Grundfos Magna / Alpha dead heading

tuffcalc

Hi everyone. I have a magna 3 and various alpha pumps as part of my radiant system.

I hear everyone saying these pumps can be deadheaded but I don't see it. Should the pump literally be stopping?

My magna 3 in proportional pressure control still pumps when all actuators are closed. Same goes for the alpha.

What am I missing?

Jamie Hall

Nothing. They can be deadheaded, though it isn't best practice. The newer pumps won't overheat as much as older ones when deadheaded.

tuffcalc

Jamie Hall said:

Nothing. They can be deadheaded, though it isn't best practice. The newer pumps won't overheat as much as older ones when deadheaded.

This is the part I'm confused about. The pump is not stopping... it continues to pump at 2050rpm (in the case of my Grundfos Magna) when all valves are closed. I figured the pump would spool down and/or stop?

Jamie Hall

It has to keep spinning, so that when a valve is opened again there will some flow and pressure drop, so that it can sense it and ramp up again (that's assuming a delta P setting). However, with the ECM drives, it will be consuming very very little power in that condition.

Steve Thompson (Taco)

Dead heading means the pump is running but all valves are off.

Standby mode is where the main power stays on to the controller but the power to the motor windings is off - hence no impeller rotation.

It's a fairly common question if you can deadhead an ECM self sensing pump or circ (sorry, couldn't resist) keeping in mind most have overload protection, intended to reduce the number of failures of deadheading at a high speed and power consumption - but the real answer is as usual "it depends".

The higher power consumption the quicker the water cooled wet rotor circ will heat up, potentially boiling the bearing lubrication in the rotor can (high temps can also cause iron oxide to stick to rotor and bearing surfaces).

Example: It's not recommend to dead head a 1,000 watt ECM operating at maximum, fixed speed. But if you dead the same circ, set for proportional pressure (where the speed and differential pressure decrease as flow is reduced) it should be OK (repeat should).

Short answer - if set on proportional deadheading should be OK. Constant pressure might be OK. Constant speed is not recommended.

hot_rod

Sounds like you have the Magna in the wrong setting? @steve thompson has a great explanation of how the various settings on ECM circs work.
Delta P functions circulators are used with zone valved applications all the time, plugged in to constant power.

Zman

It is hard not to see the irony of the Taco guys explaining the overly complex Magna. I love it!
IMO, the VR3452 is the best Delta P product in that size range. It has great range and you really shouldn't even need the manual to set it up. Super intuitive!
https://www.tacocomfort.com/product/vr3452-ecm-high-efficiency-circulator/

tuffcalc

Steve Thompson (Taco) said:

Dead heading means the pump is running but all valves are off.

Standby mode is where the main power stays on to the controller but the power to the motor windings is off - hence no impeller rotation.

It's a fairly common question if you can deadhead an ECM self sensing pump or circ (sorry, couldn't resist) keeping in mind most have overload protection, intended to reduce the number of failures of deadheading at a high speed and power consumption - but the real answer is as usual "it depends".

The higher power consumption the quicker the water cooled wet rotor circ will heat up, potentially boiling the bearing lubrication in the rotor can (high temps can also cause iron oxide to stick to rotor and bearing surfaces).

Example: It's not recommend to dead head a 1,000 watt ECM operating at maximum, fixed speed. But if you dead the same circ, set for proportional pressure (where the speed and differential pressure decrease as flow is reduced) it should be OK (repeat should).

Short answer - if set on proportional deadheading should be OK. Constant pressure might be OK. Constant speed is not recommended.

I'm set at proportional pressure. When deadheading, how slow should the pump spool down to? Is it almost idle?

hot_rod

There are plenty of You Tube videos explaining that circulator, how to set it. The Magna 3 even tell you what the 7 different modes are used for in the screen.
Unless you have high head zones, I don't think you want to be in proportional pressure mode, try Auto Adapt.

https://www.youtube.com/watch?v=fMf4KOdRgKE

Dave H_2

Or try putting in a switching relay, thermostat calls for heat....turns on the circ. Thermostat says "I'm done", turns off the circ.

The most efficient circ (ECM or not) is the one that is off!

