How would you control a Bumblebee circulator in a pure TRV situation?

knotgrumpy

In either a reverse return two pipe system or a home run setup I can understand the Alpha would sense the pressure change when a TRV opens and closes.  How would a Bumblebee know a TRV opened in order to turn the pump on since you would be dispensing with thermostats?





Put in some sort of pressure differential sensing switch between supply and return?





John Barba, are you on here?

SWEI

delta-T will work just fine

As the TRVs slow down the water flow through the radiators, the average return temperature will drop and the pump will slow down to supply less, maintaining ∆T.

knotgrumpy

What if there is no longer any return flow to measure?

On a closed home run branch there wouldn't be any returning water, would there?  On the reverse return two pipe system the return would not be flowing if the only the first radiator on loop was calling for heat and then the TRV shot down.  No water would be flowing back to boiler then.



Would a bypass be required to have flow temp to measure?

SWEI

I haven't tried a BumbleBee yet

but I'm sure they have a way of dealing with it.  I just pinged Steve Thompson on this.

knotgrumpy

Thanks

I am really interested in this.



Mark

heatpro02920

I wouldn't use delta t

I would send one sensor to the radiator and use the set point function just set the temp you want at that radiator the bee will do the rest.... I'm not sure I'm thinking of the same type of system as you I use them for radiant and rad panel fhw .....

Steve Thompson (Taco)

I think I have an answer - but...

Good question.  I think I have the answer but want to check with Mr. Barba and a few others first.



Stay tuned folks - we're on it.  Get back to you shortly.

knotgrumpy

Thanks Steve!

Thanks for all the great "How To" stuff on your website. Between Taco and Caleffi (and The Wall, of course), a person can learn a lot.



Mark

CMadatMe

In A Pure

TRV situation I would be constant circulation based on outdoor temp. Pump comes on below WWSD and off at WWSD. If a panel rad application my radiator valves would be the type with built in bypasses.



The Bee is looking to maintain a delta-t and will do so. I wouldn't want it on/off. It runs at start for a full 3 minutes before it reacts unlike the VDT which was 30 seconds.



If cast rads, same thing, bypasses and constant circulation. The pump will react to the delta-t change and speed up or down depending on the needed flow and head required at a given moment.



Would be nice if the next version had 0-10VDC and the start time was field adjustable. Three minutes is a long time. Why the difference between a 30 sec on the older VDT and 3 minutes on the Bee?

Robert O'Brien

Delta P

Delta P pumps were designed with this application in mind.

John Barba

Here's the "buzz...."

Hi gang --



This is sorta what I have in the lower level of my house. Panel rads with TRV's and a Bumblebee - with a minor variation. The panel rad in the family room -- largest load by far - has no TRV - that room runs off a thermostat. All the panel rads are piped with dedicated supply and return PEX runs off a valved manifold. There's retrofit radiant floor in my office - runs off the same manifold as the panel rads - and it operates off the family room thermostat as well.



The thing to remember about TRV's is that they modulate - they're usually either opening or closing- regulating flow through the radiator. This control method lends itself to "continuous circulation," primarily for electrical simplicity more than anything else. This application is pretty standard in Europe, and is what Delta-P circulators are designed for.



Would a Delta-T Bumblebee work in this situation? Well, why wouldn't it?



Let's say you decided to run it "continuously" (which kinda defeats to the purpose of "electrical efficiency," the most efficient circulator is the one that's off - but that's a different story for a different day). You program the BumbleBee for the Delta-T you want (20, 30, 40 - whatever you designed the system for), and then put the supply sensor on the supply piping, and the return sensor on the return piping - near the return side of the manifold and the turn it on.



The pump runs to establish the Delta-T you've programmed. With all the TRV's fully open - you're pulling a lot of BTU's out of the fluid. The return water temperature may be fairly low -- and the circulator will go faster. As TRV's close, or modulate towards closed, that means the radiators - and the system - need less flow. The return water temperature would start to increase - and the Bumblebee will slow down.



So what happens when it's warm out and you don't need much heat at the radiators? Well, if you have outdoor reset on the boiler, everything chugs along as it would if it were colder out - with one exception. The BumbleBee will run at even lower speeds when it's warmer out - like in late November - than when it's colder out, like in mid-January - and there's a Taco T-shirt in it for the first person who knows why!



