Question about Delta-T circulators.

Jean-David Beyer

If I have a delta T circulator that is set to run with a delta-T of 20F and the actual delta T is less than that, the circulator slows down, right?



What happens if  the supply to that circulator is 75F (from a mod-con on a warm day into radiant slab) and return is 74F? The circulator slows down to something or other, but it will never see a return temperature of 55F.  Will it stop altogether?



If it stops, at some point the pipes where its temperature sensors are will cool to ambient, and the circulator will remain off. Then even if the load gets too cold, the delta T will be too low and the circulator will not restart.



I suspect that my assumption that the circulator will slow down until it stops must be incorrect. But then it will not be maintaining the specified delta T.



But if there is nothing in the circulator to keep it from stopping, would it not be necessary to raise the supply temperature to one high enough that the connected load would cause a 20F drop? And would not doing this lower the efficiency of the mod-con? Mine, for example, may now get a delta-T of 5F most of the time in cold weather and 1F today when it is 45F outside. But I would really not wish to raise the supply temperature on design day from 112F to 132F into my radiant slab just to get a 20F delta T. And maybe 132F would not be enough.



Do delta T circulators make sense with a mod-con into a radiant slab?

furnacefigher15

Yes

The circulator would slow down to meet its set point if the rise is lower. But, I don't believe the motor would stop altogether, but run at a minimum speed set by the manufacture.



at 75 degree supply, what is the slab temp ?

NRT_Rob

typically

this kind of thing is a problem, including for outdoor reset curves on boilers. You can't really have a 75 degree outlet temp on a mod/con, ever, if you plan for 70 degree rooms unless you are designing for very tight boiler firing differentials, which would only make sense if you have a buffer tank or mass you're working with. even then 10 degree is a typical minimum and 20 degrees is usually better.



So, Step one is making sure the minimum outlet temp on your heat source leaves a DT to room temp that you can work with. otherwise, even if the pump runs, it may not satisfy load as flow rate could be too low. You can't follow a perfect curve for outlet temp all the way to room temp. cut that sucker off at the knees!



the VDT pump does have a minimum speed output setting. you don't set it, you just tell it to use "normal" or "lower" minimum flow. no experience myself, just reading the instruction sheet here: http://www.taco-hvac.com/uploads/FileLibrary/102-359.pdf





Further, if you only have a 1 degree dt on the loop, you have no real load.



with ANY very low temp system, the benefit of temperature reactive equipment diminishes as you do not have as much temperature range to play in, that is definitely true.

Jean-David Beyer

typically this kind of thing is a problem

But is it a problem? What is the problem with having a 1 degree delta T, providing I am getting enough heat, and providing the pump is not running so fast as to get noise in the pipes or cavitation in the circularors? The only problem I could think of is excessive power consumption in the circulator. And if at the design day, I still get enough heat, I could reduce the size of the circulator until I either stop getting enough heat, or until the heating uniformity of the zone is too non-uniform.



"So, Step one is making sure the minimum outlet temp on your heat source

leaves a DT to room temp that you can work with. otherwise,"



What do you mean? Are you saying I should raise the boiler output temperature sufficiently so that a delta-T pump can slow down enough to not overheat the room? That seems to say I would run the boiler hotter than necessary to preserve a high delta T that is not required to provide the necessary heat to the load. And that would reduce the effectiveness of the boiler by reducing the condensing and the effectiveness of the reset.



"even if the pump runs, it may not satisfy load as flow rate could be too

low. You can't follow a perfect curve for outlet temp all the way to

room temp. cut that sucker off at the knees!"



I do not understand you. It seems to me that at my present settings, I can run 75F into the slab until the outside temperature drops to 50F. At design temperature (14F), it is sufficient to drive the slab with 112F, and at 6F, 120F is enough. I do not know what the delta T is at 6F, but at 112F, the delta T is about 6F if I remember correctly, and at 75F it is about 1F. I ALWAYS get enough heat. I do not know how low a flow rate I could get away with. My only reason that I can think of to lower the flow rate is to reduce energy consumption of the circulators. But since no matter what I do, I always get enough heat, what benefit would I get from running a greater delta T just so a delta T circulator could run in its normal range?



Cut it off at the knees? Warm Weather Shutdown occurs at 70F outside temperature. That is factory default and I never bothered to change it. I suppose I could lower it to 69F or 68F or something, but do not see the point. Usually when it is that warm out, the thermostats do not call for heat anyway.



"Further, if you only have a 1 degree dt on the loop, you have no real load."



I think I agree that when it is, say, 60F outside, I have no real load. So what? I still need a little heat, and using 75F water is enough. I could go even lower, but the boiler cannot modulate down enough to get below that, and cycled more rapidly than I cared for when I tried it. At 0F, the load of that zone is 24,000 BTU/hour. It never gets that cold out. It rarely gets under 10F outside. The indoor thermostats are at 69F.



