delta T circulator

[chuck172]

Has anyone had any experience with the new Delta T circulators? I know Taco makes them, I'm sure others too.

 

[jebatty]

Does Taco have a pump curve for these? I notice from manually changing the speed on my 15-58 that gpm's do not change very much through a single zone system. Pump head increases roughly by the square of the flow increase. But I can see a real benefit to these delta-T circs in a system designed to minimize head increase as zones activate.

 

[chuck172]

Taco pump curve:

http://www.blueridgecompany.com/documents/00CurveChart.pdf

 

[chuck172]

Sorry, wrong pump curve chart!

 

[pybyr]

- it also seems to me that from an energy standpoint (electricity used over the operating life to move the water/ heat) that one of the very new (in the US) "smart ECM" (permanent magnet motor) variable speed circulators will probably save more money than a 'conventional' (AC/capacitor) motor with a "smart controller" added onto it or wired up to it. I wanted to put a variable speed control on the AC motor on my forced-warm-air heating system, but as I looked into it, any of the detailed information that I was able to find indicated that wattage does not decrease at the same rate as speed when you "throttle back" a single phase AC motor- which sort of negated some of the advantage of speed control.

ECM motors apparently do see much more of a drop in electrical usage (than conventional AC motors that have been the norm in US residential circulators) as they respond to conditions (and they're apparently even at least somewhat more efficient at full speed).

I'm not trying to 'knock' the various inherent advantages in system control that one might get from a delta-T controlled circulator-- just suggesting that if one is going to spend the extra $$$ above a conventional circ, you might want to go for an ECM one

 

[Chris S]

I had this discussion today in th esupply house. The supplier is trying to sell me teh Grundfoss Alpha ( ecm), I am using Wilo stratos- also ecm. He claimed that he Alpha can be plugged into the wall and will run around 8 watts at idle if you will, and ramp up as zone valves open without any input or actual connection to the boiler.
I questioned running 8 watts continuous vs letting the boiler control turn the pump on & off. He answered that if a trnsformer could be eliminated the electrical usage would be equal or less with the alpha. I'll field verify this to be sure.
I do know that the Wilo circs quite often run at less than 25 watts when they're pumping, so electrical savings is potentially big. A star 16 is 88w, a Taco007 is a little more than that.

 

[pybyr]

I had this discussion today in th esupply house. The supplier is trying to sell me teh Grundfoss Alpha ( ecm), I am using Wilo stratos- also ecm. He claimed that he Alpha can be plugged into the wall and will run around 8 watts at idle if you will, and ramp up as zone valves open without any input or actual connection to the boiler.
I questioned running 8 watts continuous vs letting the boiler control turn the pump on & off. He answered that if a trnsformer could be eliminated the electrical usage would be equal or less with the alpha. I'll field verify this to be sure.
I do know that the Wilo circs quite often run at less than 25 watts when they're pumping, so electrical savings is potentially big. A star 16 is 88w, a Taco007 is a little more than that.

FWIW, your supplier's claim made me curious, so I just went down cellar and connected an extra Honeywell 24 V 20VA transformer to an old appliance cord, and with an AC ammeter on it- which read 0.0 when the transformer was energized but with no load placed on the 24VAC side. I'm sure that there are some losses leading to some minor draw on the AC input, but from what I saw, it seems unlikely that it's 8 watts worth. Seems to me that smart/ ECM circs are worth a strong look, but not to the point that one leaves them running 24/7/365.

 

[Gooserider]

I had this discussion today in th esupply house. The supplier is trying to sell me teh Grundfoss Alpha ( ecm), I am using Wilo stratos- also ecm. He claimed that he Alpha can be plugged into the wall and will run around 8 watts at idle if you will, and ramp up as zone valves open without any input or actual connection to the boiler.
I questioned running 8 watts continuous vs letting the boiler control turn the pump on & off. He answered that if a trnsformer could be eliminated the electrical usage would be equal or less with the alpha. I'll field verify this to be sure.
I do know that the Wilo circs quite often run at less than 25 watts when they're pumping, so electrical savings is potentially big. A star 16 is 88w, a Taco007 is a little more than that.

