Delta T based speed control for Grundfos 3 speed pumps

[Jim K in PA]

I have read and re-read Nofossil's post of his ingenius relay mod to a Grundfos UPS 15-158. I have searched for but not found any posts with a solution to the control of the relays without a computer based solution. Have there been any "breakthroughs"?

I want to control a slightly larger Grundos pump, the UPS 26-99 and maybe a UPS 43-44. They both use the same switch control, but with different size caps. I learned about the Tekmar 157 in another post here, but thought a simpler, 3-speed/step solution may be cheaper.

Suggestions?

 

[Chris S]

The Taco 00-vdt series of pumps are supposed to be available this month, and the control is integral, with external sensors.

The Wilo stratos ecm pumps are available , same deal, different way to get there.

I was told that grundfos has their own version, but I'm not familiar with their product line.

Or did I misunderstand your question

 

[Jim K in PA]

Thanks Chris. The price of the TACO VS pumps put me off, so I went with the Grundfos UPS 3 speed units. I have learned that a Tekmar 157 should do exactly what I am looking for, and I am going to find out if the larger 26-99 pump can be controlled by the Tekmar. The pump uses the same control circuit as the 15-58, which is published as compatible with the 157. The price of the Grundfos pump AND the Tekmar 157 combined is still less than the TACO VS pump by itself, from what I have read.

 

[DaveBP]

On page 4 of this pdf is the table of circulators that are certified to work with the Tekmar 157.

http://www.tekmarcontrols.com/literature/acrobat/d157.pdf

And if you haven't run into these yet, this Tekmar pdf has some great diagrams for ways to use the 157. Just wish they had suggestions on how to hotwire it to run Gassifier/Storage setups.

http://www.tekmarcontrols.com/literature/acrobat/a157.pdf

 

[Chris S]

I don't like to buy stuff like this online, but the Wilo pump I mentioned is about $325, how much is that Tekmar control?
Chris

 

[sdrobertson]

I was looking at the taco variable speed set point pump and I ended up going with a Bell & Gossett variable speed set point as it was available and allot cheaper(Under 300.00 with shipping from Patriot Supply). I've been very pleased with this pump and it's doing everything I needed. I set it up to fast change and it speeds up and slows down pretty fast.

 

[Jim K in PA]

The Tekmar is about $225 from Pexsupply.com The Grundfos UPS 26-99 is about $240, so a total combo of about $565. I was told that the TACO VS pump would be somwhere over $600. I was not aware of the B&G;or Wilo pump alternatives. Do the B&G;/Wilo pumps include sensors and controllers? If so, it is apparent that I failed to do all my research. However, hindsight is always 20/20.

I will do more experimenting with my setup to see if adding the Tekmar will be worthwhile. I don't see it making much economic sense to tear out the almost new Grundfos to replace it with the VS Wilo or B&G;.

 

[88rxn/a]

**keeps an eye on this**.....

 

[Nofossil]

The Tekmar is about $225 from Pexsupply.com The Grundfos UPS 26-99 is about $240, so a total combo of about $565. I was told that the TACO VS pump would be somwhere over $600. I was not aware of the B&G;or Wilo pump alternatives. Do the B&G;/Wilo pumps include sensors and controllers? If so, it is apparent that I failed to do all my research. However, hindsight is always 20/20.

I will do more experimenting with my setup to see if adding the Tekmar will be worthwhile. I don't see it making much economic sense to tear out the almost new Grundfos to replace it with the VS Wilo or B&G;.

That's pretty much what I was seeing. I'm way too cheap to part with that if I can avoid it. A couple of hours and $5 worth of parts and I'm happy. Of course, it's a lot simpler when you already have a smart controller sitting there....

The Tekmar is pretty impressive in terms of capability. Wish I was enough of an Electrical Engineer to deign the solid state speed control circuit that they use inside that little thing.