Dave H.

hot_rod

Dave H_2 said:

Or try putting in a switching relay, thermostat calls for heat....turns on the circ. Thermostat says "I'm done", turns off the circ.

The most efficient circ (ECM or not) is the one that is off!

Dave H.

But you end up with a 24/7 transformer load?
40Va in many relays. Might be a toss up with the current draw in the standby mode.

tuffcalc

Dave H_2 said:

Or try putting in a switching relay, thermostat calls for heat....turns on the circ. Thermostat says "I'm done", turns off the circ.

The most efficient circ (ECM or not) is the one that is off!

Dave H.

I actually have it setup this way today... I'm running many zones on daisy chained Tekmar 306V's. The reason I'm asking is because I'd like to run the pump for a good 15 minutes after a zone closes so that it does not short cycle when various zones are opening and closing. I'm also zoning using uponor wax actuators, so sometimes they don't even trigger the endswitch before turning off. If I have the pump delay for 15 minutes it will pump into that zone while the actuator closes up (takes about 5 minutes for it to open or close).

So in my scenario above, the pump will end up deadheading sometimes (which it does)... I was wondering why the pump doesn't actually stop. I guess this is by design.

tuffcalc

hot_rod said:

There are plenty of You Tube videos explaining that circulator, how to set it. The Magna 3 even tell you what the 7 different modes are used for in the screen.
Unless you have high head zones, I don't think you want to be in proportional pressure mode, try Auto Adapt.

https://www.youtube.com/watch?v=fMf4KOdRgKE

Have read all that... here is where I am:

1. Autoadapt sends way too little flow to each loop. Around .25GPM per loop.

2. Constant Pressure (which is recommended for radiant floors) seems to use much higher power. What was concerning to me is when it is deadheading it doesn't cut power.

3. Proportional Pressure seems like a better version of Constant Pressure. I can get the same amount of flow but at lower power usage.

I called Grundfos tech support today and asked which program I should use for radiant floors zoned with thermal wax actuators. The suggestion was constant pressure, then proportional pressure, then autoadapt, then "not sure". Added to the confusion nicely

hot_rod

Guess I'm not understanding why you want the circ to run 15 minutes after the actuators close?
I think also there is a some time required for the circ to recognize all flow has been closed off, and respond.

Sounds like you have the low current draw actuators? 250- 300mA?

Unless they are in a very cold room 3 minutes till they start to open and make the end switch, 90 seconds to start closing and break the end switch. Maybe another minute or so to close completely.
What kind of load is the Magna attached to?

tuffcalc

@hot_rod

All the zones are connected to Tekmar 561 thermostats that do PWM... so I notice they are pulsing on/off in about 5 minute cycles when at room temperature. The end zone doesn't always trigger before the 561 turns the zone off.

By having the circulator stay on for at least 15 minutes, it doesn't short cycle. Previously I would catch it short cycling quite often when the is in the shoulder season. This was if it is on, even though the zone isn't fully open, it gets some flow as it closes.

Grundfos is pumping the entire home (about 40 feet up from the basement at the highest point - I have some loops/manifolds in the attic). A lochinvar WHN085 is feeding a primary loop (with an armstrong ECM pump). and the grundfos picks up from there via closely spaced tees in its own loop.

hot_rod

Keeping in mind the height of the building in a closed loop system has nothing to do with pump sizing. If the Magna was chosen based on a 40' height not gpm and head requirement, could be way too much circulator for the task at hand.
It would be unusual to see a Magna on an 85K boiler?

The fill pressure is what assures water is up 40'.
40 X .433= 17, add 5 for a fill pressure of 22 psi.

tuffcalc

hot_rod said:

Keeping in mind the height of the building in a closed loop system has nothing to do with pump sizing. If the Magna was chosen based on a 40' height not gpm and head requirement, could be way too much circulator for the task at hand.
It would be unusual to see a Magna on an 85K boiler?

The fill pressure is what assures water is up 40'.
40 X .433= 17, add 5 for a fill pressure of 22 psi.

We're bang on at 22psi. Used the same calcuation.

The magna was chosen on GPM/Height. It's slightly oversized because it can throttle down it's been working fine.