"with ANY very low temp system, the benefit of temperature reactive

equipment diminishes as you do not have as much temperature range to

play in, that is definitely true."



I guess I am coming to the conclusion that if I have a very low temperature system, and it always provides enough heat, that I might as well forget about getting a higher delta T from the system. It looks to me as doing anything to change the delta T would necessitate running the boiler at higher temperatures, and that would defeat the use of the mod-con boiler that is already in there.

Jean-David Beyer

at 75 degree supply, what is the slab temp ?

My reply disappeared.



I do not know the slab temperature. But the return is typically 74F, so I guess the slab temperature, at the level where the copper tubes are, must be about 74.5F.

NRT_Rob

the problem is this

the DT action of the pump simulates a certain average temperature (supply to return) on your given supply temp, which is set by a reset curve that is linear.



so if you have a 70 degree room, the first issue is the boiler differential. If the boiler fires on, say, a 20 degree dt, depending on the boiler you have to make sure you have at least 10 degrees between room temp and the minimum supply water temp you ask for from the boiler. Otherwise, you have a BOILER waiting for your room temp to drop before it can fire.



you might be tempted to lower the firing DT to drop your minimum temp, but I would not do this. wider firing DT equals less cycling: good. Use wider firing DT's. the 5 degrees of average temp you might lose are not worth a measurable amount of efficiency at this temp range, most likely.



so, your minimum water temp will be at least 80 to 81, I would hope. Maybe 85 or 86 with a 30 degree firing differential. Now we're already into a differential range the circ can work on in all conditions.







If instead you run a very low firing DT because you have a ton of mass or something, you can run into the other problem. when the pump is on minimum speed, you are not running the pump on DT anymore. imagine a situation where the load is steadily growing from zero and your flow is constant, and tiny, because you haven't hit the DT of the pump setting yet. Say the pump is set to 10 degree DT (though at that point, I'd say you wasted your money on a DT pump, since you're now overpumping your system)... Supply temp goes to 71, 72, 73, 74... rises with the load. but your flow doesn't change and this supply temp increase doesn't ensure you are meeting the loads at that temperature... you'll just get room temp water back because you're "trickling" the flow through, and your boiler will short cycle. In effect, your heat output is something probably approaching zero until the pump sees the DT it's looking for. how much depends on exactly how low that minimum speed is. But I'd be concerned about it without better info about the minimum pump speed.



and the problem is this conditions persists until you hit the pump differential.





my 1 degree comment was about not worrying about conditions with a 1 degree dt because you don't have a noticeable heat load so even if the system doesn't run, it doesn't matter. DT doesn't exist in a vacuum. it only occurs in response to actual load.

NRT_Rob

I should clarify a point

this can be confusing: on firing differentials, some boilers split the differential. So if you have a 75 degree setpoint and a 10 degree differential, it fires 70 to 80. sounds like you might have that.



some however would try to figure 65-75... the setpoint is a maximum value, not an average value. need to be careful with differentials and low temps as this is a big difference!

Jean-David Beyer

this can be confusing: on firing differentials

My boiler can set the differential anyway I want. On the big (slab) zone, it is set to +5F and -5F, so when the reset control asks for 75F, the supply varies from 70F to 80F. That happens to be the default setting. But I can set the upper limit and the lower limit independently. The thermostat happens to be set at 69F.

Jean-David Beyer

the first issue is the boiler differential. If the boiler fires on

"so if you have a 70 degree room, the first issue is the boiler differential. If the boiler fires on, say, a 20 degree dt, depending on the boiler you have to make sure you have at least 10 degrees between room temp and the minimum supply water temp you ask for from the boiler. Otherwise, you have a BOILER waiting for your room temp to drop before it can fire."



OK, but my differential is + and - 5F, though I could set the upper and lower limit independently around the temperature demanded by the reset curve. I try for a 69F room with a digital thermostat. So with the reset curve demanding 75F supply when it is 50F or warmer outside, I will not get the boiler to fire unless the return water is less than 70F. But the return water temperature is what? I guess it is the average of the top temperature of the slab and the bottom temperature of the slab. Say the top is 68F, and the uninsulated bottom is 58F (I made up the 58). In that case the supply temperature will drop pretty quickly to less than 70F and the boiler will fire. If the return temperature is more than 70F, it will heat the house somewhat, and in practice, it heats enough.



"you might be tempted to lower the firing DT to drop your minimum temp, but I would not do this."



I left the differential at the factory default of 5 degrees above and 5 degrees below what the reset curve demands.



"so, your minimum water temp will be at least 80 to 81, I would hope.

