Cast iron radiator at high flow rate

hot_rod

I've been playing around with cast radiators at different flow rates.

Also checking output with low SWT. I ran this one today at 8 gpm, 1" ID hose.
About 90 SWT.

I will compare heat up time at 1 and 4 gpm, 1/2 and 3/4 supply for example, as a for comparison.

Connected supply and return on the bottom connections. Within 20 minutes it was pretty consistently warm. After an hour delta was down to .6°

Even at these high flow rates across the bottom, the upper part of the sections get as warm, as the lower part.

Steamhead

An Italian Flue radiator! ISTR the internal passages of these are huge. Let's see the 1 and 4 GPM ones.

EdTheHeaterMan

I think we should send @ethicalpaul over there with his video camera, Bob. Between the two of you liking to test theories out and Paul's ability to tell the story on camera, you two can start a Norm Abrams, Bob Villa type show called "This Old Radiator" on the HeatingHelp channel. You guys will need a theme song.

https://www.youtube.com/watch?v=2pZXjb6aCy8

delcrossv

As I'm presently specing emitters for a large HW project,  I'm finding this pretty fascinating. 
Keep it up.

How noisy is it at those flow rates?

GGross

Thanks for running this test bob. I honestly haven't stopped thinking about that claim I saw here regarding cast rads since I first saw it. That one was pretty unique in that someone was running 120 GPM where 40GPM should have been the max, though the claim was that this was an easy mistake to make which I thoroughly disagree with. Any chance you have a really high flow pump just laying around there to test out?

hot_rod

Second run at .8 gpm

My take away both at .8 and 8 gpm flow the radiators heated evenly, bottom to top.
One thought was the higher, 8 gpm flow would just blast across the bottom and no temperature rising into the upper portion of the sections. I've done a similar test on 3 different cast iron radiators and had the same result. All the sections warm evenly with high or low flows.

The 8 gpm test showed 6400 BTU/hr output
The .8 gpm test showed 3700 BTU/hr output

The output difference from the BTU meter, the hydronic formula, and the pic indicate a higher output at the higher flow rate.
In both runs the gpm flow rate was a bit higher at the end, the .8 gpm jumped to 1 gpm. Possibly as the water warmed it changed? So the math and BTU meter readout are off by a few hundred btu.
I didn't correct for density or specific heat of the fluid at 90° either.

In this example, a single radiator, 30' 0f 1" piping from the boiler, a basic 15-58 is adequate for the 8 gpm test, so @70W. I used a flowsetter to choke down to .8 gpm

I have a Alpha 15-58 ordered, I'm curious to see the Watts at the 8 gpm flow.

Is the higher BTU output worth the extra gpm??

It required 1" piping to supply 8 gpm, and higher pump speed to eek out the additional 2700 BTU/hr

If a system required a 200W high head circ, I don't think the juice would be worth the squeeze.

The .8 gpm could be accomplished with 1/2 tube, possibly pex depending on the circuit length.

So, yes, output increases with higher flow, but starts to flattens out at some point. If these radiator flows were graphed I'd expect the curve to look like these examples from Idronics 23.
Fig 2-13 is an air handler coil, Fig 2-14 is a pex radiant loop.

@delcrossv no velocity noise in this radiator, it has 2" tappings, pretty wide open across the bottom sections.


One correction from the original post pic, ending ∆T on the 8 gpm test was 1.6, not .6

GroundUp

Cool test! I'm curious though, what are you using to measure flow? I've found that most radiant manifolds tend to run very close to each other, assuming they're all fairly accurate, but then the Quicksetter in the same system might read very differently. In doing the head loss calcs compared to the pump curves, I've determined that the Quicksetter is not very accurate. I've got one such system at home where the Quicksetter reads about 40% higher than both the radiant manifolds and the curve calculation, and I've got another one at a friend's shop with the same manifolds where the Quicksetter reads slightly more than half of what the manifolds show.

mattmia2

So you're using a turbine to measure flow, essentially a water meter.

hot_rod

It is the Caleffi Conteca BTU, or energy meter.
BTU meters need to be tested and certified to an EN standard, or the new ASTM E3137 standard.