FWIW, your supplier's claim made me curious, so I just went down cellar and connected an extra Honeywell 24 V 20VA transformer to an old appliance cord, and with an AC ammeter on it- which read 0.0 when the transformer was energized but with no load placed on the 24VAC side. I'm sure that there are some losses leading to some minor draw on the AC input, but from what I saw, it seems unlikely that it's 8 watts worth. Seems to me that smart/ ECM circs are worth a strong look, but not to the point that one leaves them running 24/7/365.

I believe that if you put a load on the transformer, you will find that you have some significant energy losses in the transformer itself... Also there are some losses even just with a bare coil - notice that the transformer will get somewhat warm as it sits there, (and again, more so if it is under load) - that heat is due to power being "wasted" as it goes through the transformer.

This is why the green folks urge you to pull the plug on "wall-wart" transformers used for charging cameras, cell-phones, etc. when the thing they are nominally powering is off or not being charged....

It gets very tricky when trying to measure the power draw of either a high inductance load like a transformer, or a high capacitance load as you get a phase shift between the current and voltage that will confuse most standard instruments that are designed on the assumption that I and E are in phase...

I would tend to guess that your transformer is probably radiating around 8-10 watts worth of heat due to internal losses, so the supplier claim doesn't sound that far off... Of course, unless there is some reason not to do so, I would be inclined to run the pump through a relay of some sort so that I could turn it to ZERO when it wasn't being used, but it would sound like doing longer runs at a lower flow rate might be a good money saver during the times when you can do so...

Gooserider

 

[Chris S]

Gooserider,
you obviously have a EE backround, while mine is civil, maybe should've been ME. My thought though is that relays have transformers too, or are they not energized until the relay is activated?

 

[Nofossil]

Relays have coils that draw power when the relay is activated - typically less that a watt.

My bench tests on a Taco 007 showed 78 watts at full speed and 0' head (or as close as I can get on my test stand). Interestingly, when I connected it to my NFCS variable speed controller, the power dropped to 65 watts at full speed, with the same flow rate. Maybe it's a measurement artifact, but interesting nonetheless.

At 50% power, it drew 38 watts and delivered about 1/3 of its rated flow.

 

[Gooserider]

Gooserider,
you obviously have a EE backround, while mine is civil, maybe should've been ME. My thought though is that relays have transformers too, or are they not energized until the relay is activated?

A standard mechanical relay does have a coil (I'll ignore electronic relays for this discussion) but it's a tiny one, relatively speaking, and does not function as a transformer. Essentially what you have in a relay is a momentary push button switch with a steel peice in the button and an electromagnet that will push the button when it's turned on... It's function is to use a small signal to control a much larger and / or different sort of current flow... i.e. that low voltage, low current thermostat turning on the 110v AC high power furnace motor... Think of it as a doorbell button with an electromagnet taking the place of your finger...

There are all sorts of circuits that use relays, but fairly typical in the computer world is a driver transistor that has an output sitting at a logical "high" 5 DC volts when off, and connects to an active low 0 volts when it is turned "on" by the controller - Connected to the transistor output is one side of the relay coil, and the other side of the relay coil is a 5VDC supply. If the transistor is "off" then both sides of the coil are at 5 volts, and no current flows. When the transistor turns "on" it goes to zero volts, and you get 5 volts worth of current running through the coil, which produces an electromagnetic field that triggers the switch.... The current involved is VERY small, typically on the order of milliamps, just enough to trip the switch.

The switch can be wired any number of ways depending on the relay, but usually turns on a higher current flow, frequently different voltage, and possibly switching from DC to AC or vice versa device, which is usually not connected to the coil in any way...

Bottom line is that the power consumed by a relay is almost negligible compared to that of the device it controls, and the relay will only draw power when it is active and turning something on (or off depending on the circuit)

Gooserider

 

[Chris S]

Getting technical here nofossil, but what happens to that 007 under load? When you say becnh test, I'm assuming it's on the bench, not in the system. I'll be back, have to get my meter out.

