Something I don't understand (SE)

Steve Ebels

Actually one of the many things I don't understand but for this discussion we'll stick to hydronics. Kathy says it's best if I don't wander.........?????

Here's what is puzzling me. I hear folks on this site and supply houses etc. talking about temp drop through a loop of pex or baseboard. The comment that has me in a quandry is that some are saying that if you have a low temp drop(<10*) then you are moving the water too fast to give up any heat. This just doesn't add up in my head. If you have a temp drop of, say 20 with a 1gpm flow that's your typical 10K btu per gallon circulated. With me so far? Now let's up the flow through that hunk of BB to 4gpm and say that we have only a 5* drop. That's the same 10K btu output. Out in the real world would it not be better to have a little more flow and less temp drop? Your loop of BB or pex would then provide the same output at the end as it does at the begining. Just because the temp is at or near the same on the return as it is the supply doesn't mean you aren't transfering the btu's. It means (all other things being equal) that your flow is sufficient to maintain the temp of that loop out to the end. I read a post down the wall a ways where the guy was saying that too much flow would not allow the btu's to escape. That doesn't make sense to me. Am I all wet or what? Where does this idea of fast moving water not giving up its heat come from anyway? If the pipe/pex/fin tube or whatever is higher temp than its surroundings, it has to give up heat even if the fluid is moving 90 MPH.

hot_rod

You are on the right track, Steve

Keep in mind there comes a point where moving the water too fast will cause other problems. Noise, erosion, hammering of zone valves , etc. The number I see most often for hydronics is 4-5 fps. "Below 5 fps for water above 140 degrees" according to the CDA guidelines. Cold water supply up to 8 fps.

The B&G system Sizer is a good way to watch the flow vs velocity relationship.

Slowing flow through hw baseboard can not, and will not increase the heat output, impossible.

This article from Siggy explains it best, for my simple mind. Unfortunately the graphs don't print with the article. As you have observed with baseboard and hw coil sizing charts, the faster the water flow the greater the heat output.


http://www.pmmag.com/CDA/ArticleInformation/features/BNP__Features__Item/0,2379,3760,00.html


hot rod

Steamhead

The exception would be

in a gravity conversion. Velocities there are much lower- usually under 1 FPS- because the pipes are so big. Overpumping one of these results in almost no Delta-T, since the water moves so fast it doesn't have time to exchange the heat.

To Learn More About This Contractor, Click Here to Visit Their Ad in "Find A Contractor"

RB_2

Delta T's

Steve,

Like Hot Rod said...plus

I have an article coming out in the Canadian HPAC Magazine in the New Year, which deals with this topic.

Attached graphic shows what happens to the average temperature as the delta t and the supply fluid temperature changes.

For the purpose of keeping heat transfer calcs simple the average is used. Low delta t's as a result of higher flows raise the average fluid temp. This higher average temperature in the pipes results in a higher delta t to the radiant surface at a given load resulting in a higher surface temperature and thus greater heat transfer to the room.

Increasing the load on the floor beyond the calculated heat loss at maximum load and the fluid delta t will also increase beyond the anticipated delta t....As is the case with start-ups.

One can get similar heat transfer with a 40 deg f delta t, as the 20 deg f design simply by raising the supply temperature. Halving the flow but raising the supply temp by 10 deg f on a baseboard system for example gives you very similar results but with considerable less circ horsepower.

May not mean much on a small playground system but those big boy projects really perform well and with significant savings in all kinds of stuff from cost of pipe to insulation to water treatment.

Remember any nickel, dime, quarter you can reduce from the cost of goods sold in your financial statements as a result of better design goes smack back in your pocket provided you don't lower your price!

Heaven forbid you come across an amateur that doesn't understand how to run a business you can take a partial dip into the fund to give the customer a better, more efficient and cost effective system than your low ball bowler types.

There are so many benefits to using larger delta t's specifically in distribution networks or near boiler piping. Some of the guys using our ZCP's use them for multi zone injection system which means a unit rated for 100,000 btu's at 20 deg f delta t, can act as a five zone injection system for systems as large as 500,000 btu's all with 1" or 1.25" near boiler piping in a space smaller than 28" x 48". When Hot Rod post his next Shop Assembled System Skid , imagine each of those five zone as acting as injection zones for five other systems...if you want I can post a schematic to show you how...just let me know.

I would consider it a privilege to have the chance to show you and any others so many design tricks to make you more competitive and more profitable by using these toys.

And ya know what....even if they don't fit your application (it does happens) you'll at least be able to use the design information for your own systems and business.

