Grundfos ALPHA2 AUTOADAPT function

Harold

I am choosing a pump for a rebuild of my heating system. Narrowed down to a Taco VR1816 or the ALPHA2. The existing pump is a fixed speed way oversized for actual use. Less that 40% of the total installed heating zones are actually ever used. The system has noise when the loop valves open and close. I would like to mediate that.

The Taco and Alpha2 are similar in basic function. Same controls for flow. The Alpha2 has one more function that looks like it could be useful. That is their autoadapt. I would be using constant pressure with either pump. Each of the pumps lets you set one of three speeds for constant pressure. The autoadapt has another option which is a smart control to select the proper operating support. I have not found a detailed explanation of the autoadapt except that it is "a real good thing".

The existing secondary loop pump is way over sized. There are a lot of zones in use, but they tend to be fairly random. If the Autoadapt selects "speed" and constant flow, that would seem like a valuable function. If I am understanding, rather than having to pick one of three levels, this would let the set points move about on the pump curves.

I am looking for some feedback on the Alpha2. Is that extra feature actually useful (like it seems)? Any opinions or experience with the pump solicited. Is it agile enough to deal with constant pressure and speed across the available pump speed range.

Zman

I have more experience with the Alpha than the Veridian. I think the question as to which would be better would be better answered if you posted more details of your piping.

I will say that for the typical radiant system, the "auto adapt" feature is just a made up figment of the marketing department's imagination. I have yet to meet anyone from Grundfos who can tell you how it really works.
My experience has been that In auto adapt mode with the integral check installed, the alpha has a difficult time "seeing" that a smaller zone (one or 2 radiant loops) has opened. It won't ramp up until another larger zone opens. It may work fine with baseboards as there is more flow for it to detect. I usually just switch the alpha and put it in constant pressure mode.

Harold

Thanks - just the kind of info I was hoping for.

All my heating is in the floor Pex. Lots of loops/zones and valves.

The documentation is a bit like TASMO (then a small act of magic occurs).

NY_Rob

I also tried the Autoadapt mode on my Alpha (three baseboard zones) and with just one zone open it undersupplied the heat exchanger at less than 1GPM. I now use CP1 speed with excellent results.

Harvey Ramer

Auto Adapt is best suited to a system that has a dynamic k-value. Such as a system with TRV's or Modulating electronic valves.

Here is an article that might help get a clearer picture of Auto Adapt.
https://www.phcppros.com/articles/7804-when-old-tech-meets-new-tech

Harold

Zman - If I don't cover what you want; ask.

My system is all fairly small loops valves operated by thermal actuators. 20 on the first floor. There are again that number in the basement, garage, and a potential greenhouse. None of these is used. And will likely not be while we own the house. The house is foamed in all external surfaces, the attic, and under the slab. Triple pane windows. Air circulation system.

The system is being rebuilt for various reasons. The new boiler will be a WM Evergreen 110. Replacing an Ultra 105.

The primary loop on the boiler services a DHW tank fed by a circulator. The primary loop also feeds a large buffer tank, that has a circulator feeding from the primary loop. Everything is closely spaced Ts.

The secondary loop has a circulator (the new one under discussion). Seven major Pex pipes to the various manifolds. This pump is currently a UP26-64 F. It is sized for the entire system. As I mentioned, we use maybe 40% of the actual house heating system. That is the pump I am replacing. I may also replace the circulator for the buffer tank. I will probably not touch the DHW circulator.

Controls are an old Tekmar system (T1 or 2, I don't remember - the TN4 was released just after I bought the system). This has the thermostats and controls everything in the boiler room. Some effort will be required to coordinate the Tekmar with the new boiler. The boiler does some of the things that the Tekmar does now. The Tekmar may wind up with little more than the thermostats.

I am also looking at a very unique replacement for ailing manifolds. It is the Cross Manifold units (http://blog.supplyhouse.com/cross-manifolds-are-completely-clever/). They seem to be quite new, and do not appear to have wide use (at least Google thinks that). I see a high risk here. But they resolve the issues with decaying proprietary internal parts from each manifold manufacturer.

