I hope I am not beating a dead horse.

Jean-David Beyer

This delta-T circulator question.



I admit that within a certain range of supply and heat loads, a delta-T pump can maintain that delta T. But I still do not understand why one would want to maintain a fixed delta T, or any other delta-T other than what would normally come out of a system.



I just calculated a little system that operated at 10 F delta T and at 1F delta T.

The results were as shown in the attached spreadsheet, if I can get it in here.

It shows that running with a lower delta T actually puts out more heat than at the higher delta T. A corollary is that for the same amount of heat as the 10F delta system puts out, you could run a fixed speed system at a lower supply temperature.



It says my .pdf is not a valid .pdf file. So here it is as straight text. I divided a 11 foot piece of baseboard into 11 1-foot pieces and calculated the temperature of each piece separately. I then looked up the BTU/hr/ft output from each section at that temperature. I had to interpolate (linearly) the in-between temperatures, since the charts do not show things that fine. So there as slight errors because the loss is not quite linear. But the errors should not be enough to upset the results very much.



Segment T BTU/hr/ft  Segment T   BTU/hr/ft

120                      200          120.0         200.0

119                       195          119.9         199.5

118                       190          119.8         199.0

117                       185           119.7         198.5

116                      180           119.6         198.0

115                      175            119.5         197.5

114                      170           119.4         197.0

113                      165           119.3          196.5

112                      160           119.2          196.0

111                       155           119.1           195.5

110                      150           119.0           195.0

  10                   1925                1             2172.5

DeltaT   BTU output         Delta T  BTU output

heatpro02920

Delta circs

Im not sure how much maintaining a uniform set delta T will save you in energy, but I know it helps with short cycling {does wonders for over sized mod cons}, keeping a uniform temperature inside heated spaces, and using less energy in general{when they slow down the pump it uses less energy, instead of running full all the time}...



Also for systems with multiple zone valves off of one circ, , say the system has 4 25K btu zones, when you size the system you do the math for all zones calling on a design day, but if 1 zone is calling you are going to move the water too fast, this is going to make the return temps higher and lead to short cycling and is going to make the circ run longer... {in other words, circ runs longer burner runs shorter, kind of the opposite of what we want...



No as for what delta t you want, that is going to be up to the installer to figure out when they design the system. Say I am designing a system that has a 200K btu heat loss, and it is being split up into 4 equal loss zones of 50K btu each, I will size the boiler for 200K, and each zone for 50K, now say I do this with hydro airs, the hydro air unit is going to tell me what temps and at what gpm they need for me to get 50K btu, this is where I start because I need to figure out my delta t to get this out of the coil...



so I will refer to the hydro airs instructions to see what fan speed, gpm, and temps I need to get my 50K, now here comes the problem, I don't always need 50K {only on design days}, so I use a modulating boiler controlled by an OD temp and I use a variable speed circ controlled by delta t, this makes the system possible...



I hope I explained this rite, Im not very good at putting it from the brain to the PC...



But to answer your question, on why one would want to have a set delta, that is it, to control short cycling and help with constantly changing loads and temp conditions... give the water time to get the energy into the heated space in a uniform manner...





I hope this helps..

Eastman

Keep in mind...

that discussions such as this are not pitting fixed flow versus fixed delta T.   A variable delta T does not imply fixed flow or constant delta P.  Other methods of pump control should be considered.

Jean-David Beyer

you are going to move the water too fast

I guess I still do not understand this. It seems to me that the faster you move the water, the more heat the emitter will dump into the load. The upper limit from this standpoint is where there no appreciable temperature drop from one end of the emitter to the other. At that point you get no more heat out even if you increase the flow rate. And perhaps before that, you will get noise and erosion of the pipes -- but that is a separate discussion. If you ignore electrical cost, and possible noise and metal erosion expenses, it seems to me that you want to run the circulator as fast as you can. I do not seem to get noise in my system at its current pumping rate where the delta T seems to be much less than 10F.