PeteA

@hot_rod can you test it with both the inlet and outlet on the same side. In my house nearly every radiator has the hot come in on the top and the cooler water exit the bottom but both taps are on the same side.
I've been considering getting a slightly larger pump to move a little more water through my system.
i currently have the 007e but it takes a little while to get the heat to come up through the rads. 

I put in the boiler protection valve and I’m now getting the water up to about 130 in the boiler in 3-4 minutes but I never see more than around 100 on the return manifold before the thermostat gets satisfied at 69 and shuts down.
my next move is going to be zone valves but that won’t be until spring/summer which may make the idea of a bigger pump moot.

mattmia2

If the t-stat is satisfied at 100 then you don't want it hotter than that or it will overshoot when that additional heat is radiated in to the building.

Steamhead

@hot_rod , the place I've seen over-pumped rads is in converted gravity systems. These have larger pipes feeding the rads than newer systems do, since the original motive force acting on the water is so small. Over-pumping one of these can, and does, cause the water to short-circuit through the rads. Here is an extreme example (note that we eventually replaced the original boiler some years ago):

https://www.heatinghelp.com/systems-help-center/adjusting-the-flow-rate-for-an-old-gravity-hot-water-system/

which led to further research that produced this:

https://www.heatinghelp.com/systems-help-center/sizing-circulators-for-old-gravity-hot-water-heating-systems/

WMno57

Excellent article By John Siegenthaler here:
https://www.hpacmag.com/features/is-the-water-moving-too-fast/

One question about hydronic heating that pops up over and over again is this: If the water moves too fast through a hydronic circuit, will the heat it holds be unable to “jump off” as the stream passes through a heat emitter? The answer, from the standpoint of heat transfer alone, is no. But rather than just take my word for it, let’s see why this is true.

WMno57

Higher flow rates have greater heat transfer. Period. When heat transfer drops off in a boiler, the problem is not too high of a flow rate, but steam pockets. Heat transfer does not occur to steam. You must have liquid water to have heat transfer. This is true for boilers, water cooled engines, and nuclear reactors.
Steam pockets will form if pressure drops lower than what is required to maintain liquid water. That might be higher that 212F. The more heat you are making, the more pressure you need to keep the water (coolant) liquid.
Steamhead's case study from 2014:
https://www.heatinghelp.com/systems-help-center/adjusting-the-flow-rate-for-an-old-gravity-hot-water-system/

The system had been"modernized" sometime in the 1940's. An Esso oil burner, firing 3.5 GPH, replaced the coal grates, the open expansion tank in the attic was replaced with a basement compression tank, and a circulator was mounted on the return line. At some point a Paracoil "side-arm" tankless heater was added also.

The problems started when the Esso burner was replaced with a Beckett SF. The oil company did not reduce the firing rate to compensate for the hotter flame the Beckett produces. The boiler began behaving very strangely. When it ran for a while, the pressure would suddenly begin rising very quickly. When it got to 30 pounds, the safety valve would blow and blow. And if the process went on for a while, the boiler would leak at the rear. The lady almost choked when I told her what a new boiler would cost, so I added some sealer which stopped the leaking.

I had the lady tell the oil company to reduce the nozzle size to 3 GPH and re-tune the burner to match the new nozzle. That helped, but did not solve the problem. Adding compression tank capacity got things under control, but it still didn't feel right. Massive waves of intense heat rolled off that boiler when it was running. The Beckett roared and the circulator hummed away, but the house took forever to get warm. She wasn't getting her money's worth from that 3 GPH.

When the circulator was put in, they used a 3-inch Thrush. The original 3-inch Thrush would pump 70 GPM at the very low head pressure in an old gravity system. At some point the Thrush wore out and was replaced with a 3-inch B&G. Here was part of the problem - the B&G 3-inch circ would pump 115 GPM! The water was moving through that system so fast it couldn't pick up enough heat from the boiler, or shed it in the radiators. No wonder the boiler metal overheated and leaked at the section joints. And I'll bet the sudden pressure increase was caused by bubbles of steam forming in there, because the metal was so hot.

