Q=8.01 x D x c x f x delta T [reference Siegenthaler equation 4.8]

Q is heat gain (BTU/Hour)
D is hydronic fluid density (Lbs/CF)
c is hydronic fluid specific heat (BTU/Lb/Deg F)
f is flow rate (GPM)
delta T is difference between supply and return temps (Deg F)

So more flow rate generates more heat gain.

So here is a real world example.
We had a building in Nebraska, Simple 1,500 square foot space, something like 6 300' loops of 1/2 pex. The boiler was sized right, had a 3 speed Taco pump, when the pump was on high the building would not make heat loss. The building never came up to temp.The water was moving to fast, the heat could not get off the train so to speak.
With the pump on low (still slightly over pumped) with a 15 degree delta T the system it would bring the building up to temp. With an analysis of the heat load on loop cad, the required flow was around .4 to .5 GPM, (not much) When the system was slowed down the the boiler system made the required heat load.
I know it is counter intuitive but it is a fact. Faster moving water is not necessarily better. Further buildings are moving targets as they come up to temp there needs change, thus the absolute beauty of using Delta T pumps. Yes there is an added expense, but it pegs the load and will save fuel cost and more.
Dan

Yes, what you are describing Dan is indeed correct and also consistent with the equation. A slower flow rate will result in a larger delta T. A larger delta T generates more heat gain too. So one can achieve the required heat gain by either designing for a fixed delta T or a fixed flow rate. Normally one would DESIGN for fixed delta T and then select the optimal pump (i.e., fixed flow rate and minimal operating electrical power) to provide the required flow rate. The delta T pumps are wonderful devices if $$$ are not an issue because for a fixed flow rate the ACTUAL delta T will change with the ACTUAL heat loss of the floor.

This is true and the appliance can't hold heat it does not have. It must not be given up in the first place. There have been books written, but you won't find them referenced or sold here. I am frustrated by the among of information available and even the number of on-line vendors, which seem to be under utilized.

The bigger the pump the more "potential" for heat transfer, either by increased turbulence or simple pounds/gallon heat transfer. It is true that a greater delta T will increase the heat transfer at a given flow rate, but the temperature gradient across any radiant panel or panels, will also be stretched, increasing the chance of occupant discomfort.

The reason you don't want more boiler or pump than you need is one of simple economics. It is wasteful. The way to avoid this is to get - dare I say - experienced help before picking up the first wrench.

I will say again. Installing dedicated water heaters, be they tank-less or storage type, for radiant floor heating is potentially hazardous in ways that most do not recognize nor appreciate. But hey, I am just "your typical pro who wants everything done...right.

Sorry Dan; your results and conclusions are askance with accepted physics. But I agree that it takes some mind stretching to get it. I would still bet your experience over the guys whom have only read 'THE' book.

Here is another free hint. The mass of any radiant slab will absorb a certain amount of heat energy before effecting any wall-hung thermostat. How long this process takes has everything to do with the ground temperature and insulation below the slab combined with the added load of the structure surrounding the slab, output of the heating appliance and, to a lessor degree, the flow rate of the heat transfer medium. In other words, once you get the heat away from the appliance and into the space, all else is a comfort issue. There is no set and perfect Delta T that fits every radiant floor heating system any more than there is for a poorly designed radiant panel using a water heater as a heat source.

Yes, a radiant slab will not provide any heat gain to the room or affect the thermostat until the surface temp of the radiant slab becomes greater than the room temp. Thermal mass performance can be readily modeled and certainly should be considered when designing integrated hydronic radiant floor and passive solar heating systems. We developed a DIY calculator to help people with this:

Borst Passive Solar Thermal Mass Performance Calculator

a radiant slab will not provide any heat gain to the room or affect the thermostat until the surface temp of the radiant slab becomes greater than the room temp

As soon as the the basement slab starts to warm up (ie, any btu added at all), the heat loss through it goes down and if there is any other heat source (say the floor above), the temperature in the room starts rising. Any colder than the room radiator is absorbing heat and this effect decreases as the radiator approaches room temperature. Starting with the first btu.