At the other extreme, if I lower the flow to get a really big delta T, such as so the return extremely low, the emitter will cool down to room temperature long before it even gets to the return end of the emitter. Imagine it gets to room temperature half way through a baseboard. That means the other half of the baseboard is completely useless. Of course this will not happen in any reasonably designed system. It comes kind-of close in my radiant slab zone when it is about 50F outside. In that case, I put 76F water in and it comes back about 75. I wonder how slow I would have to circulate it to get delta T of even 7F. And I expect if I did that, I would not get enough heat out of it at that supply temperature, and the heat distribution would be very uneven.



So now I run around three gallons per minute (calculated, not measured) through my zone that cannot possibly use more than 6500 BTU/hr and realistically, it uses much less than that. And the boiler will modulate down to about 16,000 BTU/h (input), so it is going to cycle pretty rapidly. If I lower the flow rate to get a 10F delta T, that emitter will consume fewer BTU./hour than it does now, so it will extract less heat from the primary (boiler) loop. So how will that lower my cycling rate?

CMadatMe

Your Not Getting It

Because your stuck on water temp. If your not removing btu/hr out of the conveyor belt your not moving out the conveyor belts btu/hr being carried.



Forget water temp it is irrelevant. It's btu/hr you want.



180 Supply

160 Return



20 Degree Delta

1gpm flow rate

10,000 btu/hr removed from the piping



180 Supply

170 Return

1gpm flow rate

5,000 Btu/hr removed from the piping



Heck



140 Supply

120 Return

1gpm flow rate



10,000 btu/hr being removed from the supply piping



110 Supply

100 Return

1gpm Flow Rate



5,000 btu/hr being removed from the supply piping





BTU is. The amount of energy needed to heat one pound of water by one degree Fahrenheit.



The definition doesn't say, "The water temp needed to heat one pound of water by one degree."



The energy in your supply piping - the energy your BB took out = The energy left over!



How do we define this:



gpm = btu/hr / (Delta-t x 500)



So that is the btu/hr you removed from the piping system. Now the baseboard or emitter requires a water temp (180) x a flow rate (1gpm) to move btu/hr sqft. Increase that flow rate and decrease the delta you end up taking less btu/hr out of the system piping.



180 Supply

160 Return

4gpm flow rate



4 x (20 x 500) = 40,000 Btu/hr removed



180 Supply

170 Return

4gpm flow rate



4 x (10 x 500) = 20,000 btu/hr removed



180 Supply

175 Return

4gpm flow rate



4 x (5 x 500) = 10,000 btu/hr removed



140 Supply

120 Return

4gpm flow rate



4 x (20 x 500) = 40,000 btu/hr removed.

Matt_67

emitter

I think what JDB is getting at is that using a Delta T pump in and of itself will not help a boiler short cyling problem when dealing with a small zone.  Slowing down the zone flow to achieve a higher Delta T causes the total BTU output of the zone to drop as he showed in his calculations.  Obviously it has other advantages as regards water temperature return to facilitate higher running efficiencies but it does not result in higher BTU output from the zone.

CMadatMe

Because Your Piped Pri/Sec

Your boiler flow rate is 7.1gpm. The reason you short cycle is because your sending more then half of that gpm back to the boiler return mixing with the return temp keeping your boiler temp elevated. You cannot pull the btu/hr out of the boiler. If you had a boiler pump that moved 71,000 btu/hr / (40 x 500) = 3.55 gpm and your system flow rate was 1gpm what would your boiler return temp be? Boiler is making 130 degree water.

Eastman

heat output

The interesting question here is what is the ideal fixed delta T for a particular system.

CMadatMe

Matt This All Started

Because I say sizing a boiler pump on a condensing boiler piped pri/sec or LLH for a 20 degree rise is really wrong. Boiler pumps should be sized for the highest rise that mfg allows. Why? It reduces the boiler side flow rate giving the system side the ability to remove all the boilers flow into the system side. JD has a Ultra 80 with a Taco 007 on it. That pump is moving 7.1gpm across the boiler HX at all times every time it's on. He cannot remove 7.1gpm to his system side so all that nice boiler water and btu/hr just created b-lines right back into the boiler return elevating the return temp.



If his pump was sized for a 40 Rise that pump would now be moving 3.55gpm at all times across the HX. A lot less newly created boiler water/ btu/hr headed back to that return.

Eastman

@Chris

"Boiler pumps should be sized for the highest rise that mfg allows."