The steam pockets were probably due to the boiler being over fired. Not too high of a flow rate.

WMno57

Too high of a flow rate can contribute to steam pockets if you are pumping away. Because you are creating a lower pressure zone inside the boiler.
This also shows up as cavitation in pumps.
https://en.wikipedia.org/wiki/Cavitation#Pumps_and_propellers
But the root problem is not flow rate. The root problem is steam pockets.

PeteA

The info given here on this site is amazing!! You guys are all true masters of your crafts not just the answers you provide but all of the supporting info that ties it all together. 

Steamhead

WMno57 said:

Higher flow rates have greater heat transfer. Period. When heat transfer drops off in a boiler, the problem is not too high of a flow rate, but steam pockets. Heat transfer does not occur to steam. You must have liquid water to have heat transfer. This is true for boilers, water cooled engines, and nuclear reactors.
Steam pockets will form if pressure drops lower than what is required to maintain liquid water. That might be higher that 212F. The more heat you are making, the more pressure you need to keep the water (coolant) liquid.
Steamhead's case study from 2014:
https://www.heatinghelp.com/systems-help-center/adjusting-the-flow-rate-for-an-old-gravity-hot-water-system/

The system had been"modernized" sometime in the 1940's. An Esso oil burner, firing 3.5 GPH, replaced the coal grates, the open expansion tank in the attic was replaced with a basement compression tank, and a circulator was mounted on the return line. At some point a Paracoil "side-arm" tankless heater was added also.

The problems started when the Esso burner was replaced with a Beckett SF. The oil company did not reduce the firing rate to compensate for the hotter flame the Beckett produces. The boiler began behaving very strangely. When it ran for a while, the pressure would suddenly begin rising very quickly. When it got to 30 pounds, the safety valve would blow and blow. And if the process went on for a while, the boiler would leak at the rear. The lady almost choked when I told her what a new boiler would cost, so I added some sealer which stopped the leaking.

I had the lady tell the oil company to reduce the nozzle size to 3 GPH and re-tune the burner to match the new nozzle. That helped, but did not solve the problem. Adding compression tank capacity got things under control, but it still didn't feel right. Massive waves of intense heat rolled off that boiler when it was running. The Beckett roared and the circulator hummed away, but the house took forever to get warm. She wasn't getting her money's worth from that 3 GPH.

When the circulator was put in, they used a 3-inch Thrush. The original 3-inch Thrush would pump 70 GPM at the very low head pressure in an old gravity system. At some point the Thrush wore out and was replaced with a 3-inch B&G. Here was part of the problem - the B&G 3-inch circ would pump 115 GPM! The water was moving through that system so fast it couldn't pick up enough heat from the boiler, or shed it in the radiators. No wonder the boiler metal overheated and leaked at the section joints. And I'll bet the sudden pressure increase was caused by bubbles of steam forming in there, because the metal was so hot.

The steam pockets were probably due to the boiler being over fired. Not too high of a flow rate.

Actually, @WMno57 , the firing rate was within the boiler's capacity.

It took me a while to figure this out too. Everything looked right "by the book". But, no matter how good the book is, it can't cover everything, can it? Bottom line, reducing the flow rate and correcting the unequal distribution thru the boiler solved the problem. And the oil company wasn't happy with the reduced oil consumption.

With regard to @hot_rod 's Italian Flue radiator, assuming it is 38 inches tall, the chart here:

https://www.heatinghelp.com/systems-help-center/american-radiator-company-catalog-1897/

shows its rating to be 56 square feet. In many gravity systems, a rad of this size was piped with 1-1/4", not 1". It would be interesting to see how it does when piped this way.

hot_rod

It is the Caleffi Conteca BTU, or energu meter.

