I do not understand ECM circulators after all...

Jean-David Beyer

People have proposed that one should use variable speed ECM circulators instead of fixed speed ones to save electrical energy. I am all for ECM, but I do not see that all the benefit can be obtained unless they run at the lowest speeds a lot of the time. I have three questions about this.



1.) If I run a delta-T pump, how do I pick the delta T to run it at, considering I run it on a system with outdoor reset? If my upstairs zone runs with a 10F temperature drop when it is  7F outdoors, and a 1F temperature drop when it is 40F outside with a fixed speed circulator in a baseboard zone, what delta T should I set the pump to? If I set it to 10F, it would go extremely slowly when it is 32F, and slower yet if it is 65F outside. Can it go that slowly? And if it could, and the zone is two rooms with the baseboard in series, would the second room get almost as much heat as the first as it does now, or would it be colder? And  how much hotter would the water have to be if it were going that slowly? (I made up some of the numbers, but they are pretty close.)



2.) Someone said that using a delta-T ECM pump would reduce short cycling, but I do not see that. Whether I run 3.0 gallons per minute with a drop of 1F, or 0.3 gpm with a much greater temperature drop, it seems to me the boiler would be putting out the same number of BTU per hour, and if the boiler cannot modulate down to that, it will short cycle no matter what the circulator is doing, provided it is not so slow that it cannot carry enough heat up the the emitters.



3.) Someone said a delta-T ECM circulator costs only a little more than a fixed speed one, such as a Taco 00 series. Since we cannot put prices here, I looked up a Taco like I have now, and a Grundfos Alpha, and the Grundfos, at least on the Internet, is about 8 times the price of the Taco. Not including labor. Since my Taco pumps are quite new and working fine, it is difficult to justify replacing them just to protect the environment, even though the environment is important to me. If a Taco failed, I might consider one of their yellow ECM ones then, if the delta-T does not fight with the outdoor reset.

Tim McElwain

My feeling about this subject

is that all the variable speed ECM type devices on both boilers and warm air work best when matched to the system they are serving. A poor system design or an existing one that is not totally compatible with ECM technology is a waste of money. By the way we are all finding out that on some of this new stuff your return on your investment may take 5 years or more to be realized. To be "Green" can cost money.



As far as boiler use goes I would recommend an ECM type circulator with zone valves as the best way to go. Again however we must work with in the parameters of the capability of the ECM unit.



Our good friend Mr. Dave Yates has a very interesting article in the January issue of Contractor magazine Page 26 called the "Biggest Loser". www.contractormag.com

Tim McElwain

I forgot to mention

I have an entire chapter in one of my manuals dedicated to motor technology Pros and Cons.

Jean-David Beyer

Mr. Dave Yates

I read his article and thought it pretty good. Reading it, and some other things, was what prompted some of my questions. I guess the biggest one is the interactions between outdoor reset and delta-T circulators. If I were designign my heating system, and my hot water (domestic) from scratch, I would do things differently, but the piping is in the walls and between the downstairs ceilnig and the upstairs floor, and buried in the concrete slab, so pretty much unchangable. Also, the boiler (and indirect) are new and work very well, compared to my 60 year-old oil burner and electric hot water heater. So If I go from an 87 watt circulator to one that runs at 14 watts most of the time would be good, but might take longer than I have to pay for itself and its installation. I could save a little by replacing the two circulators with one and zone valves. But that is money too that would not be much were I starting from scratch.



See, when it is warm out, right now I put 110F water into the baseboards, and get 109 back. And that drop is enough. They do not emit all that much at that temperature (Slant/Fin BaseLine 2000). When it was about 7.5F outside, I put 134F water into the baseboards and got about 125F back, and that was just enough.. But if I set a delta-T to 10F, and put 110F water in, chances are the circulator would stop before it could go slow enough to get 10F drop -- but I do not know this. I could not find a delta-T vs pump speed curve on the GrundFos site, and could not find anything about the Taco yellow ones.

Jean-David Beyer

I have an entire chapter in one of my manuals

Do you remember which one?

