turbulators or spiraled tube to keep it turbulent at low flow conditions. When the turbulence disappears, the energy transfer does also.

Throughout the article the point of turbulent flow is stressed.

And the "conclusion sums it up nicely that "many" ∆ problems can be resolved by proper design, component selection proper operation and maintenance. I'd suggest most in lieu of many as it seems he deals with problematic system engineering:)

His take is near identical to the Belimo position "faulty assumption at design" faulty equipment sizing, lack of balancing devices, and or skill to set the balance properly both water and air side.

He embraces variable speed at the source (chiller) as we do with mod cons.

Absent anywhere is the mention of ∆ maintaining pumps or valves. As I read it slight moving ∆ is not an issue, degrading ∆ is. But with such tight ∆ design in chilled water systems it is less forgiving. With hydronics if you design around a 20∆ and it moves either way 5, it is not a huge issue. If you design around 15 or 10∆, a 5 ∆ move is maybe what he refers to as "degrading"

Rich_49

Taco has offered a simple solution to declining Delta for heating or cooling . Problems being addressed is a good thing . maybe Mr Taylor will write us another paper .

Unfortunately nobody has picked up on one key thing contained in many of these things I have posted . Why is it that the Reynolds numbers we are used to seeing have the transitional numbers between 2300 - 4000 while in several places in these documents they are described to be between 2300 - 10000 ? Could it be there is a distinct difference between heat exchangers ( closed boundary ) and various heat emitters
(open boundary) ?

Harvey Ramer

@hot rod

Most interesting was his observation the coils that should be dropped into laminar flow conditions at low gpm rates may have turbulent flow incurred by the tight loop ends on every pass of the coil. This agrees with my thoughts of adding turbulators or spiraled tube to keep it turbulent at low flow conditions. When the turbulence disappears, the energy transfer does also.



The correct term would be Chaotic flow. Turbulent flow and Laminar flow cannot exist at the same time. When Transitional, they are jumping back and forth. However, Chaotic flow can exist in Laminar conditions due to external interference in the flow. Such as fittings or in this case, U-bends of the coil. Even still, the entrance dimension is only 10 pipe diameters, before the chaotic flow disappears and you get a full bodied laminar flow.

hot_rod

Harvey Ramer said:

@hot rod

Most interesting was his observation the coils that should be dropped into laminar flow conditions at low gpm rates may have turbulent flow incurred by the tight loop ends on every pass of the coil. This agrees with my thoughts of adding turbulators or spiraled tube to keep it turbulent at low flow conditions. When the turbulence disappears, the energy transfer does also.



The correct term would be Chaotic flow. Turbulent flow and Laminar flow cannot exist at the same time. When Transitional, they are jumping back and forth. However, Chaotic flow can exist in Laminar conditions due to external interference in the flow. Such as fittings or in this case, U-bends of the coil. Even still, the entrance dimension is only 10 pipe diameters, before the chaotic flow disappears and you get a full bodied laminar flow.

Glad you have a handle on this

https://en.wikipedia.org/wiki/Chaotic_mixing

Brewbeer

What about convective currents that set up in tubes experiencing laminar flow? Water adjacent to the tube wall cools, resulting in a density gradient between the water at the wall and the water in the tube center, inducing convection currents in the tube.

Harvey Ramer

@Brewbeer
I theorized on that as well. A full bodied laminar flow has a bad heat transfer coefficient. But if you slow the flow enough, convective currents should be produced and make a somewhat chaotic flow which would improve the heat transfer coefficient. I'm just not sure that scenario fits.

Gordy

There was a day when most everyone thought the world was flat. Yet another that thought the sun, and planets in our solar system revolved around the earth. Mere nano seconds ago in the grand scheme of time for us as humans on earth. A mind in a box is a terrible thing to waste.

Paul48

After reading that article, it would seem that the scientific community knows very little about turbulent flow. I speculate that achieving a high DT across an emitter is not because the flow has changed to laminar to the emitter. You've just moved the point of regime change to the end of the emitter.

P.S. I was going to draw a picture of it, but I broke my crayon.

Gordy

I have an extra white one you can borrow Paul.

Paul48

Ummmm..... Gordy....That's chalk.......Put the beer down!