Anyways...I'm in the office on Monday...403.236.9560.

If you want to discuss this further give me shout.

Have a great year.

RB

(By the way Hot Rod you might like to read the 22 Immutable Laws of Branding by Al and Laura Ries...you could be selling those skid packs as your retirement hobby!)

Boilerpro

Not quite, me thinks...

Its not that the water can't exchange heat, but that the water races from the inlet tapping to the outlet tapping without moving out into the radiator, either up into the top of the sections on bottom connected rads or out into all the sections when the connections are on the top and bottom on the same side.

Boilerpro

Steve Ebels

There, You said it Steamhead

"The water moves so fast it doesn't have time to exchange the heat." That's what I'm tripped up about. That is physically impossible. The statement can't be right, but I hear so many guys say it, even engineers who should know better. It's either an old wives tale or a myth that has carried down from the dead men.

If a big ole chunk of iron radiator is sitting there at 180*, IT IS producing heat, IT IS taking that heat from the water. Therefore the water is giving up its heat regardless of whether you can measure a delta T or not. It's just that at higher flow rates the delta T becomes small enough that it's hard to measure with a +/-5* thermometer or whatever is in the tool bag.

Make any sense??

Boilerpro

Flow increase often increases heat output..

When you look at flow in items such as copper tube boilers or shell and tube heat exchangers, flow rates must be very high to create turbulence in the piping. Low flow rates result in laminar flow, which allow an insulating layer of calm water to form along the sides of the pipe...reducing heating transfer. I imagine the same would go for baseboard heaters..too slow a flow will eliminate turbulence and reduce heat output. This is backed up by the design charts...heat output is always higher at 4 gpm than at 1 gpm.

Boilerpro

Steamhead

Time is the key

How much time does the water spend in the boiler or the radiator? The more time it's in there, the more heat it can exchange.

Picture the room you are in as a walk-in deep freeze, with a door on each end. If you run in one door and out the other, you won't feel the cold so much (disregarding the air-movement factor) because you weren't in the deep freeze that long. But if you walk slowly thru the room you will feel colder. You lost more heat since you were in there longer.

Boilerpro is also right when he brings up the "shortest-distance-between-two-points" principle. This obviously aggravates the situation.

In my house, the first-floor rads are connected bottom-and-bottom, but the second-floor ones are top-and-bottom on the same side of the rad. All the rads heat better with the smaller circulator. The water spends more time in them.

To Learn More About This Contractor, Click Here to Visit Their Ad in "Find A Contractor"

Steamhead

If you look at

rad runout sizing charts for forced systems, you see how small they are- a 1/2-inch pipe can handle up to 100 square feet EDR! The same 100-square-foot rad in a gravity system would have 1-1/4" or 1-1/2" runouts. The design velocity in a gravity system is much, much lower than that of a forced system.

There is very little force available to circulate the water in a gravity system as compared with forced. So they made the pipes large to keep friction to a bare minimum. This lack of friction is the reason it's so easy to overpump a gravity conversion.

Whether our 100-square-foot rad is fed with 1/2" forced-circulation runouts or 1-1/2" gravity or gravity-converted ones, the optimum flow rate thru the rad is the same. Laminar Flow may be a factor, but I'm sure this is offset by the longer time the water is actually within the rad. This, by the way, is the opposite of the flow rates you need in fin-tube baseboard or coils. The difference is that the cast-iron rad has so much more heat-transfer surface.

I can compare this principle to the long-held misnomer that if you crank the pressure up in a steam system, it will heat faster. We all know the opposite is true. The "less is more" principle also applies, up to a point, to gravity conversions.



To Learn More About This Contractor, Click Here to Visit Their Ad in "Find A Contractor"

Steve Ebels

That

Still doesn't compute to me. If the rad surface temp is a given number, say 180* for example. It's going to be radiating XXX amount of btu's whether there's 1/2 gallon or 10 gallons going through it. The point I'm making is that it's the surface temp of the radiation, (whatever form that takes) that determines the output. The flow rate makes a difference only to the point where the entire surface is the same temp. Beyond that you're wasting energy pumping too much water. Below that point you are not getting the whole rad the same temp and therefore not putting out all the heat that may be needed. I understand the comment that too much flow in an iron rad can make the water just zip right in on end and out the other, especially in a bottom feed, bottom outlet piping arrangement. What I'm talking about is the statement that in any given form of radiation, if the water is moving to fast it won't give up its heat. That don't make any sense to me.