If there's no outdoor reset, the TRV's won't open as much - to regulate flow through the radiator. Again, since there's not a lot of BTU's being taken out of the fluid, the Bumblebee will react accordingly. Will the pump dead-head if all the TRV's are closed? Yep - so will a Delta-P pump. Don't forget those things are always running, and the impeller is always turning - even when there's no flow required.



What will the Delta-T function do if all the TRV's are closed? That's a pretty rare occurrence, because TRV's aren't made to be "off," they're made to modulate - so the likelihood may be completely off the radar, but just suppose...



Let's say it's mild out - and reset says the supply water temperature needs to be 120, and you've set the Delta-T to 30 degrees. That would mean the pump would run at a certain speed in order to bring the return water back at 90 degrees. As TRV's close - as in completely closed - the BumbleBee will slow down accordingly. When the last TRV's closes completely, the pump will be running very slowly. Now there's no flow. The pump will run at that same speed until the it senses that return water temperature dropping. It's not moving, after all, so it will probably drop some. The pump will try to get that return temp back up by ramping up -- slowly. Once one of the TRV's opens up again - the sensors will get a reading of return water temp changes, and the pump will do what it needs to do.



Again, this is one of those occurrences that is mostly theoretical - and if it does happen, the time period is wicked short.



How would a Delta-P circulator handle this same situation? Look at a Delta P pump curve chart. Delta-P pumps work on a fixed pump curve -- they're funny looking pump curves, but they're fixed. Even the Alpha - it has a range where it operates, and changes are based on pre-programmed algorithms - but it's still a pump curve. If all of the TRV's are closed, the Delta-P pump will work at the spot where the low end of the pump curve intersects the 0 GPM line - it's a fair bet the BumbleBee would be running at a lower speed because the pump adjusts to what the system needs, with a constantly-adjusting pump curve. Check the Delta-P curves to see where that intersection point is.



The other thing to remember, a Delta-P pump runs the same speed in January as it does in November.





There's lots to digest here, but at the very least you guys helped me write a new blog post! I don't check here very often, but if you have any questions, please email me directly at johbar@taco-hvac.com

CMadatMe

T Shirt

In November the heat loss is generally much less than in January when you are more inclined to get to design days or pretty close. I don't need the same flows and I don't have a need to pull out a tremendous amount of btu/hr so my emitters are not sucking them out at the same rate they would be in January. Pump head is the same to a point no matter the heat loss so the only change is btu/hr pulled out of the conveyor belt. It is also less likely that multiple zones are open simultaneously.

knotgrumpy

Thanks John!

I'm thinking when the ODR lowers the water temp in November it will become increasingly harder to have the radiation 'give up' the BTU's to hit the same delta T as at design temperature, right?  EDR charts show that.



At a certain point on a warmer day I could see the pump slowing to nothing to even hit the differential.

John Barba

hasn't happened yet...

Have had a BumbleBee of various generations in my house for 3 winters now - with outdoor reset, panel rads and TRV's -- hasn't happened yet. System doesn't deliver much heat in November because not very much is needed. Nor has it happened in any of the test sites used during the 3-year development of the Bumblebee, nor have we seen it happen in any of the installation since its release.

John Barba

We've got a winner!!!

I would have hoped you got that one right Chris!!!! E-mail me your size and address and we'll hook you up...

Gordan

With all due respect...

As the TRVs throttle down, the return temperature from their circuits would DROP, not rise. The whole reason why TRVs work is that they lower the AWT across the emitter. They do this by restricting flow - same as the Bumblebee. The only way the temperature should rise is if there's a bypass and enough flow goes through the bypass to make up for the cooler (but less) flow from the throttled emitters. Your one zone that's not on a TRV is probably acting as this bypass, so it's getting more than its share of flow until the thermostat is satisfied. I guess that works well enough if that one zone is relatively large compared to the rest of the system and the piping can accommodate the extra flow. It still seems plain that the Bumblebee could interpret a "call for no heat" (that is, the TRV acting to drop the AWT and, consequently, LWT) as a "call for heat", and that the two devices would apparently be at cross-purposes.



The usual design goal behind constant circulation is not electrical efficiency, as you seem to allude. With constant circulation, electrical efficiency becomes much more important, but fuel efficiency and comfort are typically touted as the prime motivators behind constant circulation, as those factors tend to outweigh electrical energy usage in a typical system. If they didn't, constant temperature operation at highest possible water temperature would still be the optimal choice, as that leads to shortest cycles and least electrical consumption.