 

[Nofossil]

Getting technical here nofossil, but what happens to that 007 under load? When you say becnh test, I'm assuming it's on the bench, not in the system. I'll be back, have to get my meter out.

Getting technical? Here? Heaven forbid!

I actually tested it with approximately 0' head as well as 5' and 10'. In all cases, that's static head pressure since I don't have the ability to do anything more sophisticated. I don't have the power draw numbers at 5' and 10', but I don't think they were that much different.

 

[pybyr]

Relays have coils that draw power when the relay is activated - typically less that a watt.

My bench tests on a Taco 007 showed 78 watts at full speed and 0' head (or as close as I can get on my test stand). Interestingly, when I connected it to my NFCS variable speed controller, the power dropped to 65 watts at full speed, with the same flow rate. Maybe it's a measurement artifact, but interesting nonetheless.

At 50% power, it drew 38 watts and delivered about 1/3 of its rated flow.

This is interesting-- I'm curious, though- I am assuming these measurements were on the motor side of the speed control? If I'm right about that assumption, I'm curious how many watts were going into the speed control circuit (I'm assuming some inherent inefficiencies, even if modest, in the speed control's circuitry)

 

[pybyr]

I had this discussion today in th esupply house. The supplier is trying to sell me teh Grundfoss Alpha ( ecm), I am using Wilo stratos- also ecm. He claimed that he Alpha can be plugged into the wall and will run around 8 watts at idle if you will, and ramp up as zone valves open without any input or actual connection to the boiler.
I questioned running 8 watts continuous vs letting the boiler control turn the pump on & off. He answered that if a trnsformer could be eliminated the electrical usage would be equal or less with the alpha. I'll field verify this to be sure.
I do know that the Wilo circs quite often run at less than 25 watts when they're pumping, so electrical savings is potentially big. A star 16 is 88w, a Taco007 is a little more than that.

FWIW, your supplier's claim made me curious, so I just went down cellar and connected an extra Honeywell 24 V 20VA transformer to an old appliance cord, and with an AC ammeter on it- which read 0.0 when the transformer was energized but with no load placed on the 24VAC side. I'm sure that there are some losses leading to some minor draw on the AC input, but from what I saw, it seems unlikely that it's 8 watts worth. Seems to me that smart/ ECM circs are worth a strong look, but not to the point that one leaves them running 24/7/365.

I believe that if you put a load on the transformer, you will find that you have some significant energy losses in the transformer itself... Also there are some losses even just with a bare coil - notice that the transformer will get somewhat warm as it sits there, (and again, more so if it is under load) - that heat is due to power being "wasted" as it goes through the transformer.

This is why the green folks urge you to pull the plug on "wall-wart" transformers used for charging cameras, cell-phones, etc. when the thing they are nominally powering is off or not being charged....

It gets very tricky when trying to measure the power draw of either a high inductance load like a transformer, or a high capacitance load as you get a phase shift between the current and voltage that will confuse most standard instruments that are designed on the assumption that I and E are in phase...

I would tend to guess that your transformer is probably radiating around 8-10 watts worth of heat due to internal losses, so the supplier claim doesn't sound that far off... Of course, unless there is some reason not to do so, I would be inclined to run the pump through a relay of some sort so that I could turn it to ZERO when it wasn't being used, but it would sound like doing longer runs at a lower flow rate might be a good money saver during the times when you can do so...

Gooserider

Agreed on all of the basic principles you mention above- losses in the iron core, windings, reactive power, etc., some of which are indeed present even when there's no load on the transformer output-- and the cumulative effects of "phantom load" from transformers and other power supplies that are hooked up full time even if their output is not in use -- but I'm still of the opinion, until I see some hard data otherwise, that you'd have to have a larger than usual, and less efficient than normal, 24 volt transformer to be getting 8-10 watts of AC draw under no-load conditions.

Another point, that I didn't articulate before, is that there's usually a 24 volt transformer somewhere in many/ most residential HVAC installations simply to feed the thermostat(s) in the conditioned space (and relays in whatever devices the thermostats turn on to activate heating or cooling) and that transformer would still need to be there even if you did the "plug the circulator into the wall" idea that the salesperson apparently mentioned. Once you have the transformer there, with its 24 v output available, why not also use it with relays to turn off circulators when they have no reason to be running?