How far would you take this?  Surely you wouldn't specify a primary fixed flow rate that is less than the minimum flow possible in the secondary.

CMadatMe

Why Not

Just have to make sure my mixed supply temp will do the job. We do it every day in the Viessmann world. Look at a Vitodens 100 WB1B10-35, 118,000 btu/hr input, 109,000 out put. Max flow rate across the HX is 6.2gpm.



How do I get 109,000 btu/hr out put? Simple



109,000 / 500 /6.2 = 35 Degree Boiler Rise! but we don't even size for 6.2gpm



109,000 / (40 x 500) = 5.45gpm! Boiler Side



I'm just moving 5.4gpm of x water temp carrying 109,000 btu/hr to the secondary side. I'm still carrying btu/hr. So you want to run the system side on a 20 degree delta go ahead. What water temp do you need at design from your heat loss and emitter comparison? If I can't do it because my mix supply temp is too low then I'll increase by boiler flow rate and size my pump accordingly. Bet I don't need more then a 35 degree rise. I don't want to hear supply water temp either because it's more important to condense then that possible 5-8 degree water temp difference.



109,000 / (20 x 500) = 10.9gpm System Side



5.4gpm Boiler Side

10.9 gpm System Side

Need 150 degree supply temp



See attached. We both are condensing. Where it matters most is when all the zones are not calling. Say You only needed 5gpm I'd be condensing and you wouldn't

Matt_67

agreed

I agree with you with one caveat and it involves systems that were engineered and installed with higher water temperatures in mind.  I thought they didn't exist until we got involved with a weatherization program for some subsidized apartment buildings built in the early 80's.  The original systems were installed exactly to the original spec and on design day required 180 degree water and ran with an 18 degree temp drop.  In that situation you need the bigger boiler pump.

On a side note - with these firetube boilers with low heat exchanger water side head loss - how cool would it be to have a system with a single pump (no primary secondary) that had a setpoint for minimum flow rate and then used delta T to control above that rate.  

Jean-David Beyer

The reason you short cycle ...

I think it is unimportant to this discussion whether the boiler is piped primary-secondary or some other way. I hope we can agree that the no heat is dissipated in the boiler loop: it is either transferred into the system loop or returned to the boiler.



Therefore, the cycling of the boiler is determined primarily by the heat dissipated by the emitters in the system loop (since it is dissipated nowhere else), and it seems to me that the more heat dissipated by the emitters, the lower the cycling rate will be.



"If you had a boiler pump that moved 71,000 btu/hr / (40 x 500) = 3.55

gpm and your system flow rate was 1gpm what would your boiler return

temp be? Boiler is making 130 degree water."



If these assumptions are correct, the boiler would have modulated down to about 14,500 BTU/hour. My calculated system flow rate is a little under 3 gpm. The temperature drop is often around 1 or 2F, so whatever that flow rate and temperature would tell how many BTU/hr is going into the heating zone. Perhaps 500 to 1500 BTU/hour -- but I am guessing here. So it will cycle.



If I lower the flow rate in the system zone to get a greater delta-T, it will dissipate less heat, not more, and return more, not less, hot water to the boiler loop and thus to the boiler, increasing the cycling rate.

Jean-David Beyer

The interesting question here is what is the ideal fixed delta T for a particular system.

I am coming to the conclusion that there is no ideal fixed delta T for a particular system, at least for a mod-con with outdoor reset.



It is set to run with as low a firing rate as it can manage and still deliver the water temperature determined by the reset curve. And if that curve is right, it will be just barely enough to make up the heat loss of the building. And in that case, you want to get the maximum heat out the emitters you can at the temperature you supply to them, and that is at as high a flow rate as you can, consistent with not getting noise, erosion, or excess electrical power consumption. And then you  take whatever delta-T you get. All a delta-T pump can do is reduce the flow in order to raise the delta-T and that decreases the heat output of the emitters when what you want is to increase it.

Jean-David Beyer

Your Not Getting It Because your stuck on water temp

No I am not. That is why I calculated the BTU/hr/ft at the different flow rates. At the higher flow rate (lower delta T) the emitter puts out MORE BTU/hr/ft that it does at the lower flow rate (higher delta T).