Mark Eatherton

ECMs and Delta T versus Delta P

That in and of itself would tell me you are either sending water through the system at a SCREAMING rate, or your heating load is nearly non existent. Based on previous posts, I seriously doubt it is non existent, which tells me your pump is WAY bigger than it really needs to be.



ECM is not for every one, just MOST everyone. And the prevailing application is not in Delta T, but rather Delta P. In fact, all of the ones I am familiar with and have applied would require an additional control to make it work on DT. DT has its application, but it is not in a situation where you have variable flow requirements.



A delta P pump has the ability to see "holes" in the system and it increases its speed to fill those holes in with flow. If there are no "holes" open, then it idles back to a minimum position while maintaining a pressure differential between its inlet and outlet, but continues watching for changes in demand. As soon as a hole pops up, the pump increases its RPM's to maintain the calculated, required pressure differential.



Attempting to control these variable flow system circulators with differential in temperature is a bad idea in my estimate. No two circuits are the same, and their delta T varies as a function of load exposure. Choosing one temperature differential will get you into trouble.



The correct application of DT pumps, in my professional opinion, would be something like a GSHP Earth loop. As the load decreases, the loop filed pumps, which by the way DO chew up a lot of watts, can be idled back, awaiting for a load to show up. The same theory applies to a commercial water source heat pump loop. As loads increase, DT increases, requiring an increase in flow rate to maintain the required differential.



As it pertains to motor technology, ECM motors are significantly more efficient as it pertains to water to wire efficiency. I believe that by next year, all pump motors in Europe are going to be required to be of ECM design, regardless of whether they are fixed speed or variable speed. One of these days, that will probably be the case here on this side of the pond as well.



My favorite system is a combination of a DC ECM VS circulator, tied to a high mass/surface area emitter (ceilings, floors or walls) controlled by non electric thermostatic operators, connected to a ModCon boiler tied to a buffer tank to avoid short cycling. A symphony in motion and efficiency.



ME

Mark Eatherton

Think of it as...

Cruise Control for the hydraulic side of your system. It can see an approaching hill coming, and as it peaks the top of the hill, it knows to shift gears and back off the gas pedal :-)



Very intelligent pump. I suspect that in the near future, a person will be able to take a programmable digital controller, and fine tune the intelligence even more. If zone A calls, then required head pressure and subsequent flow would be X. If A and B are calling, then required pressure differential would be Y, and if all three zones are calling, then required pressure differential would be Z.



I've been playing with these programmable controllers, and their ability and capacity is only limited by your imagination. They could even be programmed to do an outdoor reset on steam based on steam's co-relating temperature, instead of trying to find an accurate pressure transducer. Cycle times can be calculated based on the outdoor air temperature, or the difference in deviation between the indoor set point and the actual indoor temperature. These are exciting times to be in the comfort business.



ME

Jean-David Beyer

screaming

"That in and of itself would tell me you are either sending water through

the system at a SCREAMING rate, or your heating load is nearly non

existent. Based on previous posts, I seriously doubt it is non existent,

which tells me your pump is WAY bigger than it really needs to be."



The installing contractor just used Taco 007-IFCs for everything; perhaps that is what they had in stock. Two of them were required by the boiler manufacturer (boiler circulator and indirect fired DHW), but the other two are for the two heating zones. The one you say is at a screaming rate is calculated by me to be 2.4 gallons per minute through a mix of 1/2 inch and 3/4 inich copper tubing. As far as non-existant, the heating load I calculate to be about 6500 BTU/hour when it is 0F outside, but the design day temperature here is 14F; so far this winter it has been below that only 3 times. So the heat load up there is pretty small. I think the low delta-T is due to both factors: low load and oversized circulator. It does not make noise, though, now that the air is out of there after replacing a noisy circulator.



"DT has its application, but it is not in a situation where you have variable flow requirements.'



Unless I remove both circulators and use just one and two zone valves, I have constant flow now. The flow upstairs (that we have been discussing) is clearly too fast. Downstairs (radiant slab) probably is too. The delta-T down there is higher than upstairs, but I cannot figure out what it is. It is 1/2 inch copper tubing in concrete slab, but I do not know the lengths or spacing of the various parts.