WMno57 said:

Higher flow rates have greater heat transfer. Period. When heat transfer drops off in a boiler, the problem is not too high of a flow rate, but steam pockets. Heat transfer does not occur to steam. You must have liquid water to have heat transfer. This is true for boilers, water cooled engines, and nuclear reactors.
Steam pockets will form if pressure drops lower than what is required to maintain liquid water. That might be higher that 212F. The more heat you are making, the more pressure you need to keep the water (coolant) liquid.
Steamhead's case study from 2014:
https://www.heatinghelp.com/systems-help-center/adjusting-the-flow-rate-for-an-old-gravity-hot-water-system/

The system had been"modernized" sometime in the 1940's. An Esso oil burner, firing 3.5 GPH, replaced the coal grates, the open expansion tank in the attic was replaced with a basement compression tank, and a circulator was mounted on the return line. At some point a Paracoil "side-arm" tankless heater was added also.

The problems started when the Esso burner was replaced with a Beckett SF. The oil company did not reduce the firing rate to compensate for the hotter flame the Beckett produces. The boiler began behaving very strangely. When it ran for a while, the pressure would suddenly begin rising very quickly. When it got to 30 pounds, the safety valve would blow and blow. And if the process went on for a while, the boiler would leak at the rear. The lady almost choked when I told her what a new boiler would cost, so I added some sealer which stopped the leaking.

I had the lady tell the oil company to reduce the nozzle size to 3 GPH and re-tune the burner to match the new nozzle. That helped, but did not solve the problem. Adding compression tank capacity got things under control, but it still didn't feel right. Massive waves of intense heat rolled off that boiler when it was running. The Beckett roared and the circulator hummed away, but the house took forever to get warm. She wasn't getting her money's worth from that 3 GPH.

When the circulator was put in, they used a 3-inch Thrush. The original 3-inch Thrush would pump 70 GPM at the very low head pressure in an old gravity system. At some point the Thrush wore out and was replaced with a 3-inch B&G. Here was part of the problem - the B&G 3-inch circ would pump 115 GPM! The water was moving through that system so fast it couldn't pick up enough heat from the boiler, or shed it in the radiators. No wonder the boiler metal overheated and leaked at the section joints. And I'll bet the sudden pressure increase was caused by bubbles of steam forming in there, because the metal was so hot.

The steam pockets were probably due to the boiler being over fired. Not too high of a flow rate.

All I can comment on is theses actual tests. I took the delta tee down to 1.6 and it was still transferring energy out of the radiator, as observed with the camera and showing on the BTU meter

I wish I had a building like the one you worked on to add a BTU meter and experiment with. It sounds like a supersized example of what I just demonstrated. I”ll see if I can get a Magna into the loop and take it to 20 gpm, just for grins🔧

Also to add a btu meter on a gravity system to see the actual flow movement velocity in feet per minute. To move 700,000 btu out of a boiler and into a radiator there must be some flow. Or as Ray suggests you kill the boiler with inadequate flow??

As @WMno57 added, something else may be going on with the heat transfer.
115 gpm in 3” schedule 40 pipe is just under 3 fps, not that high of a velocity. At 70 gpm, around 1.76 fps.

500 X 70gpm X 20 delta= 700,000 btu/ hr
500 X 115 X 20= 1,100,000 btu/ hr

Any idea of the heat load in that old building?

Steamhead

hot_rod said:

It is the Caleffi Conteca BTU, or energu meter.........

All I can comment on is theses actual tests. I took the delta tee down to 1.6 and it was still transferring energy out of the radiator, as observed with the camera and showing on the BTU meter

I wish I had a building like the one you worked on to add a BTU meter and experiment with. It sounds like a supersized example of what I just demonstrated. I”ll see if I can get a Magna into the loop and take it to 20 gpm, just for grins🔧

Also to add a btu meter on a gravity system to see the actual flow movement velocity in feet per minute. To move 700,000 btu out of a boiler and into a radiator there must be some flow. Or as Ray suggests you kill the boiler with inadequate flow??

As @WMno57 added, something else may be going on with the heat transfer.
115 gpm in 3” schedule 40 pipe is just under 3 fps, not that high of a velocity. At 70 gpm, around 1.76 fps.

500 X 70gpm X 20 delta= 700,000 btu/ hr
500 X 115 X 20= 1,100,000 btu/ hr

Any idea of the heat load in that old building?