EDIT: Actually, STEAM or gravity circulation would be the optimal choice - you can have that with no electrical usage at all!

SWEI

respect due

Ack - major typo there (and it was mid-day!)  Fixed it - thanks

Gordan

Still one issue left...

The return temperature dropping would widen the delta-T and actually cause a delta-T circ to speed up, rather than slow down.

pipeking

GORDAN

 that is if your supply temp stays the same. the pump is following the delta-t, if the return is droping the supply prolly is too. am i understanding u right?

CMadatMe

I Always Use

Radiator valves with built in bypass on them with panel rads. Page V-3, part numbers 1057 and 1058.



http://www.hydronicalternatives.com/getattachment/c87b4007-a1b7-4a31-ade3-85a9a240acf8/Overview-of-Accessories-and-Technical-Information.aspx



Your return water would be dropping much slower as the flow of the panel rad is much less then the by pass or could be the same. I don't see the pump having to ramp very much. Your still going to need to do the math and the by pass is still a TEE..



Flow is 2gpm to the by pass and the rad only wants 1gpm. Supply Water temp is 140. System designed for a 30 degree delta.



(140 x 1) + (110 x 1) = 2X



320/2 = X



X= 125



Still above my 30 delta....What has to change? Why does it change? Can't change supply water temp.

John Barba

Snapshots & movies

Hi Gordan –



Had to laugh at your opening line – can’t hear or read “with all due respect” without thinking of Ricky Bobby!



I think if you look at a snapshot in time, you’re very correct about how TRV’s affect AWT in a panel radiator. But think about what causes a single TRV to “throttle down?” The thermostatic head is getting closer to being satisfied – meaning it’s getting warmer in the room – meaning less output is needed from the radiator – meaning your reaching a point of relative equilibrium for the radiator - meaning fewer BTU’s are being taken out of the fluid – at that particular snapshot in time.



Give it a minute – and that equation will change.



Nothing remains static for very long in a panel radiator system. That TRV continues modulating in an effort to maintain a comfort level in the room. And what’s more – every other TRV in the house is doing the same darn thing, all based on their own settings, heat loads, radiators sizes, etc.



Cross purposes? If you look at one snapshot in time – yeah, you’re probably right. But heating systems are more like a full-length feature film than a snapshot. Look at what happens over the course of 10 minutes, 30 minutes or two hours, and consider what’s going on with all the radiators in the entire system (their TRV’s are probably doing different things at different times), and remember that a Delta-T circulator reads the difference between the water temperature going out to the system and the water temperature coming back from the system – key word being “system” – and will vary its speed to maintain that overall system Delta-T. If one radiator out of a dozen has a closing TRV, I can’t see that it would “confuse” a Delta-T circulator.



One more note – the circulator doesn’t care what the supply water temperature is – just the difference. Set that Delta-T for what makes the most sense for the application.



And just for giggles – since it’s about 4 degrees here in Minnesota today – I shut off the open radiator in the family room, turned my BumbleBee on continuous and let it operate on just the TRV’s on the rest of the radiators. Nothing blew up, the circulator hummed along between 9 and 14 watts (that I recorded – may have gone lower or higher while I was in the can), the boiler was happy and the downstairs (except the family room) was fine. Left it like this for about 4 or 5 hours. I don’t know what that means, but it means something…



At the end of the day, it’s kinda hard to get around the simple fact that GPM = BTUH ÷ (DT × 500).



BTW – I think you misread something. I said that people tend to like continuous circulation systems for electrical “simplicity,” not electrical efficiency. No relays, no thermostats, etc makes for a pretty simple system. That said – my own personal opinion – and it’s just that, my opinion – is that I’d just as soon turn a pump off when I don’t need any heat. With properly sized heat emitters and outdoor reset – you’re gonna have fairly long on times, anyway…

knotgrumpy

John Barba, are you on here?

Well, I guess you are.



I'm glad my question resulted in a good dialog. 



I think a really killer product would be micro Delta T circulators, John.



Imagine a fractional to maybe 2 gpm pump that could be attached to current distribution manifolds.  Each set to the particulars of the various emitters in the system.  About the price of a Sentry zone valve.



Be pretty cool.

John Barba

Mini-pump

Just need to get that machine from "Honey, I Shrunk The Kids"....

knotgrumpy

Big dreams, little pumps...