The way the system is now, there seems little point for a delta-P circulator; I would have to change the zoning to use zone valves to benefit from one of those. Right now, it seems it could be beneficial to use delta-T circulators on the two heating zones. Except for parts and labor, it would be easy to switch them (flange-mounted circulators with lots of ball valves to isolate them while changing. It would be easy to attach the temperature sensors.



I am still confused, though. I am uncertain what happens when a delta-T circulator is running with the water temperature input changing. My modulating boiler reduces the temperature of the supply water because it perceives that the heat load will go down as it gets warmer outside. If the circulator notices that the delta T has gone down (as it will because of the lower temperature), it slows down, trying to get the delta-T to go back up. At some point, will the circulator not stop completely because it cannot get the delta T back up to its set point.  Say it is 65F outside and the boiler supplies 73F water and the thermostat is asking for 69F in the rooms. Or will a delta-T circulator give up and keep running at some minimum speed if it cannot get the delta-T up to the set point?



Since my electricity bill has gone down even with 4 circulators instead of one (because the old electric water heater was removed), it is tough to tell when it would make sense to change to an ECM circulator. I would consider it if a circulator quit, or if the price of electricity went way up.

scott markle_2

Auto Adapt

I totally agree with your symphony of mechanically alignment (or whatever you called it), flow modulation by analog device is very elegant and it dovetails with Delt p pumping so nicely.



I think the future of variable speed pumping may involve integrating information about boiler firing rate, (kind of like the older style Vito 200.) Very few of these boilers (original vito 200) were set up (in this country) in the way they were probably most commonly used in europe- Direct connected to trv'd radiation with a well calibrated reset curve and no indoor thermostat.



What I don't get (and it kind of turns me off to Grundfoss for not explaining it better) is what is Auto Adapt? I'm not comfortable using something I don't understand, there are so many variable in the types of systems a pump might be connected to that it seems unlikely that one algorithm with no adjustable parameters is going to be able to make "intelligent" decisions about appropriate pump speeds. Since I'm usually incorporating other forms of "intelligence" like zone synchronization and indoor/outdoor feedback and even TRV's, it seems very unlikely that Auto Adapt is going to be able to use it's limited view of system behavior in a productive way, my fear is that it could even do the opposite, misinterpreting system behaviors based on generalized assumptions and react in a completely un-intelligent manner.



All the "pumped up" advertising for this feature insults my intelligence because it does such a superficial job of explaining how this feature actually works.



Enough of the pretty girl and human heart analogy, please give us real information, not superficial Madison Avenue hype. I'm sure this feature has it's appropriate and effective applications, but how are we as professionals to divine what this is, when the promotional materials seem to be directed at an audience that wouldn't know the difference between a pump curve and french curve. I can't help but wonder if this has something to do with a general lack of respect for the sophistication of the (American) industry.

scott markle_2

Auto Adapt

I totally agree with your symphony of mechanically alignment (or whatever you called it), flow modulation by analog device is very elegant and it dovetails with Delt p pumping so nicely.



I think the future of variable speed pumping may involve integrating information about boiler firing rate, (kind of like the older style Vito 200.) Very few of these boilers (original vito 200) were set up (in this country) in the way they were probably most commonly used in europe- Direct connected to trv'd radiation with a well calibrated reset curve and no indoor thermostat.



What I don't get (and it kind of turns me off to Grundfoss for not explaining it better) is what is Auto Adapt? I'm not comfortable using something I don't understand, there are so many variable in the types of systems a pump might be connected to that it seems unlikely that one algorithm with no adjustable parameters is going to be able to make "intelligent" decisions about appropriate pump speeds. Since I'm usually incorporating other forms of "intelligence" like zone synchronization and indoor/outdoor feedback and even TRV's, it seems very unlikely that Auto Adapt is going to be able to use it's limited view of system behavior in a productive way, my fear is that it could even do the opposite, misinterpreting system behaviors based on generalized assumptions and react in a completely un-intelligent manner.



All the "pumped up" advertising for this feature insults my intelligence because it does such a superficial job of explaining how this feature actually works.



