Less Heat Transfer When Pumping More Hot Water Through A Coil??? 5

[BronYrAur]

Back in September, I asked a question if over-pumping a coil could ultimately result in less heat transfer. See this link to view the old thread.

http://www.eng-tips.com/viewthread.cfm?qid=351624

It was told to me by a old-timer pipefitter who said that if you over-pump a coil, the water won't have as much delta-T and therefore you won't get as much heat transfer. I of course thought I knew better because this guy was forgetting that more flow at a smaller delta-T would still give you heat transfer.

I was pretty convinced with the replies to that old thread until I read an article in PME Magazine. See attached. The article branches out to a couple of different topics, but one of them is about not getting heat transfer because of the flow being too large. I highlighted the statement that confuses me the most. I contact the author to see if he had any scientific backup for this, but he didn't. It just seems to be a phenomenon out there that you run across through experience (like my pipefitter).

Does anyone have any experience with this?

 

[Compositepro]

The statement that more flow results in less heat transfer is just hogwash. He obviously does not understand how the heating system works well enough to explain his observations. We can't explain it either without knowing all the relevant details. However, well established scientific principles do not get discarded because a heating contractor can't explain why something happens.

 

[dvd]

I'd be curious about the flow regime and whether transition from laminar to turbulent flow enters into this. The boundary layer will be much thinner in the turbulent regime and won't pick up the heat. Seems like Pr and Nu numbers will shed some light on the heat transfer.

 

[crshears]

In another thread I asked if an effect like this could be due to a reduction in 'loiter time'...

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[LittleInch]

In both of the instances mentioned in the article, it is clear there is a complex network of heaters and heat exchangers. Probably also some bypass valves or other devices which either direct flow back to the return without going through a HX, or are somehow diverting water away from the heater(s) which need them. Also don't forget that a pump running flat out or beyond its curve will end up with a lower DP, hence those HXs with a higher DP will be starved of supply compared to one which is not. Those balancing valves are there for a reason which is to equalise the flow through the outlet system. What he probably did was end up with a lot more of the flow going through one or two heat exchangers at the expense of ten others which were then starved of flow.

In essence this is not a thermodynamics issue, but a system water flow issue. As the flow reduces due to higher DP across the balancing valves or the bypass is closed, more flow and with a higher DP can go through the heater(s) which people want to use. The key is to look at the DT across the boiler. If this starts going down despite more heat being required, the issue is water balancing of the flow, not a new thermodynamic law(!).

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Dougt115]

The water follows the path of least resistance.

As LittleInch stated it sounds like a circulation problem. The valves are not for temperature control on the older systems they are for pressure balancing between the different zones by equalizing the line resistances. By fully opening all of the valves the water took the easiest path, hence the shortest, and the zones furthest away got colder.

 

[LittleInch]

Assuming this is the case, there is a maximum amount of heat able to emitted from a single radiator or AHU depending on surface area / air velocity. Hence whilst the temperature of the few HX's with lots more flow than they should have got would have been emitting more heat than normal, the other HX's had a lot less flow and hence less heat output. Overall this meant less heat out and a low differential temperature across the boiler.

Simple when you think about it, but that's why the good systems should be regulating the max flow into each sub system or HX to prevent this sort of thing happening by persons who don't understand the issues of complex water networks...

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[IRstuff]

"loiter" time is a bit nebulous, but you can be more rigorous about the analysis.

Consider that the air flow has a limited heat extraction capacity, i.e.,
P_air = h * A_hx * dT_hx in watts.

The water stream has a certain level of mass transport,
vel * A_pipe * ρ, which confers a power when multiplied by σ * dT_water, so the power flow in the water is expressed as
P_water = vel * A_pipe * ρ * σ * dT_water

P_water > P_air typically, since the exchanger efficiency is always less than unity. Note that P_water can be arbitrarily large, if vel is made arbitrarily large, yet P_air is limited to a maximum value where its dT_hx is essentially unchanged throughout the exchanger. Therefore, there is indeed a range of velocities where the water is increasingly cooled, but beyond a certain velocity, the amount of cooling is saturated, and the apparent effect is cooling efficiency is decreased. That is only partly correct. The cooling efficiency, as measured by h, is unchanged, it's just that you're dumping way more heat than the exchanger can handle.

Consider a simple example, say, P_water = 1 kW, P_air = 500 W. Now increase vel by 3x, so P_water = 3 kW, but P_air is probably only going to be about 750 W. So, in the first case, you've removed 50% of the heat from the water, but in the second case, the heat removed is only 25%, which means that the water temperature in the second case only drops by 50% of the temperature drop in the first case. So, it looks like the efficiency decreased, but the actual heat removed was increased, so technically, the efficiency went up in the exchanger, but there was more heat brought in than could be removed.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[LittleInch]

In summary I guess the answer is actually "Less heat transfer when pumping more hot water thorough the coil next to the one in your room...."