Zman

Sounds like the Alpha 2 would work well. I would try it on CP2.

SuperJ

@Harvey Ramer your article is pure gold. My system works in a similar way. (I think I disagree a bit with the buffer tank comments though. )

AutoAdapt works well for me with TRV panel rads. It's basically an auto sensing proportional pressure mode. Grundfos has a pretty detailed white paper floating around, showing the math behind it.

For problem systems, it would be nice if Grundfos offered manually selectable proportional pressure curves like Taco though. Grundfos only offers selectable constant pressure curves on the Alpha2.

I'm curious about the "not seeing" one zone open scenarios mentioned. Is the symptom inadequate heat in the zone, or inadequate flow to keep the boiler happy? With my TRV's/buffer tank the flow does drop what off in warm weather (when just my basement office needs heat), but for me it's a good thing. I get a huge temperature delta on the rad (leaving water feels like it's basically at room temperature), healthy boiler cycles that are an hour apart, and a nice constant heat from the rad (no pumps or valves cycling to interrupt the carefully metered flow of heat into my basement office). I can sit next to the window and have the heat loss perfectly balanced instead of feeling cold until the heat cycles again.

Harvey Ramer

SuperJ said:

@Harvey Ramer your article is pure gold. My system works in a similar way. (I think I disagree a bit with the buffer tank comments though. )

AutoAdapt works well for me with TRV panel rads. It's basically an auto sensing proportional pressure mode. Grundfos has a pretty detailed white paper floating around, showing the math behind it.

For problem systems, it would be nice if Grundfos offered manually selectable proportional pressure curves like Taco though. Grundfos only offers selectable constant pressure curves on the Alpha2.

I'm curious about the "not seeing" one zone open scenarios mentioned. Is the symptom inadequate heat in the zone, or inadequate flow to keep the boiler happy? With my TRV's/buffer tank the flow does drop what off in warm weather (when just my basement office needs heat), but for me it's a good thing. I get a huge temperature delta on the rad (leaving water feels like it's basically at room temperature), healthy boiler cycles that are an hour apart, and a nice constant heat from the rad (no pumps or valves cycling to interrupt the carefully metered flow of heat into my basement office). I can sit next to the window and have the heat loss perfectly balanced instead of feeling cold until the heat cycles again.

The problem with the buffer tank is it interrupts the indoor feedback loop to the boiler. In a direct piped system as I described, the boiler receives indoor feedback via the variable flow rate.
If a buffer tank is absolutely needed in a system like this, I would suggest piping it inline between the system return and the boiler return. This would add mass to the system while not disrupting the feedback loop.

TrevorLambert

I have two Alpha2 pumps for my hydronic floor heating system. I was encouraged to get them by the supplier, but from what I'm reading here it seems like they don't do anything more than a cheaper pump would do. Each pump is serving a separate zone, which has no demand sensitive valves. There are manual valves, of course, but none that operate in response to changing conditions. Am I correct in my interpretation that the Auto Adapt feature is pointless in this kind of system? I'm looking to get another pump to circulate water/glycol through an underground passive geothermal loop, I'm guessing another Alpha2 would be a waste of money? The supplier who sold me the first ones told me they use much less energy than a conventional pump, but is that the case in a simple system with an unchanging load?

NY_Rob

^ Even though you're not using the DP or Auto features you can save on electricity costs with an ECM pump vs. a standard pump.
For example my Alpha runs at approximately 20 watts to push 8.4gpm, my Grundfos 15-58 (non-ECM) pump on speed 1 pushing 8.6 GPM burns approximately 60 watts.