Zman

It sounds like your circulator is way too big. The height of the building has nothing to do with circulator sizing.
I like to visualize a Ferris wheel when thinking about this. The water going down on the one side pulls the water on the other side up. The circulator is sized to offset the friction loss in the pipes and components at required GPM, nothing else.
A typical properly designed residential heating system requires < 5 PSI to operate correctly. trying to move the water much faster than that is an incredible waste of energy. The system curve of a radiant system gets very steep at that point, you can try to push harder but all you do is waste electricity.
The delta t on the supply vs return is a good indication of proper flow through a radiant system. A 10 degree difference under steady state conditions is a good target. If the delta is <10 degrees you are paying too much to move the water.

Zman

What model magna do you have? The high head models may not want to turn down low enough to be dead headed.
How many radiant loops do you have? What is the tubing size and length?

tuffcalc

Zman said:

What model magna do you have? The high head models may not want to turn down low enough to be dead headed.
How many radiant loops do you have? What is the tubing size and length?

It is a Magna 40-120F.

There are 32 loops. I'm not 100% sure how long each one is. Probably between 100 - 200ft each.

It's setup for constant pressure at 11.5ft/head (5.0psi). Autoadapt really cranks it down to 4.9ft/head. I might just let it run on autoadapt for a while and see how it goes.

On autoadapt its running between 10 and 40 watts (depending on how many zones are open). On constant pressure it is running between 37 and 50watts ... so power difference is negligible.

I'm thinking it might be better to leave it on constant pressure so the zone can respond faster (more gpm = more BTU) since the difference in electricity is so small.

tuffcalc

The Grundfos manual says the following:

Constant pressure
We recommend this control mode in systems with
relatively small pressure losses.
The pump head is kept constant, independent of the
flow in the system.

Proportional pressure
This control mode is used in systems with relatively
large pressure losses in the distribution pipes. The
head of the pump will increase proportionally to the
flow in the system to compensate for the large
pressure losses in the distribution pipes.

Is the system I'm describing a high pressure loss system or low pressure loss system (the "primary loop" from the boiler runs on its own pump and feeds into the "secondary loop" which goes to all the manifolds that have thermal actuators on them ... I hope I described this right).

Zman

I would say that the constant pressure mode you are set to is probably pretty close. My experience with auto adapt and radiant is that it has trouble detecting when one smaller zone opens and underperforms at those times. If it is working for you, keep using it.
Proportionate control is rarely the right setting for radiant systems as most of the resistance is in the panels rather than the distribution piping.
Measuring the supply and return delta t is a great way to see how the system is performing. Ignore results until the water has had time to circulate for a while. A cold high mass slab will have a huge delta at the beginning of a cycle.
As with all things heating, more is not always better. A delta of 10 is great, a delta of 5 will give you almost identical BTU output will be no better in terms of comfort or efficiency.

tuffcalc

Zman said:

I would say that the constant pressure mode you are set to is probably pretty close. My experience with auto adapt and radiant is that it has trouble detecting when one smaller zone opens and underperforms at those times. If it is working for you, keep using it.
Proportionate control is rarely the right setting for radiant systems as most of the resistance is in the panels rather than the distribution piping.
Measuring the supply and return delta t is a great way to see how the system is performing. Ignore results until the water has had time to circulate for a while. A cold high mass slab will have a huge delta at the beginning of a cycle.
As with all things heating, more is not always better. A delta of 10 is great, a delta of 5 will give you almost identical BTU output will be no better in terms of comfort or efficiency.

Thanks.

I don't have supply/return gauges on the manifolds ... just ordred some now. I'll install those and see what the temps are.

SuperJ

With any of the modes you first need make sure the system is balanced for the correct flow. You can do this by temperature delta (tough on high mass systems), or by flow indicators (which can have dubious accuracy). A lot of manifolds have a somewhat hidden adjustment below the actuator to restrict flow.

AutoAdapt doesn't replace balancing, and IMHO is mostly targeted at constant circulation systems with modulating control valves (TRVs).

With infloor you're going to need similar amount of differential pressure regardless of the number zones (assuming similar circuit lengths) (but varying amounts of flow depending on the number of zones calling), Which is what constant pressure does.
If you have undersized distribution piping to your manifolds then proportional pressure might work better.

I think the magna pumps have a couple other modes as well that may or may not be useful, and some are capably of using external sensors.