In a gravity system, the flow rate would be less than that. Not sure by how much. Which is exactly the point- with gravity piping, it's way too easy to over-pump. The resistance to flow is so much less.

Not sure I still have the heat-loss calc we did, but the replacement boiler, a Solaia SL-7175, has a Net rating of 181,800 BTUH.

WMno57

Steamhead said:

Bottom line, reducing the flow rate and correcting the unequal distribution thru the boiler solved the problem.

Two changes, Flow Rate, and Unequal Distribution (internally to the boiler).
My theory is the problem was solved by fixing the unequal distribution, and down firing.
I don't believe increasing flow rate causes "short circuiting" in a piping or hydronic distribution network. I think increasing flow rate in a hydronic system, equally increases flow in all parts of that system. I don't believe there is a tipping point or point of no return where beyond a certain flow rate, flow is reduced in restricted parts of the system.
Maybe I'm wrong. I'm not understanding the unequal distribution in the boiler. Are you saying the unequal distribution was corrected by reducing flow? Could the unequal distribution have been caused by a steam pocket? Down firing could have eliminated the steam pocket, and restored flow.
Sigenthaler gives several reasons why over pumping is to be avoided. But reduced heat transfer (both at the boiler end and the emitter end) is not one of them.

hot_rod

It is the Caleffi Conteca BTU,

Steamhead said:

hot_rod said:

It is the Caleffi Conteca BTU, or energu meter.........

All I can comment on is theses actual tests. I took the delta tee down to 1.6 and it was still transferring energy out of the radiator, as observed with the camera and showing on the BTU meter

I wish I had a building like the one you worked on to add a BTU meter and experiment with. It sounds like a supersized example of what I just demonstrated. I”ll see if I can get a Magna into the loop and take it to 20 gpm, just for grins🔧

Also to add a btu meter on a gravity system to see the actual flow movement velocity in feet per minute. To move 700,000 btu out of a boiler and into a radiator there must be some flow. Or as Ray suggests you kill the boiler with inadequate flow??

As @WMno57 added, something else may be going on with the heat transfer.
115 gpm in 3” schedule 40 pipe is just under 3 fps, not that high of a velocity. At 70 gpm, around 1.76 fps.

500 X 70gpm X 20 delta= 700,000 btu/ hr
500 X 115 X 20= 1,100,000 btu/ hr

Any idea of the heat load in that old building?

In a gravity system, the flow rate would be less than that. Not sure by how much. Which is exactly the point- with gravity piping, it's way too easy to over-pump. The resistance to flow is so much less.

Not sure I still have the heat-loss calc we did, but the replacement boiler, a Solaia SL-7175, has a Net rating of 181,800 BTUH.

I wonder if the dead men knew or calculated flow rates, certainly there was some reasoning behind 3” vs 2 or 4” pipe? No doubt the gravity systems worked. Air removal must have been trickery with any horizontal piping at all. Wasn’t that the motivation behind Gil Carlsons application of circulators in gravity systems. Certainly some of them didn’t perform as expected😗

The flow can never be 0 gpm, all heat transfer will stop at 0 gpm flow, regardless of the driver

WMno57

As some of you know, I'm not a hydronics guy, I'm a car guy.
Years ago, an old wives tale amongst race car engine builders was that if a thermostat is removed from an engine, it must be replaced by a restricting orifice. Sometimes this orifice is a gutted thermostat. The street knowledge for doing this was if the coolant flows too fast, it doesn't stay in the radiator long enough to cool off.
The thinking now is the cooling problems were due to cavitation of the water pump. The thermostat or restricting orifice keeps enough pressure on the water pump impeller to prevent cavitation. Once you have cavitation you get less flow. So the cooling problems were never due to the water going through the radiator too fast.