CMadatMe

Boiler Pump

Forget the min boiler btu/hr. It's the flow rate of the boiler

Pump that kills you.

Eastman

@Chris

"Where it matters most is when all the zones are not calling."



Don't get me wrong.  I totally agree there is an incentive to have a higher delta (higher than the secondary) on a fixed flow primary connected to a variable flow secondary system.  But having mismatched flows and deltas at design conditions is a compromise  --clearly on a more typical day (where most of the dollars are spent) the beneficial trade off is synchronized flow with fewer zones calling and hence synchronized deltas.  And that is good.  But, and maybe this is where we disagree, the theoretical ideal would be to synchronize over the entire allowable flow range of the two subsystems.  This is never done in practice, but it's interesting to speculate how this could be accomplished.  This is why I don't see any advantage to purposely selecting flow/delta boiler combinations that are entirely outside of the system design parameters.  For example, why would one purposely select a mismatched primary/secondary for a single zone fixed flow system unless necessary?

Jean-David Beyer

Forget the min boiler btu/hr. It's the flow rate of the boiler Pump that kills you.

How could it? It neither adds heat to the system, nor does it remove heat from the system. The burner adds heat, the emitters remove it. The circulators just move the heat around.

CMadatMe

Guess

You haven't read any of my post. Flow rate moves the

Btu/hr created. The boiler pump is fixed and no matter

where the boiler fires flow rate is the same. The

Boilers burner uses the temp rise as it's target. If the

Boiler min is 16k. Then you only need to move

1.6gpm across it on a 20 Rise. Why cant't I move less?

SWEI

Surely you wouldn't specify

It matters not which side of the system has the higher flow rate.  Think of closely-spaced tees (or any other hydraulic separator) as an bidirectional continuously variable transmission.  This gives you the freedom to select an optimal flow rate in both source and load loops.

Paul48

In your case

"The circulators just move the heat around"

They move it so fast, it returns to the boiler with little change..

Jean-David Beyer

They move it so fast, it returns to the boiler with little change..

Let me rephrase what you said: they move it so fast it returns to the boiler with little TEMPERATURE change, but it has given off more BTU than had it flowed slower with greater TEMPERATURE CHANGE. It has given off more heat with high flow than low flow, so more of the heat produced by the boiler is moved into the zone than if the flow rate were lower with more temperature change. And increasing the amount of heat given off by the emitters reduces the cycling rate because more of the generated heat is given off to the load and less of it (the heat, not the temperature) is returned to the boiler.

Paul48

Perfect!

For an oil-fired cast-iron boiler.

Jean-David Beyer

You haven't read any of my post.

Sure I did.



"Flow rate moves the Btu/hr created. "



It may and it may not. It depends on whether or not the load will accept it. And that depends on the gradient beween the water on the water side of the emitter and the air on the other side of the emitter. And the greater the delta-T of that water, the less the gradient as you get to the return end of the emitter. As my table shows.



"The boiler pump is fixed and no matter

where the boiler fires flow rate is the same.'



Its flow rate is the same: yes. The water flow rate.



"The Boilers burner uses the temp rise as it's target.



Not the Ultra 3: it mainly uses the supply (to the system) temperature as its target. And it adjusts its firing rate to assure that, slowing down as it gets up to it.



" If the Boiler min is 16k. Then you only need to move



1.6gpm across it on a 20 Rise. Why cant't I move less?"



First the practical reason: W-M supply a fixed speed circulator with the unit. If they wanted the installer to diddle that flow rate, they would not bother to supply a circulator at all; they would let the installer pick the appropriate circulator and save themselves the cost of it.



Second, and I do not know if this is necessary to prevent flashing in the heat exchanger, but it might be: it holds only about 3 quarts of water, and the burner can put 80,000 BTU/hour in there,  so its temperature can change very rapidly. So they may want this high flow to safeguard the aluminum heat exchanger. They say if you run their indirect (I do) I should use the same size circulator to run that. They sure seem to want that rate of flow through that heat exchanger.