So,
when we look at the cost of the delta T pumps they are about $100.00 more per unit over a standard pump.
Agreed added expense, but look at the gain,The design target is improved. The water velocity will decrease as building target temp is acquired.Over pumping is not a solution, it is counter productive. I am guilty of over pumping, using a delta T pump provides a good work around especially if you are providing a DYI system where there will be variables beyond control or expectations. The new Taco Bumble Bee pumps are an ECM motor, 45 watts at max, and will turn down to next to nothing, so there initial cost can be recovered by increasing building performance response, less electrical use, better fuel management as your pump is optimizing what it needs and they last for ever(ceramics). I am just impressed with the option, the additional cost, well I suppose it is client related.
We have been designing around these for some time now. A good radiant system has room in the budget for a few extra dollars to place some smart parts in, out door reset for instance would fall in the same category.
I am of the same mind as Morgan on the whole tank less water heater work around, it may save a few dollars up front, but more often than not they will burn more fuel and not last as long. There are to many good options using a real boiler be it condensing or non.
Dan

I still say when this slab is up to temp it won't matter that he's using a water heater to anyone unless he keeps shutting it off and trying to heat it up from cold again. this heater can handle this heat load with acceptable delta-T. end of story. it's just a startup in the worst possible conditions (near design condition)

Dan, you're dead wrong on being able to pump a circuit so fast that you slow down heat transfer. You will NEVER reduce heat transfer by increasing flow. never, ever, ever. that does not mean that more flow is always better! but too much flow can NEVER be the cause of an underheating problem. If your building suddnely started heating when you slowed down a pump you got hit by coincidence, or there was a truly unique flow issue in your system.

more flow lowers delta-T, so it LOOKS like you aren't transferring heat as fast. but you are. you're pushing more gallons, at lower dt, and moving at least as much and usually more BTUs by raising flow.

I will dig deeper and find the numbers,
But as a basic example, on my solar collectors evac tubes, if I run the water through at a higher flow rate I will recover less BTU in a day. They preform better at a slower rate of flow.
Pumping water at 3 gpm through the panels will result in less heat gain to my storage tank, Pumping water at 1 gpm will increase the gain in my storage tank.
Dan

It might be that it looks like that because the tank is staying more stratified. You can get a smaller amount of hotter water with a lower flow.

Dan. NEVER means NEVER. you might not be able to get a TEMPERATURE you want (lower Delta-T at higher GPM means lower temp rise) but you will ALWAYS increase BTU production/transport by increasing flow, unless you are at a maximum value already, say a heat source's output max. in which case, you will transfer exactly as many BTUs at the higher flow rate, at a lower delta-t.

you cannot slow down solar collection by speeding up a pump. all you can do is reduce the temperature rise across the panels, because essentially you are diluting the energy in more water. which you're doing anyway once it gets back to the tank. the only problems with overpumping are not BTU related: energy use, noise, erosion, etc. so it may not help enough to justify raising the flow.. in fact in most cases it won't, which is why I design very low energy low flow systems... but there are times that it will.

sanity check yourself dan. if you go slow through your array, say, in at 100 and out at 120, average temp is 110. do you think you will get more, or less solar BTUs into the water if you double the flow, send it in 100 and take it back at 110, average fluid temp 105?

Ok,
so I have talked and looked harder at this.
We are actually talking about
Laminar value
Turbulent value http://www.efm.leeds.ac.uk/CIVE/CIVE1400/Section4/laminar_turbulent.htm
Reynolds number http://en.wikipedia.org/wiki/Reynolds_number

U tube video http://www.youtube.com/watch?feature=endscreen&v=eIHVh3cIujU&NR=1
look at the 10:00-15:00 section
Look at 21:49 section this is the real image to see.

In essence you are correct BTU load will increase with flow, but again the ability for the heat exchange is diminished as Turbulent value increase.
It is a relationship between the Laminar value and the turbulent value.
If the water flows slowly absent of turbulence it has a clean form across the entire core of the pipe. When the turbulence increases due to velocity the laminar wall is pushed to the outside of the pipe, the heat exchange between the turbulent section (core) of the pipe and the laminar section (outer wall) is diminished. Your BTU loaded in the turbulent area is not transferring to the Laminar area efficiently.
I am not a silence guy, just the messenger.
Not arguing that BTU is not increased with flow, just that BTU is not as accessible it would be with out turbulence. Over pumping = Turbulence= diminished heat transfer
Delta T control combined with proper pipe sizing, will help minimize turbulence.
That is all I am saying,
Dan

you do not have that correct either. turbulent is better for heat transfer. a laminar flow layer is, technically, insulative, the opposite of enhancing conduction. not by enough that in a residential radiant floor anyone should care, but that's the fact. that's an argument our new PE friend and I were having in the other thread the other day.

that is not the core of the discussion we were just having, but it is one more point in favor of "you cannot diminish heat transfer by overpumping". you can certainly waste energy, but you cannot reduce heat transfer in any circumstance by over pumping.

the core of the discussion we were just having is that it is impossible to slow down heat transfer over time by raising flow. there is no case where you were moving or transferring X BTUS at Y flow, and would move less than X BTUs at some flow greater than Y.