Enough of the pretty girl and human heart analogy, please give us real information, not superficial Madison Avenue hype. I'm sure this feature has it's appropriate and effective applications, but how are we as professionals to divine what this is, when the promotional materials seem to be directed at an audience that wouldn't know the difference between a pump curve and french curve. I can't help but wonder if this has something to do with a general lack of respect for the sophistication of the (American) industry.

scott markle_2

Auto Adapt

I totally agree with your symphony of mechanically alignment (or whatever you called it), flow modulation by analog device is very elegant and it dovetails with Delt p pumping so nicely.



I think the future of variable speed pumping may involve integrating information about boiler firing rate, (kind of like the older style Vito 200.) Very few of these boilers (original vito 200) were set up (in this country) in the way they were probably most commonly used in europe- Direct connected to trv'd radiation with a well calibrated reset curve and no indoor thermostat.



What I don't get (and it kind of turns me off to Grundfoss for not explaining it better) is what is Auto Adapt? I'm not comfortable using something I don't understand, there are so many variable in the types of systems a pump might be connected to that it seems unlikely that one algorithm with no adjustable parameters is going to be able to make "intelligent" decisions about appropriate pump speeds. Since I'm usually incorporating other forms of "intelligence" like zone synchronization and indoor/outdoor feedback and even TRV's, it seems very unlikely that Auto Adapt is going to be able to use it's limited view of system behavior in a productive way, my fear is that it could even do the opposite, misinterpreting system behaviors based on generalized assumptions and react in a completely un-intelligent manner.



All the "pumped up" advertising for this feature insults my intelligence because it does such a superficial job of explaining how this feature actually works.



Enough of the pretty girl and human heart analogy, please give us real information, not superficial Madison Avenue hype. I'm sure this feature has it's appropriate and effective applications, but how are we as professionals to divine what this is, when the promotional materials seem to be directed at an audience that wouldn't know the difference between a pump curve and french curve. I can't help but wonder if this has something to do with a general lack of respect for the sophistication of the (American) industry.

scott markle_2

Auto Adapt

I totally agree with your symphony of mechanically alignment, flow modulation by analog device very elegant and it dovetails with Delt p pumping so nicely.



I think the future of variable speed pumping may involve integrating information about boiler firing rate and pump speed, (kind of like the older style Vito 200.) Very few of these boilers were set up (in this country) in the way they were probably most commonly used in europe- Direct connected to trv'd radiation with a well calibrated reset curve.



What I don't get (and it kind of turns me off to Grundfoss for not explaining it better) is what is Auto Adapt? I'm not comfortable using something I don't understand, there are so many variable in the types of systems a pump might be connected to that it seems unlikely that one algorithm with no adjustable parameters is going to be able to make "intelligent" decisions about appropriate pump speeds. Since I'm usually incorporating other forms of "intelligence" like zone synchronization and indoor/outdoor feedback and even TRV's, it seems very unlikely that Auto Adapt is going to be able to use it's limited view of system behavior in a productive way, my fear is that it could even do the opposite, misinterpreting system behaviors based on generalized assumptions and react in a completely un-intelligent manner.



All the "pumped up" advertising for this feature insults my intelligence because it does such a superficial job of explaining how this feature works.



Enough of the pretty girl and human heart analogy, give us real information, not superficial Madison Avenue hype. I'm sure this feature has it's appropriate and effective applications, but how are we as professionals do divine what this is when the promotional materials seem to be directed at an audience that wouldn't know the difference between a pump curve and french curve. I can't help but wonder if this has something to do with a general lack of respect for the sophistication of the (American) industry.

scott markle_2

Auto Adapt

I totally agree with your symphony of mechanically alignment, flow modulation by analog device very elegant and it dovetails with Delt p pumping so nicely.



I think the future of variable speed pumping may involve integrating information about boiler firing rate and pump speed, (kind of like the older style Vito 200.) Very few of these boilers were set up (in this country) in the way they were probably most commonly used in europe- Direct connected to trv'd radiation with a well calibrated reset curve.