Howard

2 more cents

I'm with you, SE. It always seemed to me that a radiator or section of finned tube will generally give up a little more heat if flow is increased. BTUs per gallon will decrease but gallons per minute makes up for that and a smidge more. An exception would be if excess velocity causes undesireable flow patterns with areas of low surface temperature. I'm trying to remember if this was an issue with cast iron baseboard.

Bob Bona_4

I

hear what Steve is saying, and what HR and Steamhead are too. Here's a thought. Say if you campared it to yourself passing through a door slow vs. fast. That makes sense. Now, what if there was another of you, nerves connected somehow, right behind your "pass" through that door. Like water. Not one part of water goes through, but many parts, each fresh part a renewed temperature.

I guess think of a train (flow) of water through a rad, with every car that passes through a rad a new source of giving off heat. Theoretically, would speed matter in the heat transfer, if the supply temp was held constant?

It always intriques me how we all think differently, and how one person sees things a way others have to think about a diffent way to make sense to them.

Am I making sense?:)

To Learn More About This Contractor, Click Here to Visit Their Ad in "Find A Contractor"

hot_rod

Laminar flow

Your describition is true. As I understand it laminar flow would be nearly impossible in small hydronic jobs. Below Reynolds 2300 is where the flow turns laminar. You're talking a fraction of a gpm to obtain this. I'm not sure you could get a pump to flow this small amount and still overcome the head of the circuit.

I've been told engineers can "play" with flow and Reynolds numbers in large commercial applications when fluid travels long distances, perhaps between buildings, to reduce the heat loss in the piping.

Larry Drake wrote about Reynolds numbers recently, check www.petroretail.net, I think, for this article. Modern Hydronic Heating also goes into this in good detail.

hot rod

To Learn More About This Contractor, Click Here to Visit Their Ad in "Find A Contractor"

Steve Ebels

I see what you're driving at Robert

However there is a Caveat' in there somewhere. As far as lowering the flow to get a higher delta T, producing the same btu output. You can reach a point of no return also.

The heat emitter has to be able to drop the water temp that far in the first place. A panel rad system would do this nicely. Just oversize the rads so they can drop the water temp 30, 40 even 50* instead of 20. Then you could use a very low flow to get the same btu output. The problems come when you are dealing with all the variables a person runs into on a job. Not enough space, "I don't want that big thing in my house" type stuff.

Also bear in mind that raising the system temp drops burner efficiency. Maybe burner eff. is the wrong term but the uptake is that a boiler firing into 130* water is more efficient than one firing into 180.

Everything is a trade-off and has to be taken into account when designing a system.

I still don't understand where the idea that fast moving water doesn't give up as much heat as "slow" water comes from. The heat transfer is the same you just don't see the huge delta T. Maybe it's just a terminology thing and I'm misconstruing what is being said.

I'm going to bed. It's too late to be thinking this hard for someone as feeble minded as myself.

Maybe I'll give you a call Monday. I have a job coming up I'd like to pick your brain on.

Bob Morrison_2

a couple more cents worth: time is not the issue

Can you get more heat (increase the rate of heat transfer) by slowing the flow rate? The answer is "no", practically speaking. (Is that the underlying issue in this discussion?) But you can increase the fluid delta T by pumping a slower flow rate.

For hydronics, the driving forces for heat transfer rate is surface temperature and heat emmitter surface area: Heat gain rate BTUh = UA(Twater-Troom) + Radation Effect. Time is not included in the equation. for a horizontal pipe in a room the contribution from radation and natural convection are roughly equal. On the fluid side, Heat loss rate = 500 x gpm x delta T. To maintain the same heat loss rate, the delta T or gpm can be varied. However, varying the gpm does not affect the heat loss rate significantly.

For example at 4 gpm, typical finned tube radiation output is 610 BTUh/LF, while at 1 gpm the output is 580 BTUh/LF, at 180 degrees entering. The flow rate has minimal affect on the heat transfer (as long as turbulance is maintaned). The slower flow rate will result in greater temperature drop, though.

For residential applications, lower system water temperature provides greater benefit than pumping a slower flow rate, in my opinion.

Bob Morrison

David Efflandt

Similar to car radiator

A perfect analogy is heat transfer of a car radiator. Two people I know thought that removing the thermostat would enable their car to run cooler, in both cases they were wrong. One was a Mercedes 240D diesel used for a summer trip through South Dakota, another was a Taurus SHO used for racing. In both cases removing the thermostat resulted in too much flow and too little heat transfer (overheating). Neither would have overheated with the thermostat in place.