 

[Gooserider]

(SNIP)
Agreed on all of the basic principles you mention above- losses in the iron core, windings, reactive power, etc., some of which are indeed present even when there's no load on the transformer output-- and the cumulative effects of "phantom load" from transformers and other power supplies that are hooked up full time even if their output is not in use -- but I'm still of the opinion, until I see some hard data otherwise, that you'd have to have a larger than usual, and less efficient than normal, 24 volt transformer to be getting 8-10 watts of AC draw under no-load conditions.

Another point, that I didn't articulate before, is that there's usually a 24 volt transformer somewhere in many/ most residential HVAC installations simply to feed the thermostat(s) in the conditioned space (and relays in whatever devices the thermostats turn on to activate heating or cooling) and that transformer would still need to be there even if you did the "plug the circulator into the wall" idea that the salesperson apparently mentioned. Once you have the transformer there, with its 24 v output available, why not also use it with relays to turn off circulators when they have no reason to be running?

I think you are underestimating the power lost in that transformer - 8-10 watts is NOT a lot of power on a large scale - if you work it backwards, at 110VAC, 8-10 watts = 70-90mA which isn't a lot of current to be dropping... (I=P/E)

On the 24v transformer, you are right that it is hard to avoid having one, but the question is how big does it need to be? The fewer things you have hanging off of it, the smaller the transformer you can use... However I agree that it also makes sense to turn off circs that don't need to be running - I'd probably want to use a latching relay just on the grounds that I wouldn't want to be spending power to hold the relay in either state... The only question is if there is some reason why those circs shouldn't be turned off, or if one would spend more money on the hardware to turn them on and off, as opposed to the juice used by leaving them on....

Gooserider

 

[pybyr]

I think you are underestimating the power lost in that transformer - 8-10 watts is NOT a lot of power on a large scale - if you work it backwards, at 110VAC, 8-10 watts = 70-90mA which isn't a lot of current to be dropping... (I=P/E)

On the 24v transformer, you are right that it is hard to avoid having one, but the question is how big does it need to be? The fewer things you have hanging off of it, the smaller the transformer you can use... However I agree that it also makes sense to turn off circs that don't need to be running - I'd probably want to use a latching relay just on the grounds that I wouldn't want to be spending power to hold the relay in either state... The only question is if there is some reason why those circs shouldn't be turned off, or if one would spend more money on the hardware to turn them on and off, as opposed to the juice used by leaving them on....

Gooserider

Not to get totally sidetracked on transformers, but as this discussion emerged here, I looked at the specs-- the "classic" Honeywell 24 v transformer that has been used for years is rated at 20 VA output (for practical purposes, lets treat VA as watts, even though I realize that there are distinctions). Newer/ larger 24v standard residential HVAC transformers are rated for 40 VA output. Even the basic-standard-technology E-I pattern laminate core transformers are pretty efficient (90+ % is apparently not unusual) from anything I've previously read or recently tried to re-check.

If you're going to do a system that includes zone valves- which I expect would probably be the case if someone is using a variable-speed circ (the ability to modulate the pump does not seem to be likely to be useful if you're doing a pump-controlled primary-secondary approach rather than a valve-controlled approach) then I expect that the electrical demand of a relay or relays to shut circs off when not needed is going to be really small & inconsequential compared to the electric demand of multiple zone valves.

 

[Nofossil]

If you're going to do a system that includes zone valves- which I expect would probably be the case if someone is using a variable-speed circ (the ability to modulate the pump does not seem to be likely to be useful if you're doing a pump-controlled primary-secondary approach rather than a valve-controlled approach) then I expect that the electrical demand of a relay or relays to shut circs off when not needed is going to be really small & inconsequential compared to the electric demand of multiple zone valves.

If you're really concerned about power consumption, you're almost certainly going to be using zone valves in a supply / return configuration unless you have radiant heat. The circulator per zone approach and the primary / secondary design both use dramatically more power - hundreds of watts during operation. Transformer loss is inconsequential compared to that.