It will happen.  Look at what happened to boilers...



BTW - Gennady is showing off a BumbleBee in this most unlikely post:



http://www.heatinghelp.com/forum-thread/145170/garden-hose-for-gas-line



He has a link to a nice gallery of his work and features a Bee in one of them.

SWEI

ECM micro circs

are probably in our future.  There are a couple in the EU market but they aren't sold here.



I'm envisioning something that runs on 24 VAC and tops out around 2-3 Watts.  Small enough to mount on a manifold, internal ∆T control plus some kind of network connection.  Full BACnet or LON would be overkill here, but an inexpensive field bus akin to 1-wire that would allow control and monitoring would be great.  Imagine a smart micro LoadMatch...

Paul48

Technology

I think companies like Taco are only handicapped by the market itself. They have the resources to produce very advanced technologies. Getting them accepted in the market is another story. People don't change what they've done for years.

Solid_Fuel_Man

Very well

said!

knotgrumpy

That sounds right

I stopped in a supply house yesterday to order a few things. Mostly a forced air supplier with a little area for hydronics. They had a stack of 3 speed pumps.



I asked if he knew anything about the BumbleBee or the Alpha. Never heard of either. Said guys here just "keep it simple" and that I probably knew more than most techs in area.



Scary, since I sell carabiners and ropes.



Change is hard.

CMadatMe

Think About This

There are roughly 12-16 guys that regularly post, help and share. I'll call them the ground troops in the trenches. You get a few manufactures that respond when necessary such as in this thread and a rep hear or there. Other then Bob Eck I don't think I've seen or heard from anyone that works in a local supply house.



This site is such a valuable tool for someone that WANTS to learn and it's a shame it isn't taken advantage of more. I work for a rather large HVAC distributor. I could walk into any one of our branches and not one counter guy or salesman ever visit this site nor others like it such as the Taco Neighborhood nor would they even know about them. Why? It's not because they haven't been provided the information, it's because supply houses are changing into McDonalds and Burger King. Point to the number you want I'll give it to you? What's the guy down the streets price? I'll match or beat it. Why is this?



In my opinion it's because the majority of contractors do the same thing. What do you want Mrs Smith, I'll get it..What was the other guys price? I'll match it. In this sea of hydronics there are few fish that get it compared to the thousands of minnows.



Until the minnows evolve and there are enough fish, technology in our industry will continue to evolve as fast as a snail moves.The biggest reason for the explosion of high efficiency and few evolved changes we have seen has been 90% driven by the consumer forcing the change in contractors not contractors working as an industry to make the change.



Thank you to the internet and sites such as this because without them we would still be stuck in 50's technology.



I'm done ranting now..

Gordan

Not sure why that would be

The supply is maintained relatively constant by the boiler.

Gordan

Not sure

Chris, not sure whether this is a reply to my post above. If so, I did specifically point out that the system return temperature could be expected to potentially go up with a bypass. This would then be interpreted by the controller as a signal to slow down, if there's enough flow bypassed by the TRVs. Eventually, the circ and the TRVs would settle into an equilibrium. This seems like a good combination and, in fact, might be a better match for systems that are pumping through the boiler, depending on just how quickly the circulator responds (you wouldn't want to underpump through the heat exchanger - it would be nice to have some sort of "min GPM" and "max GPM" parameter that you could set to your heat exchanger's requirements.)

CMadatMe

Your Right

The circ would slow down and your delta across the rad would increase since it has the larger pressure drop. So what happens when the delta gets larger? Your panel rad thermostatic is going to open wider and allow for more flow which means flow through the bypass decreases.



This demonstrates why the Bumble Bee in my opinion is a better pump then an Alpha. It reacts to changes in the system btu/hr delivery, the Alpha does not.

hot_rod

There is a point

when the heat output from a convector drops quickly at low supply temperatures. Output charts show this clearly. So if you are running a system on ODR and forcing a ∆T would you not at some point run out of "driving temperature" and virtually no heat output from the panel rads or fin tube? Even in mild weather conditions buildings need some heat.



It seems manufacturers like Jaga and Runtal have keyed into that concept and now offer small fan assist rads to be used at low supply temperatures, realizing that forcing the convection (air movement across the fins) can compensate for the low temperature operation.



Solar, GEO, and mod cons are ideal low supply temperature generators. To leverage their efficiency low temperature operation is the key.