WMno57

@Steamhead
OK, I read through your case study again. I think I answered my question about unequal distribution, and how you fixed it.
I was having a problem visualizing the two sided boiler. So one side had lower flow and was getting too hot? You increased the flow to the hot side, and that resolved the problem?

hot_rod

WMno57 said:

As some of you know, I'm not a hydronics guy, I'm a car guy.
Years ago, an old wives tale amongst race car engine builders was that if a thermostat is removed from an engine, it must be replaced by a restricting orifice. Sometimes this orifice is a gutted thermostat. The street knowledge for doing this was if the coolant flows too fast, it doesn't stay in the radiator long enough to cool off.
The thinking now is the cooling problems were due to cavitation of the water pump. The thermostat or restricting orifice keeps enough pressure on the water pump impeller to prevent cavitation. Once you have cavitation you get less flow. So the cooling problems were never due to the water going through the radiator too fast.

That seems like a more logical, and provable concept, less flow= less heat transfer. But good to ask the questions, look for other answers or causes.

EBEBRATT-Ed

Who can predict the water flow through a given piece of equipment without testing?

We here conflicting statements which don't always reflect the btu/water formula

Most say laminar flow or lazy flow reduces the heat transfer and turbulent flow increases the heat transfer.

Boiler and radiation do not act the same. We all learned in school that heat goes from "where it is to where it aint"

In a boiler the water is heated and the heat is transferred to the water through the cast iron.

Water in the radiation is transferred from the water to the radiation

Steamhead

WMno57 said:

@Steamhead
OK, I read through your case study again. I think I answered my question about unequal distribution, and how you fixed it.
I was having a problem visualizing the two sided boiler. So one side had lower flow and was getting too hot? You increased the flow to the hot side, and that resolved the problem?

No, the opposite. One side had way more flow, and that was where it was overheating and leaking- right near the return connection.

As a car guy, I'm sure you understand "dwell" in old-school breaker-point ignition. You need a certain amount of dwell to build up the magnetic field in the ignition coil, so when the points open and the field collapses, there will be enough energy in the "collapse" to generate a spark.

Same concept here. If the water is racing through the boiler fast enough, it won't stay in the boiler long enough to pick up the heat. Not enough dwell, as it were. In this case, the wildly oversized circ and the unequal distribution combine to produce this result.

@hot_rod , we'd have to locate some design info that was used to design gravity systems. I'm sure it's out there.

Oh, and having had that experience, I've been able to solve some other systems' problems. One that comes to mind was a converted gravity system with ~700 square feet of radiation. The old B&G 100 circ had been replaced with a Taco 0010, which has almost exactly the same performance curve. The lady complained that the bedroom was too hot, so I helped her turn the shutoff valve clockwise. But the room got hotter!

I'm sure you all see what the problem was now- reducing the flow made the water distribute through all of the radiator, which increased its output.

I replaced the 0010 with a 007, opened the valve, and the problem went away.

Remember, when dealing with a gravity conversion, you have to think like the Dead Men did, and try to mimic the flow they designed the system for.

WMno57

Very interesting. Thanks Steamhead.
I guess its back to the drawing boards. Got to come up with an explanation that squares with both what you observed, and what Siegenthaler is saying.
I did find this:
https://lambdageeks.com/mass-flow-rate-and-heat-transfer/
Way over my head, but it seems to mirror what Siegenthaler is saying.

WMno57

From the automotive world:
https://shop.championcooling.com/articles/high-flow-water-pump

A high flow, or high volume water pump will help increase flow, helping to move coolant faster. One manufacturer of high flow water pumps is Edelbrock, so we reached out to our friend Smitty Smith to get the low down on why and how these pumps work.

"Generally speaking, higher flow and higher pressure is beneficial to the cooling system," he told us. "Higher pressure in particular raises the boiling point of the fluid, thus decreasing the likelihood of vapor forming in the coolant. This results in better cooling, rounder cylinders, cooler valves, etc. Higher flow rates (up to a point) will increase the heat rejection from the radiator."

With all of these benefits, it made us wonder if these high performance pumps will help with overheating. Smitty said, "If the problem is not coolant flow related, a water pump upgrade will not improve performance significantly. Meaning, if you have a poor fan setup, adding a better water pump will not fix the problem." But he also says, "The idea that too much water flow causes overheating is a myth." That means that having a high flow or high volume pump doesn't cause the water to pass through the radiator too quickly for it to be properly cooled.