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[MatthewL]

IMHO, it sounds just like what dvd is referring to. Think of the difference in heat transfer when a system goes from nucleate to film boiling or condensation. The heat transfer rate goes up to a peak in the nucleate regime, then starts to fall back until a steady state is reached in the film boiling regime. If the temperature in the coil "tips" the heat transfer into film from nucleate, then the heat transfer rate due will drop and if you extend the still pool approximation* to a flow setting, the drop in heat transfer rate due to boiling regime will actually exceed the increase due to flow, resulting in a net decrease in overall heat transfer, resulting in the cold outlet becoming warmer, even though the flow of heating liquid has been increased.

*In the still pool scenario, a wire is immersed in a fluid and heated with an electric current, starting with the wire at or near the boiling point of the liquid. As the current is increased, (read flow), the fluid around the wire heats and begins to boil, starting with nucleate boiling which increases the heat transfer, but then transitions to film boiling quickly as nucleate boiling is in an unsteady state. This causes the temperature in the wire to heat rapidly, but reduces the amount of heat transferred to the surrounding wire, typically until the thermal overloads trip, since film boiling stabilizes around 1000 F above the boiling point of the liquid. I have seen this demonstrated, and once the current (read flow) is stopped, the system comes back quickly to the quasi-stable side of the boiling curve (nucleate boiling), which allows the current to be restarted.

Matt

Quality, quantity, cost. Pick two.

 

[MisterE2]

Sadly I almost wholeheartedly agreed until I remembered that there is a simple case that can prove the poor pipefitters point, flow over a backwards facing step has exactly this happening, the fluid passes over a step and then forms a weakly recirculating zone, basically heat transfers less inside this area, likewise, if you're getting flow separation, and a lot of it, it will work to insulate rather than assist in heat transfer, see the link or google, "flow over a backwards facing step" or "flow separation"

 

[Compositepro]

And increasing flow will increase the circulation in these "dead zones" and increase heat transfer. I'm sure that with careful design you could come-up with a geometry where heat transfer went down with fluid flow, but let's stay in the real world.
"Air curtains" on doors reduce heat transfer through a door by stopping a strong convective flow (cold air moving one way at the bottom of the doorway and warm air moving the other way at the top).

http://www.airtecnics.com/en/Technology/36/air-curtain-advantages-and-benefits

These effects are not significant in real heat exchangers and discussing them is generally a distraction from solving the actual problem. Many people get too easily distracted.

 

[MisterE2]

Eventually with more fluid flow you would get more heat transfer and a more consistent DT along the boundary. But until then, there will be flow separation, the fluid will be plagued by recirculating zones and vortices building up and detaching from any area that has a deformity or odd geometry. IE flow detachment; on the right side of image, not that intuitive to allowing for additional heat transfer

 

[IRstuff]

A vortex in this picture would improve heat flow, since it allows fluid to circulate over parts that would otherwise have stagnant flow.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[MisterE2]

I'll agree w/ that on the right hand side, it encourages mixing of the entire fluid stream, with fully turbulent flow there's no argument there will be more heat transfer and a fully mixed fluid. I think just past the transition point to turbulent flow, with recirculating zones and weakly recirculating and detaching vortices, it could help to create pockets of lower temperature fluid.

 

[IRstuff]

I would that's not that different than if there was laminar flow at that point. I don't think the issue is with specific points in the flow or exchanger behaving poorly, but rather, it's a question of the consolidated performance of the system from output to input. The transition to turbulent flow, in the totality should be an overall improvement in heat exchange.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[moltenmetal]

Just so that we can put a nail in this ugly thing's head: given a constant temperature water source, a single coil will emit MORE heat the faster you pump water through it. This phenomenon is continuous from laminar through transitional into turbulent flow of the water in the coil. The water exit temperature will increase with increasing flow, increasing the delta T driving force for heat transfer to the surroundings by radiation and convection. In addition, the water-side film coefficient will increase with increasing flow. As the exit temperature increases, the fraction of sensible heat available in the flowing water stream used by the coil will decrease with increasing flow, but the actual amount of heat emitted by the coil will increase. There is of course a limit: the delta T can only get so big, and the film coefficient only so high, within the practical limits of pumping, and of course you will rapidly approach that limit such that big increases in flow actually affect the heat emission rate very little. The water-side film coefficient will rapidly become high enough that it no longer controls the rate of heat transfer as well.

"Loiter time" has nothing to do with it whatsoever. Heat transfer is a RATE phenomenon. The molecules are always moving, vibrating, transferring their vibrational (heat) energy to the molecules of their container. Don't confuse a transient state operation with a continuous one! When people talk about adiabatic processes as processes where there isn't sufficient time for heat transfer to occur, which makes sense when considering a transient state operation, it confuses people a bit I think. In a continuous flow operation, that's a rubbish concept: what they really mean is that there is insufficient AREA for significant heat transfer to occur through. The heat IS being transferred- just not enough of it to affect the temperature of the flowing medium significantly.

 

[FredRosse]

MOLTENMETAL IS CLEARLY CORRECT. In general, higher flow rates, all else being unchanged, will always increase heat transfer. I have had a career of fluid flow and heat transfer for over 45 years, and fundamentals are very clear. I think the first answer, from Compositepro sums things up perfectly as well.