SuperJ

ECM pumps are much more efficient even when running in constant curve mode. In your application you would select one of the fixed speeds. I don't think the Alpha2 is disproportionately more expensive than any other comparable ECM pump.
You're right than AutoAdapt isn't really suitable for your specific application, but the Alpha isn't limited to just AutoAdapt.
A lot of systems with a pump on each zone are over pumped, sometimes going with a single delta P pump and zone valves can be more efficient.
If you end up micro zoning your circuits at the manifolds (assuming that's relevant to your system) in the future auto adapt may have some value.

Harold

I have been perusing pump curves and reading reviews. I finally believe I am going with some Taco VR1816 rather than the Alpha2. I am just not getting what the actual advantage the ALpha2 has.

However, while doing research, I have come across a third alternative that looks interesting. The AquaMotion AM55. From what I read, the company is an offshoot of a big manufacturer. They have only been around for 2 or 3 years. But I am intrigued by the pump curves. They also seem to have a pretty robust pump.

I believe that is what I am going to use.

Harold

I have been perusing pump curves and reading reviews. I finally believe I am going with some Taco VR1816 rather than the Alpha2. I am just not getting what the actual advantage the ALpha2 has.

However, while doing research, I have come across a third alternative that looks interesting. The AquaMotion AM55. From what I read, the company is an offshoot of a big manufacturer. They have only been around for 2 or 3 years. But I am intrigued by the pump curves. They also seem to have a pretty robust pump.

I believe that is what I am going to use.

Harold

I have been perusing pump curves and reading reviews. I finally believe I am going with some Taco VR1816 rather than the Alpha2. I am just not getting what the actual advantage the ALpha2 has.

However, while doing research, I have come across a third alternative that looks interesting. The AquaMotion AM55. From what I read, the company is an offshoot of a big manufacturer. They have only been around for 2 or 3 years. But I am intrigued by the pump curves. They also seem to have a pretty robustly built pump. They have a wider range of head and flow than the other two I had been looking at.

I believe that is what I am going to use.

Harold

I have been perusing pump curves and reading reviews. I finally believe I am going with some Taco VR1816 rather than the Alpha2. I am just not getting what the actual advantage the ALpha2 has.

However, while doing research, I have come across a third alternative that looks interesting. The AquaMotion AM55. From what I read, the company is an offshoot of a big manufacturer. They have only been around for 2 or 3 years. But I am intrigued by the pump curves. They also seem to have a pretty robustly built pump. They have a wider range of head and flow than the other two I had been looking at.

I believe that is what I am going to use.

SuperJ

The biggest difference is the Grundfos in AutoAdapt automatically selects a proportional pressure curve, on the Taco VR1816 you can select from one of several fixed proportional/constant pressure curves.

This can be good and bad. You can set the Taco to match your design and forget about it. It will work great. The downside is you MIGHT be leaving a bit of efficiency on the table. It is possible the proportional pressure curve you select on the pump tracks your system close enough that there is no value to optimizing further.

The Grundfos can respond to a changing system curve and automatically adjust it's target proportional curve accordingly. The downside of this, is that anecdotally it appears that if may be possible to be fooled by outlier some conditions (low flow zone with high head loss??).

They are both good pumps, and the Taco gives you a bit granular control. So there is no bad choice in this case assuming the pump can meet your system design flow and feet of head (at the same time). I wouldn't hesitate to use a VR1816, but would be a little hesitant to use an unproven pump.

hot_rod

Harold said:

I have been perusing pump curves and reading reviews. I finally believe I am going with some Taco VR1816 rather than the Alpha2. I am just not getting what the actual advantage the ALpha2 has.

However, while doing research, I have come across a third alternative that looks interesting. The AquaMotion AM55. From what I read, the company is an offshoot of a big manufacturer. They have only been around for 2 or 3 years. But I am intrigued by the pump curves. They also seem to have a pretty robustly built pump. They have a wider range of head and flow than the other two I had been looking at.

I believe that is what I am going to use.

That is a fine pump, I think the AquaMotion team actually developed the technology 30 years ago, they have a deep bench with many years of experience in circulator development and manufacturing.