Here's some relative numbers just to put it all in perspective.

Activated relay: 0.3 - 0.8 watts

Not bad. Ten to twenty activated relays would draw about the same power as either of these:

20VA 24V transformer at 70mA standby current: 8 watts
Old-style Honeywell zone valve, when open: 8 watts

8 watts 24/7 works out to 70kWh for a full year, or about 30kWh for a five month heating season. At $0.12/kWh that's about $3.60.....

Now for the killer:

Taco 007 circulator: 80 watts

Every one of these that you have costs you 300kWh (or $36) per heating season, divided by the percentage of time that it runs. And that's the 007 - a pretty small model. I look at these primary / secondary systems with a big primary circulator, where you might average three or four circulators running all the time and I wonder. It looks to me like it would be pretty easy to be looking at $2000 or more in power consumption during the 20 year life of the system, just to run the circulators.

 

[Gooserider]

That is my biggest issue with a lot of the stuff that I see Siggy and friends putting out in the way of designs - probably very nice for comfort, but not so great for electrical power conservation...

I'm trying very hard to minimize the number of circs that I have in my setup design, and then reduce the number that need to be running at any given moment, and / or make them system responsive in some way to control the power demand. Looks like I'll end up with a bunch of zone valves, but that isn't as bad... What I'm kind of surprised is that nobody seems to have a zone valve design that only draws power when changing state... i.e. motor to open, then turns off with the valve staying open until it gets a signal to close, at which points it motors to close, and turns off until it gets a signal to open, etc. It doesn't seem to me that it would be any harder / more expensive to build than the current style which seems to require constant power to maintain it's active state, and does a spring loaded reversal if it looses power...

One of the questions I have about these temperature sensing pumps, is whether they have a readily controllable external input as to what temperature you want them to hold at any given moment... (You might want a different temperature if charging the DHW tank vs. feeding the house loops...) Also can it be controlled as to which way the pump speed changes in response to a temperature demand change? Will have to look at the spec sheets... It might even be easier if I either used an NFC motor-controller on a standard pump, or (if they make them) an ECM pump that just has a remote speed setting function and use a NoFo controller to adjust it's speed depending on need...

Gooserider

 

[Nofossil]

What I'm kind of surprised is that nobody seems to have a zone valve design that only draws power when changing state... i.e. motor to open, then turns off with the valve staying open until it gets a signal to close, at which points it motors to close, and turns off until it gets a signal to open, etc. It doesn't seem to me that it would be any harder / more expensive to build than the current style which seems to require constant power to maintain it's active state, and does a spring loaded reversal if it looses power...

Gooserider

The Taco EBV does exactly what you suggest. The problem with zone valves is that the signal to close is the removal of power. With no power, how do you close the valve? Hence the spring. The Taco EBV stores enough power to close the valve after power is removed. It consumes essentially NO power, open or closed. I love them.

 

[Gooserider]

What I'm kind of surprised is that nobody seems to have a zone valve design that only draws power when changing state... i.e. motor to open, then turns off with the valve staying open until it gets a signal to close, at which points it motors to close, and turns off until it gets a signal to open, etc. It doesn't seem to me that it would be any harder / more expensive to build than the current style which seems to require constant power to maintain it's active state, and does a spring loaded reversal if it looses power...

Gooserider

The Taco EBV does exactly what you suggest. The problem with zone valves is that the signal to close is the removal of power. With no power, how do you close the valve? Hence the spring. The Taco EBV stores enough power to close the valve after power is removed. It consumes essentially NO power, open or closed. I love them.

Those do look like nice units and would work well for most of my design so far - however I did have a couple of places where I was looking at three-way zone valves to control whether water went through a heat exchanger or bypassed it... The Taco listing only showed two way valves, though they may have similar design three ways somewhere else on their site.

I like the idea of using a cap to power the return cycle - I'd been thinking a bit more of doing a different wiring / signalling style - i.e. power one way to open and the other to close, or possibly a signal wire to tell the motor which way to turn...