Actually...more flow rate=more velocity=larger Reynolds number=more turbulence=more heat transfer=more BTU heat gain.

More turbulence breaks up the laminar boundary layer near the wall that inhibits the heat transfer. The flow transitions from laminer flow to turbulent flow between 2300 and 4000 Reynolds number. However, once you have transitioned to turbulent flow, increasing the flow rate does NOT increase the heat transfer any further. However, increasing the flow rate will continue to increase the BTU heat gain per equation 4.8 because you are simply conveying more heat into the slab by virtue of the higher heat conveyance rate (i.e., flow rate).

The delta T in equation 4.8 (or just in general) can also be calculated very accurately by only knowing the flow rate, slab surface temp, indoor room temp, and slab heat transfer coefficient (which can be calculated by knowing the slab thickness and tube spacing).

I am very happy to finally see a discussion about Reynolds number and laminar/turbulent flow on this forum...I was starting to come to the opinion that no one on this forum would let physics or math prevent them from designing a system  

Did you bother looking at the video?
It counters what you just said, and was prepared by Iowa Institute of Hydraulic research.
They counter your point with science. The turbulent core rides past the the laminar wall.
All that bubble water whips by.

Dan, this is basic hydronics. laminar flow = lower heat transfer as the boundary layer insulates. turbulent flow is more efficient at transferring heat to the wall of the pipe.

laminar is more effective for transit... i.e. the water is pushed more easily, less resistance. but HEAT TRANSFER to the wall of the pipe is reduced.

As one can see here , heat transfer keeps increasing with increasing Reynold's number. Ie, increasing flow rate keeps increasing the heat transfer (liquid to pipe), even after you achieve turbulent flow. If you want to visualize it, the turbulence keeps making the laminar boundary layer thinner.

Posted By Blueridgecompany.com on 08 Jan 2013 02:49 PM
Did you bother looking at the video?
It counters what you just said, and was prepared by Iowa Institute of Hydraulic research.
They counter your point with science. The turbulent core rides past the the laminar wall.
All that bubble water whips by.

Been imbibing any o' that bubble water?  (Can I have some?)

Seriously- I never read "The Book" on hydronics, but I do have a degree in physics. Yer dead rong on this- it seems you've turned it on it's head! Turbulence is YOUR FRIEND when it comes to increasing the heat transfer rates, just as everyone keeps saying. Higher velocity results in reduced the heat transferred per gpm of flow, but the overall heat transfer increases monotonically (and far from linearly) increasing with flow.

But overpumping is a waste of power here. (The heat from the power used by pump itself is a measurable fraction of the heat being delivered to the room!)  When the radiation is massive and the heat source is limited, raising and lowering the temp of that mass simply takes time.  If the slab is uninsulated (or inadequately insulated) that only adds to the response time.

Overpumping carries some wear & tear costs on the boiler too, but those costs are comparatively low when the "boiler" is a tank type hot water heater.  The bigger issue with non-condensing tank type hot water heaters is sub-100F return water temps from the slab resulting in copious destructive condensation on the  flue & heat exchanger, not wear from the excessive flow.

Once the flow completely transitions from laminar to fully turbulent flow (at a Reynolds number of about 4000), the heat transfer coefficient essentially stair steps up and flatlines. However, the BTU heat gain keeps increasing with increasing flow rate per equation 4.8. Yes, the Reynolds number also keeps increasing with increasing flow rate, but the heat transfer coefficient essentially stops increasing once you have transistioned to fully turbulent flow...at least to the degree that one could measure any perceptable heat transfer coefficient change. Once you exceed a Reynolds number of about 4000, it is purely the increased flow rate (not the increased Reynolds number) that is resonsible for the increased BTU heat gain.

Interesting , lets get back to the real world.We are talking about residential radiant heating.Too add to oversizing pump and velocities, besides the waste of energy, fluid velocities can cause piping failures from internal erosion and this should be avoided at all costs.