What I don't get (and it kind of turns me off to Grundfoss for not explaining it better) is what is Auto Adapt? I'm not comfortable using something I don't understand, there are so many variable in the types of systems a pump might be connected to that it seems unlikely that one algorithm with no adjustable parameters is going to be able to make "intelligent" decisions about appropriate pump speeds. Since I'm usually incorporating other forms of "intelligence" like zone synchronization and indoor/outdoor feedback and even TRV's, it seems very unlikely that Auto Adapt is going to be able to use it's limited view of system behavior in a productive way, my fear is that it could even do the opposite, misinterpreting system behaviors based on generalized assumptions and react in a completely un-intelligent manner.



All the "pumped up" advertising for this feature insults my intelligence because it does such a superficial job of explaining how this feature works.



Enough of the pretty girl and human heart analogy, give us real information, not superficial Madison Avenue hype. I'm sure this feature has it's appropriate and effective applications, but how are we as professionals do divine what this is when the promotional materials seem to be directed at an audience that wouldn't know the difference between a pump curve and french curve. I can't help but wonder if this has something to do with a general lack of respect for the sophistication of the (American) industry.

bob_46

Scott

Scott, read this.

Brad White

ECM use on CV Systems

At first blush, I would agree that ECM use really would apply only in a dynamic system (one with TRV's on ALL radiators to provide a response-worthy change in pressure and flow).



However, sometimes things happen regardless.



In our current house we have a 1995 vintage W-M Gold series at 100 MBH input, 82 MBH output. Heat loss is around 48 MBH, oh well.



I had two Wilo Stratos Eco's in the crates already and their "curves" (areas of performance), suited two different functions, boiler circulator and the old gravity main circuit.



Setup is P/S with 8 gpm through the boiler, serving through an EarthLee u-shaped custom header Lee Brooks and her family fabricated for me. Off of this are a few direct existing fin-tube/convector zones (Taco 003 and a 006), plus the original 3" and 2.5" gravity HW main house zone.



The boiler circulator is set at 4 feet of head and moves about 8 gpm based on the temperature spread and boiler output when all zones are calling. When satisfied, the delta-T narrows to about 8 degrees on an average day and there is a Taco PC700 ODR control with differential to make the math tough. This circulator draws about 11 Watts, just under 1/10th of an Ampere which is as low as I can get anywhere.  At night, when the boiler is essentially off, this is said to drop to 50% but I have not measured that yet. This circ runs off an outside temperature sensor.



Similarly, the main house gravity circuit runs of the same ODT sensor as the boiler circulator, essentially constant circulation. A Taco 3-way iValve gives ODR control to the main zone and there are no TRV's on the radiators -at least this year. Thus nothing to respond to, yet.



Based on temperature drop, very close to 20 degrees, often 18 degrees, I am moving about 4 gpm against what has to be 3 feet of head. Absurdly low. This Wilo is drawing about 8 Watts, about as low as it can go. It too is in Auto, but I have not taken an amp reading after hours. (I actually do have a life   Do I actually know the head? No, but it is what it is. My point being, these circulators, while expensive, do draw lower amperage out of the box than a Taco 003 (same as a 006 by the catalog), or the Grundfos 25-15 which is 25 Watts.



I already owned the circulators, so the payback was not a determinant, but rather one of control and adjustment. I can still tweak these manually, but that is more for when I am bored. I am still using less electricity, but not enough to show an ROI under ten years. What I find interesting is that the PSC circulators each draw more amps than both of the Wilo's put together. Fractional, not a big deal, but it tells the future in some ways.



Mind you the Grundfos Alpha Auto-Adapt may do the same thing better, but the Wilo Eco's seem to have settled on their different duties just the same.

NRT_Rob

we don't bother with autoadapt

you don't need it. we love the constant pressure modes: predictable performance is key. if we can, the 8 watt max low fixed speed setting sure is nice too.



consider that an alpha has a MAX power draw of 45 watts. Often, it can run at half that or less.



that means you're saving 50-70 watts . If you figure a 75% run time for an outdoor reset system, that's almost 550 hours a month.



watts x 550 x your cost per KWH = monthly savings.



we're finding in many markets it may make sense to use alphas for boiler pumps.



In markets with cheap electricity, it only makes sense if it allows you to drop the PBV.

Unknown

ECM Made Simple?

You all knew I was going to pipe up with my 2 cents worth - here's the take on delta PV ECM control, the most common setting (somewhat similar to Auto Adapt).