Gooserider

 

[pybyr]

Ran across this very content-rich slideshow by Sigenthaler-- so far I've only gone through the segments on ways of better understanding and using hydraulic separation, and the need to maximize efficiency of circulators and reduce over-pumping (some interesting observations in this-- and it looks as if the next few years will be a time of considerable innovation in circs for the North American market):

http://www.slideshare.net/cooppower/siegenthaler-presentation-small

 

[Gooserider]

I agree, it's a very interesting presentation, though I'm only about half way through it so far, and it looks like he is starting to get the idea of reducing circ count - and otherwise paying really serious attention to electrical efficiency as well as the water side design...

I do like the looks of the new stuff that he is showing as well. The point he makes is excellent that we are essentially pushing the limits on what can be achieved for efficiencies on the combustion side of things, and getting the heat from the firebox into the system. Where we really need to start looking now is how to move that heat to where it is needed most efficiently.

Gooserider

 

[pybyr]

Goose-- since you seem interested in perhaps blazing new trails in both topology and technology - here's an idea I had a while back, but that I don't have the experience/ depth to run with and try out -- what if one ran valves with 'stepper motors' (with the valve shaft either direct-driven from the stepper motor or via cogged belt and pulleys); also include some sort of "position verification" (maybe an end switch so that the valve can be run to one full position, then signal to control hardware and software that it's reached that setting, and then stepped back there in repeatable increments to the desired setting. In what I'm envisioning would be that each valve would have an onboard 'driver' circuit for these purposes of running the stepper motor, and all of the valves could then be tied together & back to a main controller by some form of shared serial cable. Due to stepper motors' precise control-ability and repeatability, the valves would not be limited to "open or closed" but could be modulated in between for partial flow settings where that's useful (such as radiant floor feed or boiler return protection). Each stepper motor on each separate valve would have a way of distinctly identifying itself to a main system controller - much as one-wire temp sensors can be strung together and share one connection back to a controller. In fact, then, perhaps use the same serial "backbone cable" for connecting various sensors for thermostats and sensors for heat sources and heated zones. Use the same serial cable (or a conductor in the same bundle) to make DC available for each stepper motor's controller. Maybe also develop and include some remote AC relay-like devices that could also be controlled by the same serial bus but be located out at circulators to turn them on and off (or, better still, vary their speed as needed). Anyway, food for thought, I hope.

 

[Gooserider]

Goose-- since you seem interested in perhaps blazing new trails in both topology and technology - here's an idea I had a while back, but that I don't have the experience/ depth to run with and try out -- what if one ran valves with 'stepper motors' (with the valve shaft either direct-driven from the stepper motor or via cogged belt and pulleys); also include some sort of "position verification" (maybe an end switch so that the valve can be run to one full position, then signal to control hardware and software that it's reached that setting, and then stepped back there in repeatable increments to the desired setting. In what I'm envisioning would be that each valve would have an onboard 'driver' circuit for these purposes of running the stepper motor, and all of the valves could then be tied together & back to a main controller by some form of shared serial cable. Due to stepper motors' precise control-ability and repeatability, the valves would not be limited to "open or closed" but could be modulated in between for partial flow settings where that's useful (such as radiant floor feed or boiler return protection). Each stepper motor on each separate valve would have a way of distinctly identifying itself to a main system controller - much as one-wire temp sensors can be strung together and share one connection back to a controller. In fact, then, perhaps use the same serial "backbone cable" for connecting various sensors for thermostats and sensors for heat sources and heated zones. Use the same serial cable (or a conductor in the same bundle) to make DC available for each stepper motor's controller. Maybe also develop and include some remote AC relay-like devices that could also be controlled by the same serial bus but be located out at circulators to turn them on and off (or, better still, vary their speed as needed). Anyway, food for thought, I hope.

Interesting idea, and sort of vaguely what I was thinking in terms of, but I was hoping there would be off-the-shelf hardware that I could do it with... One of the things I'm concerned about, is that for better or worse, I have to get whatever I build past the local code inspectors, and I suspect that will be harder with home-brew approaches than with stuff that I can show him spec sheets w/ UL labels for...

Gooserider