Smitty continued, "There is a whole lot of science behind heat exchangers, but the bottom line is that more flow will never cause a decrease in cooling performance.

Steamhead

WMno57 said:

Very interesting. Thanks Steamhead.
I guess its back to the drawing boards. Got to come up with an explanation that squares with both what you observed, and what Siegenthaler is saying.

I don't have Siegenthaler's book in front of me. How much space does he devote to designing gravity systems?

WMno57

I don't have his book either.
I'm just going off his HPAC magazine article.
https://www.hpacmag.com/features/is-the-water-moving-too-fast/

hot_rod

Steamhead said:

WMno57 said:

@Steamhead
OK, I read through your case study again. I think I answered my question about unequal distribution, and how you fixed it.
I was having a problem visualizing the two sided boiler. So one side had lower flow and was getting too hot? You increased the flow to the hot side, and that resolved the problem?

No, the opposite. One side had way more flow, and that was where it was overheating and leaking- right near the return connection.

As a car guy, I'm sure you understand "dwell" in old-school breaker-point ignition. You need a certain amount of dwell to build up the magnetic field in the ignition coil, so when the points open and the field collapses, there will be enough energy in the "collapse" to generate a spark.

Same concept here. If the water is racing through the boiler fast enough, it won't stay in the boiler long enough to pick up the heat. Not enough dwell, as it were. In this case, the wildly oversized circ and the unequal distribution combine to produce this result.

@hot_rod , we'd have to locate some design info that was used to design gravity systems. I'm sure it's out there.

Oh, and having had that experience, I've been able to solve some other systems' problems. One that comes to mind was a converted gravity system with ~700 square feet of radiation. The old B&G 100 circ had been replaced with a Taco 0010, which has almost exactly the same performance curve. The lady complained that the bedroom was too hot, so I helped her turn the shutoff valve clockwise. But the room got hotter!

I'm sure you all see what the problem was now- reducing the flow made the water distribute through all of the radiator, which increased its output.

I replaced the 0010 with a 007, opened the valve, and the problem went away.

Remember, when dealing with a gravity conversion, you have to think like the Dead Men did, and try to mimic the flow they designed the system for.

So I could move 700,000 btu/hr out of a gravity boiler, into radiators with 1 or a few gpm of flow? Since slow is better for some reason only in gravity systems?? doesn't add up. And without design numbers it is speculation. IMO.

hot_rod

WMno57 said:

I don't have his book either.
I'm just going off his HPAC magazine article.
https://www.hpacmag.com/features/is-the-water-moving-too-fast/

Over the years I flow tested radiators, radiant fin tube, air coils. One common denominator is all see an increase in output at higher flows. Certainly no reason to flow above 5 fps in the piping. So here are typical flow rates in copper tube with velocity numbers.

Steamhead

hot_rod said:

......So I could move 700,000 btu/hr out of a gravity boiler, into radiators with 1 or a few gpm of flow? Since slow is better for some reason only in gravity systems?? doesn't add up. And without design numbers it is speculation. IMO.

Think of it this way: Slow was all you had in the gravity era. Your motive head was the difference between warmer water and cooler water. And that's not much. With such small motive head, velocity might be as low as 1 FPS or less. The more BTUs you had to move, the bigger the pipes you'd need, to keep frictional resistance down.

We're so circulator-oriented now that it's hard to picture this. But to the Dead Men, it was "how it was done". And their stuff worked.

In the system I wrote about, there were two sets of mains coming off the boiler. IIRC one was 3-1/2" and the other was 3". The supply connections to the boiler had not been changed- only the returns. So that was the only place that encountered maximum flow rates. We ended up with 45 GPM, which gave us a ΔT of 20° or so. And it worked.