Is not the heat moved the product of the temperature change and the rate of flow? If I double the temperature change by cutting  the flow in half, I come out even. But if I must lower the flow more than that to double the temperature change, I will actually transfer less heat. My chart shows that cutting the temperature change from 10 to 1 (90%) causes an increase of heat transfer of almost 13%.

hot_rod

I follow your logic Jean-David

In the case of heat emitter output, lets say baseboards. We all agree heat loss is a linear slope. ODR marches that very close, read any telemarketer essay on that subject. When you force a fixed delta T on a low temperature system, especially, what happens below 130; is a fast drop off in output. Very low flow, forced by the delta T being fixed really limits output from a convector. I've seen that plotted with a simulation program to show the fast droop at the low end.



Our goal should be to design around low supply temperatures. Good for condensing equipment, great for solar assist, and great for GEO powered hydronics.



At last years RHC round table I threw out 120F as a design temperature. Robert Bean stood up and suggested 100F. Those high bar Canadians!



Knowing all boilers now require ODR, it seems we will be running low supply much of the time. So I don't see how forced delta T fits where we are headed?



Most mod con boilers have PWM signals built in now. Soon we will have circs able to take that signal and modulate the boiler circ. viessmann was way ahead on that concept with Vitodens why has it taken other so long:)



So modulate the boiler pump based on firing, and if you have a distribution circ, modulate that on ODR.



Once the delta T pumps can be used either way, you have an ideal circ for boiler return protection for wood and pellet boilers ideally.



Perhaps a product developed for a market that doesn't exist until a small delta P is in the offering?



Really the distribution is in charge, not the boiler. A good read on thermal equilibrium in I dronics 12. An example of where a boiler would run, if it didn't have limits controls with under sizes, exact sized and over sized distribution connected to it

Jean-David Beyer

Perfect! For an oil-fired cast-iron boiler.

Can you explain why a mod-con with outdoor reset would be less perfect as far as this discussion is concerned? I happen to be more interested in mod-cons since that is what I have, but it seems the heat transfer issues are the same.



For example mod-cons seem to be piped primary-secondary and cast iron boilers seem to be piped more conventionally, and I do not see what difference if the cast iron boiler is fired with coal, gas, LP, or oil. I doubt that most of them modulate unless they are industrial size, and most of them probably do not condense either.



But just for laughs, one can imagine piping an oil-fired cast-iron boiler with a primary-secondary setup. (I cannot imagine anyone really doing that on a small residential system.) You might wish to have the boiler loop set up somehow to keep the return temperatures high enough, which would be easy enough if you had a modulating burner, or some temperature-controlled mixing valve between supply and return of the boiler loop. Now if we assume that boiler is oversized and therefore cycles too rapidly, how to we arrange to get sufficient heat demand on the system loop to remove all that heat? By slowing down the flow through the emitters? My table still shows that that lowers the heat removed from the system, not increases it.

Paul48

It's Your Horse

If you're happy with it, that's all that counts.No one can change your mind about the things you have come to believe, so it's a dead horse.

Jean-David Beyer

I follow your logic Jean-David

I am glad somebody does, because otherwise I would have to take seriously the possibility that I have a severe mental block that weeks (it seems) of discussion here, by professionals, has failed to clear up for me. It has not even convinced me that I am wrong. In fact I have been considering it and wondering how to identify it and fix it.



And it is not that I must "win" on this: there is no prize either way. I was just trying to figure out what advantage a delta-T circulator would have in my system and I could not figure out any. I can see the advantage of an ECM circulator, but I could not see a use for a delta-T model. If I were going to replace my two system circulators with zone valves, it would sure make sense to put in a delta-P ECM model though. And perhaps I will if those Taco 007-IFC units seize up or whatever their failure mode is.



On my Ultra-3, I use the ODR to modulate the firing rate and use fixed-speed circulators in the secondary (system) loop. The contractor decided on circulators instead of zone valves because he said zone valves were unreliable. Perhaps they were when he was a little boy. Anyway, they were wrong on too many other issues and are now my former contractor.



For my boiler using the ODR on the boiler firing rate seems the thing to do. Why heat the boiler water higher than needed only to mix it down for the load.