 

[Chris S]

As I understand it, the Taco has external strap on sensors for supply & return temps. You set the control for your desired delta T & then the control does the rest. Not sure about pricing, but they are supposed to be available this month starting with 008-VDT around $300 I heard

The Wilo is an ECM motor, again as I understand it a "better" more efficient type of motor. This circ runs on head - you set the head you want, and then it will adjust as zones open and close. also arouns $300

Can anyone explain in laymans terms why an ECM is more efficient?
Chris

 

[Nofossil]

Can anyone explain in laymans terms why an ECM is more efficient?
Chris

It uses less energy to do the same amount of work ;-)

 

[Chris S]

Well, that would be the definition of more efficient, and... I read it on the internet soooooo
It must be true

 

[sdrobertson]

I was told that the TACO VS pump would be somwhere over $600. I was not aware of the B&G;or Wilo pump alternatives. Do the B&G;/Wilo pumps include sensors and controllers?

The B&G;circ came with one sensor and the controller already installed and ready to go. I ran the sensor about 100 feet and its seems to work well. Just opened the box, wired it up and ran the sensor.

 

[BrownianHeatingTech]

The Wilo is an ECM motor, again as I understand it a "better" more efficient type of motor. This circ runs on head - you set the head you want, and then it will adjust as zones open and close. also arouns $300

Control based upon temperature and other factors is an option with the Wilo Stratos pumps, with the addition of an auxiliary controller.

Can anyone explain in laymans terms why an ECM is more efficient?

It allows direct control of the motor by a microprocessor system, which allows fine control of the speed of the motor. The control is accomplished by changing the wave that the motor rides, rather than lowering voltage or using other power-wasting techniques. Measurement of the deformation of the waves by the motor can give load feedback to the microprocessor, which is how the Stratos pumps maintain a fixed head pressure (they actually operate based upon torque/rpm ratios, rather than directly measuring water pressure with a failure-prone transducer).

If you look up how a DC motor works, and what the commutator does, you should be able to see what I mean. Just imagine removing the mechanical commutator, and replacing it with a switch, then manually flipping the switch back and forth to change the polarity of the coils. By varying the rate at which you flip the switch, you can vary the speed of the motor, without wasting any power to do it (other than the tiny amount used by the drive electronics).

Joe

 

[Redox]

An ECM motor is basically any motor that is running on a drive supplying a PWM output. They adjust the width of the pulse and the number of pulses to match the needs of the motor at any speed. The real reason for the savings is due to the "affinity laws"

http://en.wikipedia.org/wiki/Affinity_laws

Basically, it says that at 1/2 speed, you will be moving 1/2 the water at 1/4 the pressure and 1/8 the power with a given impeller size. Multispeed motors are an economical compromise, but a drive allows a perfect match to the conditions all the time. Since you rarely need full flow, variable speed can save a lot of electricity.

Chris

 

[BrownianHeatingTech]

An ECM motor is basically any motor that is running on a drive supplying a PWM output. They adjust the width of the pulse and the number of pulses to match the needs of the motor at any speed.

That's incorrect. Pulse width modulation is how you control a typical motor, by "faking" the power to the motor in order to change its speed. This is now the Tekmar controls can vary the speed of a traditional (PSC, etc.) circulator.

An ECM is a completely different beast. I normally dislike Wikipedia as a source, but the article on AC motors is actually pretty good: http://en.wikipedia.org/wiki/AC_motor)

Joe

 

[Redox]

An ECM motor is basically any motor that is running on a drive supplying a PWM output. They adjust the width of the pulse and the number of pulses to match the needs of the motor at any speed.

That's incorrect. Pulse width modulation is how you control a typical motor, by "faking" the power to the motor in order to change its speed. This is now the Tekmar controls can vary the speed of a traditional (PSC, etc.) circulator.

An ECM is a completely different beast. I normally dislike Wikipedia as a source, but the article on AC motors is actually pretty good: http://en.wikipedia.org/wiki/AC_motor)

Joe

How is my statement wrong? An ECM is still running on variable pulses of current supplied by a drive circuit. It isn't a PWM VFD in the strictest sense of the word, but my statement was deliberately broad in order to simplify the issue for those that may not completely understand how motors or drives operate. This is what I do for a living and I've been told I'm pretty good at it.

I was trying to make another point about variable speed/flow as I was rushing out the door this AM. I am just trying to be helpful here for those that may not understand something as well as I do and maybe actually learn something in the process. I don't pretend to know everything and I don't have time to start or continue arguments.

Chris

 

[BrownianHeatingTech]

How is my statement wrong? An ECM is still running on variable pulses of current supplied by a drive circuit. It isn't a PWM VFD in the strictest sense of the word, but my statement was deliberately broad in order to simplify the issue for those that may not completely understand how motors or drives operate. This is what I do for a living and I've been told I'm pretty good at it.