It is a question of balance. Experienced racers that remove their thermostat, replace it with a fixed orifice that keeps the coolant flow within a usable range, fast enough to not boil and slow enough to transfer enough heat.

David Efflandt

Similar to car radiator

A perfect analogy is heat transfer of a car radiator. Two people I know thought that removing the thermostat would enable their car to run cooler, in both cases they were wrong. One was a Mercedes 240D diesel used for a summer trip through South Dakota, another was a Taurus SHO used for racing. In both cases removing the thermostat resulted in too much flow and too little heat transfer (overheating). Neither had any overheating history with the thermostat in place.

It is a question of balance. Experienced racers that remove their thermostat, replace it with a fixed orifice that keeps the coolant flow within a usable range, fast enough to not boil and slow enough to transfer enough heat.

Jeff Andle

heat and velocity...

forget the flow rate for a minute.

Physically heat loss is a result of the temperature difference and temperature drop is a result of heat loss.

That gallon of water gives the same heat transfer per minute at a dead standstill as it does at 100 miles an hour (ignoring pipe turbulence, etc.).

The effect of high flows is only a matter of how long THAT gallon of water is in the loop transfering heat and losing temperature. Balancing this is that at twice the flow rate twice as many of these gallons are transferring heat in the same unit of time.

The diminishing increase of net heat transfer comes about because the average temperature in the loop is higher at higher flow rates. The individual gallons of water loose less heat and drop temperature less but more of them do it and more of the pipe is closer to the source temperature.

Jeff Andle

surface temperature decreases from 180 along the pipe

Assume a 20 degree drop along a 200' circuit at the design flow rate. You are radiating from a 180 temp at the inlet, 160 at the output and around 170 near the midpoint. The cooler surface temperature is the "insulating" layer of heat depleted water in a laminar flow.

Oversimplifying this for twice the flow, the surface temperature drops less (oversimplifying "half" as much) and the outlet is 170, etc...

I think the statement about moving too fast to give up it's heat is an oversimplification of these principles.

To borrow from the walk through freezer, one person walking through the freezer would give up an amount of heat and come out cold. If they ran through in half the time they would not be as cold (and would not have heated the freezer as much), but TWO people running through in half the time would heat it as much (or more).

steve gates

I wonder if

the comments from days of the past come from some long run that gave up it's btu's to early and there was none left at the end, hence a larger pump to get heat at the end of said run?

Howard

This concept is totally losing me. Over X amount of time, 5 pounds of water give up 4 deg F or 20 pounds of water give up 1 deg F. Either way it's 20 BTUs per X time, right? How in the world does BTU output decrease when flow is increased? The car radiator example above goes contrary to what makes sense to me. Doesn't the car thermostat modulate to open on rise to increase flow. Increasing flow had better not reduce heat transfer. (Strange stuff happens when you approach the speed of light but we're talking fractional hp here)
Somebody set me straight--plain english preferred

Arthur

Car Radiator heat transfer

I had an old van once and as I was having problems trying to keep it cool also removed the thermostat and it actually ran hotter still. Then removed the thermostat housing and fitted a lead orifice washer and after that she ran cooler then finally fitted an Air scoop to defect the air up though the rad and that fixed her. I came to the concusltion that the pump was causing caviation and so reducing the flow of water rather than increasing it.
The point about heat loss is to also remember the time factor we measure KW or BTU's over a period (usually an hour, ie KW/hr,BTU/hr) so that the rate of heat loss though a freezer would be same whether you ran or just ambled though, Note I said rate not the amount, If you took 2 min to walk and 1 min to run the heat loss would be 2xbtu's the run amount of (xbtu's) even though the rate was the same. xBTUs/min

Arthur

Re'Something I don't understand

SE, I'm inclinded to agree with you.I guess one of the reasons heating guys have tried to keep the flow and return temps of lower delta t's is that it was not considered good to have too big a temperature drop from the top of the boiler to the bottom very especially with cast iron boilers. I have seen a boiler crack at the back when a teacher set the pump stat at boiler temp so that the boiler got to 180 and then the stat called out to the pump 'time to get up' The stinking hot water in the boiler when down to the school while the freezing 'cold' stuff from the school came back to the 'hot' boiler. Result one great crack in the back section

Mark_4

I think if I'm understanding this right , you need to think of the flow in relation to having several radiators connected, not just one. It seems to me that the guy who walks through the freezer will be much colder walking into the second freezer , the guy who runs through the first , will have "left over" heat to give up!O.K. now my feeble mind is messed up too!