It does not matter for this discussion, but how would you hook up the ODR to the secondary pump(s)? The colder it is out, the faster those pumps run?  Because the slower they run the slower something like baseboard would heat. and the faster they run the more heat you get (but not forever: asymptotically), but my guess is that might get you 2:1 turndown . Would not more turndown get you a baseboard quite cold at one end and warm on the other with complaints about uneven heat? And to get the same heating, would not a higher supply temperature be required to get the same average heat output as you would get with a greater flow rate? I do not doubt that electronics could be devised to run such a thing, but it would be complex because of the non-linearity of the difference between outdoor temperature and system flow rate(s).

Jean-David Beyer

If you're happy with it, that's all that counts.

Actually, I disagree with that.



I am after understanding, not patronizing.



We are back to that other thread where it is clear we disagree, but I am unclear where the disagreement is. So we can go nowhere.

Eastman

@SWEI

Hydraulic separation gives you the freedom to select optimal and suboptimal configurations.  Why would one select the suboptimal? 

SWEI

suboptimal?

who said anything about suboptimal?

Eastman

me

That is what my post to Chris is about.  Flow rates do matter and compromises are necessary. 



Quoting myself:

"This is why I don't see any advantage to purposely selecting flow/delta

boiler combinations that are entirely outside of the system design

parameters." 



There are flow/delta combinations that are anything but optimal.  Just like there are transmission/engine combinations that are not optimal.



Again:

"For example, why would one purposely select a mismatched

primary/secondary for a single zone fixed flow system unless necessary?"

hot_rod

So is it worth

An extra 100 bucks to install a 'Bee. I like the fact it is a more efficient ECM motor, I like the digital display, and the features the microprocessor offer. It may or may not offer increased comfort, depending on your system and how you program it. So for me yes, I think many customers see the value.



The biggest unknown with this brand or any "smart" circ is how long will they last. If dirty power takes out the processors, like it does on high efficiency boilers, furnaces, etc then the value may not be there. Should they be powered from UPS devices like all our computers are to keep the power clean? We will know in 5-10 years from now.



I'm replacing all my circs with various brands of ECM based versions. I like the energy savings and the "bells and whistles". It in the eye of the beholder, I suppose

Jean-David Beyer

So is it worth An extra 100 bucks to install a 'Bee.

It may be an extra 100 bucks to install a bee if you are putting in a new system, but if I  have a perfectly good 007-IFC removed to put one in, it is whatever the retail cost plus the labor to have the bee installed. And for that I would like the hope of a payback greater than the amount of the electricity saved.

SWEI

and if the 007 was oversized

the Bumble Bee curve is more like an 008.



(flogging another horse here) Taco, how about some other pump end options for the Bee?  A choice of 006, 008, or 0010/12 hydraulics anyone?

Jean-David Beyer

and if the 007 was oversized

One of my 007-IFCs is oversized. Not so large as to cause noise, however. I have to look at the thermostat to see if there is a call for heat unless I put my ear right on the baseboard where the supply pipe enters the house. Or I could touch it.

CMadatMe

Only If It's

Used in one of its 4 Fixed Speeds. It will reduce the flow of the zone as the emitter takes less btu/hr out of the conveyor belt and the delta begins to shrink. So in the swing months Oct, Nov, Mar, Apr when the emitter isn't pulling out btu/hr she slows down. You could equate it to doing ODR. So you could perceivabley move just a gallon or less. Your delivering to the zone what the emitter can convect across itself, giving you the best chance to send cooler water back to the boiler. In JD case, just the essence of him walking in the room is probably putting out more btu/hr then his board.



With a delta-t pump its the flow that changes. It's the only none constant when your aiming stick is fixed. With a delta-p pump it moves the same gpm across the zone every time it comes on because the pressure drop in the zone never changes.

heatpro02920

age old question, P vs. T vs. a properly sized fixed

I use all 3 where the system calls for it, Delta T circs like the VDT and Bumble bees are great for a running 1 zone that never changes {no valves or adjustments} like a single hydro air, std loop of baseboard, or single radiant circuit. Delta P is a good choice for systems with multiple zones off of one circ or zones with bypass valves, ect where the pressure will change when the demand changes... They seem to work well on primary loops with multiple zones. And fixed circs, are great for when you just need a pump, like an out door stove, low cost hydronic system, ect...



I think if they could make a bumble bee a little cheaper they would be good for 95% of the residential needs, but at twice the price of a 007, a lot of times when a circ goes bad it just makes more sense to throw in a 007...