An ECM runs on pulse frequency modulation, and they are controlled by computer interfaces (since the drive circuits are included with the pump).

The difference is pretty important in this context, because there are a lot of PWM controllers (Tekmar, etc.) designed for operating standard pumps, and folks might mistakenly try to connect one to an ECM, or think that a standard pump with PWM controller is the same as having an ECM.

I am just trying to be helpful here for those that may not understand something as well as I do and maybe actually learn something in the process. I don't pretend to know everything and I don't have time to start or continue arguments.

Hence why I said, "that's incorrect," rather than "hey, stupid, that's so wrong!" If we didn't all make mistakes, there wouldn't be erasers on pencils...

Joe

 

[foxt]

I've actually been playing with a tekmar 157 on my GARN, and have one observation to offer. The data brochure, under the "Turn On" section, states that the variable speed circ will run at 100% for a selectable period of time before operating at variable speed. It goes on to state in a different section that it does this to ensure that the system can overcome whatever head may exist from drainback systems, and to ensure that source temp water has reached the sensors. I interpreted this to mean that, when initially enabled, the control will run the variable speed circ at 100% and then switch to operating within the control paramters of the delta-T settings.

Well, it appears that what the control actually does is, when initially enabled, if the sensor temps are such that the variable speed circ should run, it will go ahead and run at 100% for the selected time and then switch to varying the speed. If the sensor temps are such that the var speed circ doesn't need to run, it never starts up. This makes a difference to me because I am trying to use the control to manage a 20* delta-T across the GARN side of my hx. I started with the source and storage sensors on the in/out of the GARN side of the hx. If the system hasn't called for heat in a while, that delta-T at those fittings is zero. Given the way that the control operates, when the system calls for the GARN circ, the control actually won't turn the circ on because the delta-T is 0.

I'm currently getting around this by placing source sensor on the GARN itself, and placing the storage sensor on the out side of the hx. I'll have to think about a way to move the sensor back to the hx, but it may not be worth the effort ....

Anyone else have any hands-on experience with the 157? Am I missing something?

Tom

 

[Jim K in PA]

I'm currently getting around this by placing source sensor on the GARN itself, and placing the storage sensor on the out side of the hx. I'll have to think about a way to move the sensor back to the hx, but it may not be worth the effort ....

Anyone else have any hands-on experience with the 157? Am I missing something?

Tom

Tom - thanks for that excellent feedback. I read the instructions and suspected that very problem would come up. If you ran your pumps 24/7 it's not an issue, but that is not efficient and not what I want to do either. I think I am going to use a DPDT relay, a timed relay, or some other manner of primary pump control on the HX side. If the 157 is the primary control, you get exactly what you found. If the pump turns on and runs for 30-60 seconds, that will be more than enough time (for my system at least) to get the warm water to the HX, and then let the sensors and Tekmar take over. I have not yet taken the time to work out the details or hardware needed.

 

[foxt]

Jim,
Yup, if I decide it's worth the trouble, I'll buy or build a relay timer and have that control the var speed circ for the first 60 seconds of every demand cycle, then switch over to the 157. I *think* that essentially replaces the need for a non-zero value in the 157's var speed startup cycle, but I haven't really thought that through yet ...

Tom

 

[Deering]

Jim,

I'm still thinking my way through this one, but instead of using a plate heat exchanger, why not go with a reservoir-type exchanger with a coil such as an Amtrol Boilermate? Say a 60 gallon WHS-60ZC. (broken link removed to http://www.amtrol.com/boilermate.htm)

This way the heat you supply from your Garn doesn't have to exactly match up with the demand - your Garn supply pump could run at a single high speed until the Boilermate demand is satisfied rather than trying to meet the more rapidly fluctuating demand from the zones in the house - just size the pump for your maximum demand. In the mornings when demand is high from the Boilermate, the Garn supply pump would just cycle on more frequently. And your zone return would be buffered by the Boilermate so your house boiler wouldn't be shocked by low temp return water. This would also give you a bit more effective heat storage capacity.

Obviously this would take up more space. I'm not sure about costs - would this be much more expensive than plate hx's and variable speed delta T pumps? It seems like it might be a bit simpler and more foolproof...like the Garn itself.