Maybe 85 or 86 with a 30 degree firing differential. Now we're already

into a differential range the circ can work on in all conditions."



I sure would not want that much differential. I had something like that with my old oil fired non-reset boiler and the overshoot driving my 50 ton (estimated) concrete slab was seriously uncomfortable. Also, in my smaller baseboard zone, the expansion and contraction noise would be unpleasant. Running 85 to 115F water on anything warmer than design day (where I now run 112F) would cause overshoot and reducing the condensing ability of the boiler. I see no benefit to doing this.



"If instead you run a very low firing DT because you have a ton of mass

or something, you can run into the other problem. when the pump is on

minimum speed, you are not running the pump on DT anymore. imagine a

situation where the load is steadily growing from zero and your flow is

constant, and tiny, because you haven't hit the DT of the pump setting

yet. Say the pump is set to 10 degree DT (though at that point, I'd say

you wasted your money on a DT pump, since you're now overpumping your

system)..."



To take last things first, I am probably overpumping my system. I have a Taco 007-IFC driving about 750 square feet of radiant slab (about 32 feet x 24 feet) in 5 circuits, one under each room. I do not know the length or spacing of the 1/2 inch copper tubing. I have about 50 tons of mass (thick concrete slab). My guess is that you are saying unless I increase the differential of the supply water, wasting the ability of the modulating condensing boiler, I will not benefit from a delta-T circulator. I suspected that already, but would like to understand the benefit of high delta-Ts so seemingly recommended for hydronic systems. My impression is that the delta-T around an hydronic loop is something that you might as well wait to observe, not something specifically designed for. Design to keep the maximum flow less than 4 feet per second, and probably more than 2 feet per second and be done with it. Take the delta T you get. Providing you get enough heat, of course. And if you do not, increase the heat emission by having more area, or higher temperatures. But even if you do this, you still take the delta T you get.



"my 1 degree comment was about not worrying about conditions with a 1

degree dt because you don't have a noticeable heat load so even if the

system doesn't run, it doesn't matter. DT doesn't exist in a vacuum.

it only occurs in response to actual load."



Well, when it is 55F outside here, and my delta T isabout 1F  it may not have, from the point of view of the boiler, a noticeable heat load, it does matter if the system runs or not. Putting 75F water into that slab will keep the thermostat comfy at it s set point, and if the system did not run, it would get too cold. So there is an actual load, but it is very small when it is warm (>50F) outside.

NRT_Rob

mostly incorrect

assumptions on firing differential. you need to think average temperature, not peak temp.

Jean-David Beyer

you need to think average temperature, not peak temp.

That sure does not agree with my experience.



When I had my old boiler with no reset, it put water between 130F and 140F into the slab. Yes, this was wrong, but that is what it did. I left it as the original homeowner had it, and I did not know any better. I certainly should have lowered the temperature to between 120F and 130F, perhaps. My guess is that it was already condensing, and lowering it would have made it worse. The main thing I did with it was open up the differential. It was set at about 135F +|- 2F and I raised it to +|- 5F to reduce the very rapid cycling. Like on for 45 seconds, off for 120 seconds or something like that. I might have increased the differential a little more, but when I set it to 15F, the pressure relief valve in the make up assembly would leak. I did not know, yet, the the old-style expansion tank needed draining a couple of times a year. I figured that out by my second year here.



So what happened? The thermostat (set at 68F) called for heat. The slab started to warm up. Eventually the zone (there was only one zone at the time) got warm enough and the thermostat was satisfied. The hot slab continued to heat the house for 4 to 8 hours and the temperature went up to 75F or a little more. That was just plain uncomfortable. Then the house started to cool down. By the next morning, it could get down to 65F or so, which was too cold. The thermostat had already started to call for heat, but it would take 4 hours or so to start warming up again. Now the average temperature of the house may well have been 68F, but the peak temperatures (both hot and cold) were not acceptable.



For the downstairs zone (the one with the slab), the cycling rate is quite slow. When it is really cold out, the thermostat will call for heat for 12 to 18 hours straight, and the boiler will fire and modulate almost all the way down and just run. It cycles more rapidly than that when it is warm out because the boiler will not modulate down far enough, but still it is of the order of one cycle per hour.



With a differential of 10F on the boiler and the reset running the

supply temperature close to what is needed, the thermostat (now set to

69F) never goes below 68F, and it rarely makes it up to 71F except on

days like today where it has gone up from 50F outside to 71F in about 3

hours.