Harold

I have gone through a whole bunch of Caleffi tutorials. They are excellent and a real contribution for Hydronics learning. I am, as I am sure you have noticed, not a pro. But, I have been studying.

I have gotten the files for the original system design. The head for the entire system is about 7 feet and the flow about 12 GPM. Looking at just the main floor, it is 6 GPM and 6 ft. I am going to try to append a set of curves to this post.

Looking at the published curves for the AquaMotion AM55.

There are fixed and variable speed options. I believe I want variable. An excerpt from the manual regarding variable speed "The pump follows the proportional pressure curve that has been selected. The differential pressure is controlled electronically by changes in the input current."

Looking at the curves for the AM55, it appears to me (but here is where I need some pro input) that the pump should work in my system. It can satisfy the full system requirements. In either fixed or variable settings. The top fixed curve seems excessive for the system. The first variable one down on the graph would appear to completely serve the system, but seems likely to use excessive power. If it could handle a low initial head, the third setting down also looks possible, depending on how it behaves at low head/flow. I just don't really understand how they might behave. If nothing else, I can fool with the curve selection and see what happens. They have a display that can show speed and wattage used. But that information would not seem useful for monitoring how things are flowing in the system. I think it would be a lot more useful for tuning things if they displayed something like head and gpm. Their microprocessor should be able to easily provide that information from the monitored current it used for the existing display. But they don't. And I don't really want to add a couple more expensive pieces to the system to monitor things like that. Or figure out how to convince their microprocessor to tell me what I want to know.

OK. You have had another excessively long question. I try to be as complete as I can.

I look forward to input.

hot_rod

If the system needs 12 gpm at 7', i don't see how this circulator fits? or are you considering two zones and two circulators?

In Idronic 16 we show how to develop a system curve. Plot the system curve over the pump curve to get the OP operating point, and see how well the circulator matches the system demand.

Here is an example of two fixed speed circulators, a flat curve, and steep curve, both providing about the same flow at @ 10'.

The next slide 6-3 shows how a flat and steep curve circ would perform in a zone valved application, 3 blue lines are system curves with valves 1,2, & 3 on. You can clearly see how the flat curve option is the better choice with on/ off zone valved systems.

lastly 6.4 shows how a perfectly flat curve would be ideal. Differential bypass valves or proper V/S circulators will get you close to that ideal condition.

idronics 16 will give you a good understanding how circulators behave and how to best select one.

SuperJ

I'm not trying to push a certain pump, but your 12gpm@7fthd is within the AutoAdapt range of the Alpha 2.



And is almost within range for the Taco VR1816 on the low proportional pressure curve. You may be able to tweak your design to make it work (a bit more delta T, with a higher SWT, and lower flow in each circuit?)

Harold

HI. A bit more information on what I am thinking. I checked about the pump graph. The curves for the variable speeds are actually the curves that will be followed by the processor in the pump. So, worst case, any of the top 3 curves would work for 10 gpm at 7 ft. And might be usable for the whole system since the total system design is for 12 gpm at 7 ft. Depending on what valves are open. For just the main floor (which will be my only use) it is 6 gpm and 5.7 ft. If I have not screwed up, that would be fine. I intend stubbing out for a second parallel pump if the rest of the system needs to be activated and needs the extra flow.

I don't have direct information, but I suspect the variable curves are to conserve energy at the low end. Which makes sense to me. Some experimentation with the installed system will be called for to see what happens at the low end. But I have the feeling that the manufacturer has thought this out.

I have seen the Taco data, and it almost gets to the required operating point for all of the three modes. It was actually my reference device while doing this. The ALPHA2 could hit the design point. At fixed speed. I remain reluctant to use their auto control. I do not understand how or why it functions.

I have a differential pressure sensor across the circulator in the existing system. I will look at using that for something.

I have also been told that I should have a relief bypass between the source and return of the secondary loop. Ostensibly to prevent damage if something fails. But I don't want that device interfering with the smart pump. And If I maintain the continuity of the secondary circuit (as I intend), there is really nowhere this would be useful. Sounds wrong. Any thoughts on this?