Delta PV simply means whatever the pump is set at (10ft for example), it provides 10' of head at max speed and half of that at dead head or zero flow (5').  The important thing is to remember this pump has an inclining pump curve, from left to right, the opposite of a constant speed pump (there the head goes down as flow increases).



So, this is a benefit as a decrease in system flow shows a decease in friction loss (so it's a good thing to reduce the pumps differential head) and as zones close there is no increase in pump pressure (that causes unbalanced, high flow "on" zones) and does not require a by-pass valve.  As zones and the system are satisfied the BTU load drops, requiring fewer pounds of water to provide heat.



ECM at full speed will draw about 50% to 60% of what a standard wet rotor pump, mostly due to the permanent magnet rotor.  The HUGE savings (up to a confirmed 95% electrical savings) occur during low load and low flow conditions.



Expect to pay 2 1/2 to 3 times as much as a standard commercial pump (not including a VFD or transducer) and 3 to 4 times as much as a standard residential circ.  Electrical payback (not including system savings) I've seen has ranged from 8 months to 4 years.



The only app I know of where variable flow pumps do not make sense are constant flow applications.



I've either confused a bunch of you or helped (I hope for the latter!).

Tim McElwain

JDB

in my manual Advanced Electric Ignition Systems Vol I Chapter Five is on ECM Blower Motors in that same manual there is an addendum at the end of the manual called Motor Technology.

Nanuk

ECM Ciculator on MOD/CON Boilers

This discussion of Variable pumps did not include how the work with the modulation control of a mod/con boiler.



It is been our experience that you should have a Hydro Separator or buffer tank between the Mod/Con boiler and Variable flow through the system zones. If you have been around a while you may remember the hunting problem with 4-way mixing valve controls when both sides were not hydrolically balanced. the same thing happens with some mod/con boilers when the flow starts to vary through the heat exchanger.

 

To get consistent and accurate modulation from the boiler you need the same throughput of water over its entire modulation range. If you change the speed of the water flow, you change the exchange rate of heat into the water. if you have a variable pump working on delta  t and a boiler modulating control working on delta t one will always be playing catch up to the other. It is like juggling as one goes up the other goes down to compensate.



If you want to use a variable speed pump with a modulating boiler they need to be interfaced so that the boiler knows the throughput and modulates the burner accordingly. Some of the boilers on the market now have this technology but they are on the higher end of the price scale.

Gordan

Do tell - I've always wondered about this

Even without variable speed pumping, you can wind up with variable flows through the boiler if the boiler is not hydraulically isolated from the system. For instance, the Prestige install manuals allow for multi-zone systems that are pumped straight through the boiler; zones opening or closing would affect flow. Differential pressure bypass valves are specified to ensure that adequate flow is maintained, but that doesn't mean that the flow is constant, so I think that this issue would be present even with constant duty circulators. Which boilers or controllers have you seen to be susceptible to this confusion?

Brad White

I would clarify one sentence

Gordan. You wrote, "For instance, the Prestige install manuals allow for multi-zone systems

that are pumped straight through the boiler; zones opening or closing

would affect flow".



Opening and closing any valve on a directly-coupled system would affect flow. (It has to or why do it?). But the key as Nanuk was saying is, the flow rate varies slowly enough so that the combustion rate can adapt to it. When flow varies suddenly (as on a small-tube Giannoni HX), you will flash into steam if the combustion rate does not drop in concert.



Viessmann Vitodens matches the flow rate to the combustion rate, I believe by closely monitoring the internal delta-T and anticipating this based the rate of change. My conjecture but it makes sense.



The TT heat exchanger and the new Knight are fire tube boilers with larger vessels so are more forgiving of flow rate to combustion rate.

NRT_Rob

the problem he's talking about

if if there is feedback between variable pumping and boiler output. imagine a delta T control circ, boiler changes output, pump changes speed in response, which changes boiler output, etc.



with fixed speed pumping there is no issue: the boiler just modulates to the demand it has and you're done.



with ECM pumping on constant pressure modes w/a bypass, as long as your minimum flow is adequate, there are no issues as well.

Gordan

Thanks, Brad and Rob

I missed the "delta T" part of the variable pumping discussion.