Problems occur only with the upstairs baseboard zone that requires 6500 BTU/hour only when it is 14F below the design temperature, and the boiler will modulate down only to 16,000 BTU/hour. I did raise the differential for that zone so as to be +8F and -7F; i.e., 15F total instead of 10F. I also lowered the maximum firing rate of the boiler when that zone was running to 55% instead of default 94% so the servo control would not overshoot. It is underdamped when running such a small zone. I also set the minimum supply temperature to 110F instead of about 80F that I intended to use. These changes enabled me to reduce the cycling rate to about 5 cycles per hour in most of the heating season. And that zone heats only about 4 hours a day in that zone because by then the sun shines in the windows, and heat drifting up from downstairs can maintain the temperature, so the thermostat does not call for heat much in the afternoons. I could increase the differential up there still more, but I would have to do it by lowering the minimum temperature. I would not want to raise the maximum temperature because that would reduce the condensing. And a wider differential would make more expansion and contraction noise as the water temperature changed.

NRT_Rob

hmm

Your room temperature swings were because you were so far above your temperature requirements in total, mostly. you overcharged your slab by the time the thermostat satisfied... simple thermal lag situation.



If your slab can take a large portion of the output of your boiler though, then you are in the high mass situation I noted and a wider differential won't help... that's a variable situation amongst slab people depending on the size and loop layout of the slab (i.e., how fast it can dissapate heat on startup).



though wider differential and a PWM thermostat could tame the slab better than trying to do it on a firing differential alone, if you are not actually seeing any short cycling issues in moderate weather then I would say your slab is doing a fine job of buffering and definitely ignore my ramblings about wider dts to reduce cycling.



I have to admit I have no idea why raising your minimum temp on the baseboard would reduce cycling. raising your minimum lowers your differential and should only increase cycle frequency. what am I missing?

Jean-David Beyer

raising your minimum temp on the baseboard

From the boiler's point of view, running the baseboard at 80F when it is warm out (say 60F) and I am trying to heat to 69F inside sinks so little heat from the boiler that its temperature rapidly rushes up to the upper limit, even though it is 10F higher than the lower limit. The boiler, though it is the smallest in the product line, will only modulate down to 16,000 BTU/hour, and that zone may perhaps need only 1000 BTU/hour or less in that temperature. Running the baseboard at 80 or 85F will just not put enough load on the boiler, so it had very rapid cycling; perhaps 10 or 12 cycles per hour. Raising the minimum temperature from 80F to 110 F means that more of the minimum amount of heat the boiler can produce will be dumped into that zone. Thus, the temperature increase from the lower limit will be slower because more of it is going into the rooms and less returning to the boiler. So the boiler takes longer to reach the upper limit.



A side effect is that the thermostat is satisfied sooner and the thing can shut off entirely, which is certainly a lower cycling rate -- if you choose to look at it that way.



If my goal were to just reduce the cycling rate, I could let the boiler run up to 180F and the cycling rate would go down even more, and the temperature would go up and satisfy the thermostat sooner. I might get away with that with the low mass baseboard. It would never work with a 55 ton concrete slab. But if I let the supply get over 130F, condensing would pretty much stop, and since I get all the heat I need with the supply under 131F even on design day, it seems to be a mistake.



Remember the main goal is to get enough comfortable heat. If I can reduce the cycling rate also, that is OK. It is difficult for me to believe that giving up the condensing just to lower the cycling rate is the right trade-off.



Back to the original question. If I get the right amount of heat at the lowest supply (and therefore return) temperature, why should I change things to get more delta-T in the heating loop(s)? I am thinking that the delta-T is something interesting to observe, but not something to design for. Flow rates, low temperatures, sufficient heat emitters, are things to design, but then just take whatever delta-T you get.



And sure, if I could have gotten a 40,000 BTU/hour boiler instead of the 80,000 BTU/hour one I did get three years ago, I should have. I am glad I refused the 105,000 the contracting company tried to sell me, "to provide a margin of safety."

NRT_Rob

something is missing here

Oh, I see what you mean. You shifted your whole reset curve upwards, not just the firing differential, I misunderstood what you were talking about. that makes sense. you could have done the same thing with a wider differential and a lower curve, but six and one half dozen or the other, no big whup.



higher dt is just less pump energy. me, I'd just use an ECM circ and run at half power all the time. when ECM DT is available that will be interesting indeed

Paul48

J-DB

I don't think a delta-t circulator is a good choice for radiant.It's not your typical emitter,and to try to get a noticable delta-t when attempting small changes in slab temperature probably would not work. I have to admit, I was shocked you run a 75* supply temp., but you know what works for you. What's the measurement for that......candle-power? :-)

Zman

70 degrees?

How are you heating the room with supply temps lower than 70?  Even if you are only heating the room to 65 and you have a ton of surface area the floor target would have to be at least 70. At the coolest the return water would be 70 and the supply water 80.In my opinion you are just wasting circulator energy by going that low.