Does all of this make sense with my additional information?

hot_rod

If the Alpha fits, try it on AutoAdapt, switch to other modes if needed

Not sure about a relief? A pressure differential valve?
Should not be needed with a VS circ

Big Ed_4

The relief bypass was used on single speed pumps to prevent dead ending ..

NY_Rob

Harold said:

I have a differential pressure sensor across the circulator in the existing system. I will look at using that for something.

Well, don't be too disappointed when you see the Alpha's displayed GPM flow rate is significantly different from your actual direct/observed flow rate.
With that in mind... use the Alpha's flow rate indicator as a relative measuring device- i.e. it indicates 2GPM on speed I with one zone open and 4GPM with two zones open... so it is adjusting flow based on head and both zones are somewhat balanced.

Harold

Any input for what I have said about the AM55? That is the one I am focusing on at the moment.

If I am understanding things (provided curves) properly, the AM55 should easily handle the main floor of the house. And I will leave stubs for a parallel unit if I ever need it for the system as designed. I have not had the tools to measure actual temperature/flow values for the main floor.

There are significant areas that are included in system sizing (e.g., garages), that will never be heated. As my wife will tell you (at the drop of a hat) I lean seriously to over doing things. There are more areas that are likely to loose heat at less than calculated rates. And has Winter South facing solar gain through glass designed to accept solar gain. The building is heavily insulated and half of the basement area is a buried foundation wall. With foamed insulation between the concrete and internal space. One smaller pump may well do for the entire house.

I am still not comfortable with the ALPHA2 (I have seen references to an ALPHA3 with extended electronics - but have not found any information). I don't like what limited information (actually, lack of) I have found about their auto adjust. What I have seen seems to say that the pump will find a setpoint after "learning" how the system is used for a week or two. But in my mind, this should be a continuous calculation/adjustment process as demand varies. Not a "learned" response. They really don't describe what goes on.

Anyway - how about the AM55?

Harvey Ramer

Auto Adapt does change it's curve settings automatically. It does it over time measured event sequences. It doesn't happen instantly.

There are a few reasons why Auto Adapt can succeed or fail in a zone valve application.
If the relationship between flow rate and head loss is well suited to the pump curve in each zone, Auto Adapt will do fine.
That, however, is often not the case. Zoned systems do not have a dynamic k-value per se. They have a fixed k-value that is staged with zone valves. The valves are instant on and off, where-as the pump uses time segmented k-value measurements to adjust it's curve. If you have a combination of low head loss zones and high head loss zones, the pump will likely forsake the requirements of the high head loss zones in favor of a more efficient operating point, created by the low head loss zones. Which zones win would depend on a complex equation of individual and combined runtimes. In a nut shell, squeaky wheel gets the grease.

Steve Thompson (Taco)

OK - I can't hold back any more ;-). And Hot Rod, I won't make a case for delta T (but I could).

My "concerns" about any auto mode circs (I am not saying they are a bad thing either - these are just old school thoughts):
- Smart devices taking over our thinking and we loose our trouble shooting logic, or general system understanding
- How long the device takes to "read" the system
- Is the device reading the system at start-up or normal operation
- Does the device need to re-read the system after a power off
- Can the proportional re-set curve go up (some devices have the curve only going down)

hot_rod

I think we are all still on a learning curve with AI. How to adapt to so many different system and components to customer expectation and variables. Even Alexa doesn't have all the answers.

The "perfect" answer or product is not yet and may never be available. What fun would it be for us if "perfect" pumps, boilers, thermostats, etc were already available

The goal for the designers and installer of smart products is to have a understanding of the application and limitations of the various choices and solutions.

I for one feel the basis laws of thermodynamics still apply and win at days end.

Harold

Just to get back to the thread participants. I am going with the AquaMotion pumps.

I really don't believe in what is being described in the Alpha2 is useful. The whole thing about it thinking about the system for a week or so and then being able to adjust to system demands is to me not believable as a useful function.