There's probably a fatal flaw in my logic somewhere, and I'd definitely appreciate hearing what it is since I'll be facing the same distribution challenges you have when I install my system.

And BTW - great photos and videos of your install! I've gone through them several times. You're an inspiration. Many thanks!

 

[jebatty]

Didn't notice this discussed, but be cautious in estimating flow rates and btuh with a multi or variable speed pump. In general pump head increases by about the square of the increase in flow. You need to calculate your pump head at a target gpm/btuh delivery rate. From this calculation you can estimate +/- gpm/btuh delivery rate as pump speed changes based on the flow chart for the pump.

For example, if your low btuh requirement was 40,000 and your pump head at 4 gpm was 6' (4 x delta-T of 20 x 500 = 40,000) , a Grundfos UPS 15-58 on LO would meet this demand. From this you cannot assume it will move 9 gpm on MED or 13 gpm on HI. Very roughly, at 8 gpm pump head would increase to about 24', or off the chart for this pump.

Based on the same initial assumption, if flow increases by about 1.4 to 5.6 gpm, pump head would double to 12', which the 15-58 would almost deliver on MED (56,000 bthu). If flow would increase by 1.5 to 6 gpm, pump head would increase to about 14', which the 15-58 would deliver on HI (60,000 btuh, and which is pretty close to maximum delivery from this pump).

 

[Jim K in PA]

Deering - The boiler mates might be able to hold some reserve Btus for transient peak demands, but the ones I (briefly) looked at the specs for don't appear to be big enough for the continuous higher loads that a baseboard system needs. The FP HX will handle high loads with low head losses if sized large enough. It is the tradeoff between transfer rate vs head loss that I am dealing with.

Jim - I agree with what you are saying, which is why piping, pumps, HXs and controls should be chosen carefully with peak demand considered, plus whatever margin of "safety" you wish to have for additional loads or lower water temps. That is why I sized my system and chose my pumps as I did. I am using a Grundfos UPS26-99 for my house loop, and that is the pump I wish to vary the speed of. Most of my demand is readily met at "medium" speed, but there are long periods of time where slower speeds are justified, and then those peak demand spikes where kicking it to max throughput would be desireable.

 

[foxt]

Something else I am seeing while playing with the 157 on the GARN circ - when allowing the 157 to vary the circ, I believe I can observe what I can only describe as horizontal stratification in the GARN: there will be a 10-15 degree delta between the front and back of the GARN. Originally I thought that I was seeing this because of a difference in the accuracy/design of my various temp sensors, but now I have ruled that out. I'm by no means a hydronics expert, and am wondering if what I am seeing can be due to the way the 157 is managing the circ.

I have the 157 set to manage a 20* delta T across the GARN side of a fp hx that mates the GARN to my existing distribution system. On warm days when the demand is light, the 157 lets the circ run at what I have set as min circ output, which is approx 10% above the circ output required to overcome the head of the GARN loop. The 157 is set to run the circ at 100% for 30 secs when it first comes on. Under high demand conditions, it usually varies between 100% and around 50% (I haven't religiously watched it, I just notice that it seems to be in this range when I remember to look at it). Under light load, it does ramp down to min and sit there. And after a while at the slower speeds, I start to notice the output temps from the GARN falling below what the front gauges read, sometimes as much as a 15 degree difference. Just for grins, I ramp up the speed of the circ to 100% and within minutes, the temps within the GARN equalize to within a couple of degrees.

Am I deluding myself, or would it be possible that with almost 2000 gallons of storage across a cylinder 10ft long and 6ft in diameter, the circ can run slow enough to setup what seems almost like a short circuit at the back of that cylinder? I'm wondering if that circ is just pushing cold return water into the GARN at the base of the rear, and sucking it right back out of the supply fitting that's only 3-4 feet above it? I took a few pictures of the inside of the GARN before I filled it, and I recall wondering why that return fitting didn't connect to a pipe that would deliver the return water to the front of the unit to ensure good mixing ...

If I am correct in my uneducated guess, then I need to find a new min setting for the circ - have to add in a factor to ensure that the return water velocity is sufficient to mix. Or, I could go back to first principles with the GARN, keep it simple, and just run the circ on/off ....

What do you guys think?

Tom