WMno57

More food for thought. Automotive radiators can't short circuit. Two types, down flow and cross-flow. In both types the coolant must pass through the core.


https://www.speedwaymotors.com/the-toolbox/differences-between-downflow-and-crossflow-radiators/28799

In a traditional CI household radiator the water can short circuit across the bottom, but eventually the heat migrates throughout. Hotrod's pictures show this heat migration occurring from left to right over time.

hot_rod

Steamhead said:

hot_rod said:

......So I could move 700,000 btu/hr out of a gravity boiler, into radiators with 1 or a few gpm of flow? Since slow is better for some reason only in gravity systems?? doesn't add up. And without design numbers it is speculation. IMO.

Think of it this way: Slow was all you had in the gravity era. Your motive head was the difference between warmer water and cooler water. And that's not much. With such small motive head, velocity might be as low as 1 FPS or less. The more BTUs you had to move, the bigger the pipes you'd need, to keep frictional resistance down.

We're so circulator-oriented now that it's hard to picture this. But to the Dead Men, it was "how it was done". And their stuff worked.

In the system I wrote about, there were two sets of mains coming off the boiler. IIRC one was 3-1/2" and the other was 3". The supply connections to the boiler had not been changed- only the returns. So that was the only place that encountered maximum flow rates. We ended up with 45 GPM, which gave us a ΔT of 20° or so. And it worked.

A gravity system would be considered a thermo-siphon I assume? What I understand is the vertical and the temperature difference drives the speed of the fluid. A 10 story building with 190 SWT, 68F falling down would be much faster, than a 5 story building . How fast, is it measurable in fps? The lower building would get smaller diameter pipe I assume.


If all this is true, odr or variable speed flow was a byproduct. As return warmed, space was getting warm, the speed of the thermosiphon slows.

At days end is a room needs 20K of energy to maintain temperature, somehow the gravity powered system moved that amount. So someone calculated piping, fittings, vertical and horizontal to assure it all worked.
I've also read the upper radiators needed to be valved down to drive sufficient flow to lower radiators. So some limitations to adjustability from room to room or apartment to apartment.

Hence the double hung thermostats ;0

Steamhead

Yes, thermo-siphon is the technical term. I know I've seen some specific design info somewhere, just have to find it. And yes, the flow rate varied with the water temperature.

PRR

Increased heat transfer with higher flow rate: in AIR transfer we assume that a 3/8" layer is "dead air", in un-blown systems. That's a starting number with many variables. But the primary advantage of a fan is to "break up" or "scrub" that dead-air layer so heat flows better.
Water has viscosity and I assume a similar syndrome. Higher flow is less "dead space" and better transfer.
There's also studies on water-cooled engines (same problem in reverse) where dead-water layer is noted and sometimes scrubbed. But the best of these are 1940s aircraft engines and much of that was not published at the time ("loose lips...") and is fragmentary today.
If you go totally bonkers: you can heat with just a pump, no fire. Some hot-tubs do this. I had one, when I filled it, I let it pump non-stop for a couple days and it got hot. A 1 horsepower pump is about 2KW or like a small domestic tank-heater. If it just pumps around an insulated covered tub, the heat accumulates.

Steamhead

@PRR ,that is true for a more-recent system designed for forced circulation. But gravity systems, as I've said, were designed for much lower flow rates. That being the case, they are way too easy to over-pump.

EBEBRATT-Ed

IMHO the worst thing you can do to a gravity system is put a pump on it....if you don't have to.

I Installed a Smith BB14 as a replacement on an old gravity system. I talked with Smith about it first. I asked them what I would gain by installing a circ pump if the house was heating ok as is and they had no answer.

Now if the house would not heat due to old clogged gravity piping that would have been another story.

dronic123

@EBEBRATT_4 : do you think you can hook up something like a Weil Mclain or a Slant Fin residential CI ( say BTU's 100-150) and would it work on a gravity system without a circ/pump?---or is the pot of water too small?

Steamhead

dronic123 said:

@EBEBRATT_4 : do you think you can hook up something like a Weil Mclain or a Slant Fin residential CI ( say BTU's 100-150) and would it work on a gravity system without a circ/pump?---or is the pot of water too small?

Probably not. The internal resistance would be too high, so you couldn't get the proper flow through it.