My whole system has many (around 45) small valves. With different loop lengths. I simply do not believe the Alpha2 will do something useful.

The AguaMotion appears to have a more robust construction and easier repair features. Although the likelihood of any of the pumps I have looked at failing very often seems low. I would be happier if the AquaMotion could move between curves (operating area) on demand. Maybe that would have no utility, but it would seem to me that variable speed with multiple possible curves (or working area) allowing something more like the area view presented in the Alpha2 could work to some advantage. I would also like the AquaMotion pump to use displays for parameters useful to the owner and system manager (maybe pressure and flow). Power use, by itself, is not (to me) interesting and uses up a display. The night setback feature would not be interesting to anyone with reasonably high mass systems. I have no idea how that market segments; but I won't need the function. Another display taken.

After all of this, and things I would prefer to be different, it is the AM55 pump.

Ending status for things and then we can leave this thread alone.

The very interesting manifolds I asked about. It seems to be an exceptionally clever system. Quiet, operates valves sequentially (so levels load effects, possibly reducing any intrinsic noises in the system), and uses ball valves to turn flow on and off. The problems: fairly new start-up, expensive, too long to fit in many places. A 10 valve unit needs at least 32". Speaking with the manufacturer, they did say that the manifold could be mounted vertically. This would make it feasible to put in cramped areas. In the vertical position it may not really support those (I think fairly useless) bouncing red flow rate sticks. But as a retrofit, I don't think it is worth the effort of that much repiping. But I would like to have them. It will likely exceed the maintenance free lifetime of most other manifolds.

Just an aside; the WM EVG 110 boilers seem a bit scarce.

I also had a conversation with my contractor about the ugly pipes. He suggested insulation to hide the ugly pipes. I had considered that and a can of Brasso while working on the replacement system. The bare pipes do provide a bit of heat to the unheated basement. But from a cost perspective, we are going to reexamine insulation more closely. Part of the basis for not replacing pipes is that the seepage appears to be a function of the bad Glycol (or just any Glycol), and has been stable for a long time. I don't want to put Glycol in the main system. Less maintenance and much lower chance of leaking again. To keep only water in the main system, I could add a heat exchanger to any of the freeze prone circuits and isolate them to use Glycol. Expensive; but not likely to happen.

Still have to cut and paste pipes and air supply for the new boiler. But that is straight forward.

Drains and refurbishment for air extractors so they do not do pigeon imitations all over the equipment.

A few more shutoffs and drains.

Drain equipped with acid neutralizer.

I am not sure about a device to regulate flow from source to return. Maybe something with a high pressure trigger point in excess of what is possible in case the pump goes crazy. Unlikely; but possible I suppose. And if it did exceed pressure, how would I know? No alarm or electrical shutoff on any I have seen (which is quite limited). I could see some utility for a fixed speed pump (with the vast over capacity of the one I had) But messing with flow within the active range of the smart pump would seem to me to be a bad idea. Confuses the pump.

Integration with old TekMar system. It has all of the thermostats.

And the weather gods have been watching. After weeks of very unusual high temperatures (and a couple of weeks of air rated dangerous to human life), it has now been getting colder than normal. Gee; maybe I need heat.

So, thanks for your help and insights. I really appreciate it.

And for those that know what this means: So long and thanks for all the fish.

hot_rod

I would also like the AquaMotion pump to use displays for parameters useful to the owner and system manager (maybe pressure and flow).

I like seeing the power in Watts, it gives you an idea where the circulator is operating. The GPM flow readout has not proven to be very accurate, it is a calculation not actual measurement, don't bet the farm in it.

Temperature would be a good number to display, I'd rather see that displayed. Very easy to integrate a thermistor to read that.

Harold

Displaying pressure would, I think, be useful to see the response to loops opening and closing. Flow, not so much. As noted it is a calculation from (I think) rotor power consumption. Pressure would require a sensor; so probably not a favorite of the manufacturer.