For years, decades really, companies like tekmar have schooled us on the near ideal reset curve, which almost perfectly matches the heat loss of a building. If you plot that "ideal curve" along side a curve with a forced ∆T the heat output "droop" is considerable at low supply temperatures. Pay attention to that with convection type emitters. Not so much on a radiant slab where the heat transfer is mainly radiant and MRT works to your favor.



I agree there is a time and place for ∆T function pumps, I can't see them being the ideal logic for all circ applications. It behoves the installer and designer to run some numbers, use the data and experience from the pioneers and base the component selection on the combination of the above.



To say one technology is better, or smarter than the other seems to limit your ability to offer both

Gordan

I get that the room will be getting warmer

And therefore, for any given AWT, the emitter will have a harder time rejecting heat into the room. But I'd expect those differences in room temperature to be small relative to the differences between the "ideal" supply water temperature (leading to no throttling) and the system supply water temperature - after all, the TRV is keeping the temperature in the room relatively constant. It's a dynamic system, to be sure, and it's not easy to mentally keep track of all the variables. But there are two kinds of dynamic systems: balanced ones, which typically have some sort of negative feedback that pushes everything to equilibrium, and unbalanced ones that feature positive feedback of some sort that leads to extremes. Imagine the case where the outdoor reset curve is set too high for the conditions so all the TRVs are throttling. In that case, the system delta T has to widen in order for the rooms to not overheat. Flow would drop rapidly with a dP circulator but wouldn't a dT circulator actually up its duty point in order to bring the delta T back to the configured value? This is what I mean by stating that the two devices would appear to work at cross purposes, whereas a dP circulator naturally and directly responds to the TRV's action in the correct way. Admittedly, I'm using a contrived example to illustrate my point, but I believe that it is a valid point: left to their own devices, the TRV (without bypass) and the dT circulator are not a natural pairing.



In a more realistic setting, we would be dealing with an optimized reset curve, so those TRVs should remain relatively open. But, by maintaining a fixed dT, doesn't a dT circulator force the system supply to be bumped higher to achieve the same AWT and, therefore, equivalent heat output from the emitters? And it does this by restricting flow, which is equivalent to what would happen if you bumped up the reset curve on a dP-circulated system and let the TRVs do the flow restriction. In both those cases, the circs will operate at a lower duty point than if the reset curve were optimal with a floating dT. And in both those cases, the same system dT will result. And in both those cases, it would seem that the benefit of a reduced system return water temperature would not extend to the boiler return water temperature because, having constant flow through the boiler loop, constant heat output, and a constant supply setpoint that it's maintaining, it follows that the boiler return (mixed boiler supply and system return) would be the same in both cases. The same is true with pumping through the boiler. So then, the question is: does it make sense to bump up the supply temperature in order to obtain a lower return temperature, but at a lower flow? And to what extent, if any, does it make sense to do that?

Gordan

Apples and oranges

I'm not sure how you can claim that a dP circ combined with two-way TRVs does not react to changes in the system btu/hr delivery. The TRV does this locally for each zone or emitter, and the dP circ provides the aggregate flow required by all the zones. dP circulators and pressure bypasses are not a typical combination because those two devices work at cross-purposes... either the pressure bypass would never allow the circ to meet its pressure differential setpoint or the circ would never allow the pressure bypass to get activated.



Now, a pressure bypass (or three-way TRVs) with a dT circ is a different story. Particularly the latter. But that's not the application we were discussing; in addition, see my response to John for more questions about the benefits of maintaining a constant dT in the system. (Maybe dT-v would be a good addition to the algorithm, allowing the target dT to increase with the circulator's duty point?)

John Barba

You're right...

...it is a contrived example, and contrived examples ALWAYS make one's point ;-)



I mean, who would contrive an example that disproves his point?



Delta-P was designed for this type exact application - and it works very well, there' sn o dispute there.. But Delta-T works well in this situation, too. As I said, I lived with the exact example you mention yesterday, just to see what would happen - a fairly realistic setting. Was it an entire heating season? No, just one freakin' cold day in Minnesota. The circulator seemed to be just fine with it all...because one TRV does not a system make. There are a bunch of them...and water temperature doesn't make them throttle, room temperature in the vicinity of the operating head does. And I'll bet dollars to donuts my ODR curve is probably a bit high for the real heat loss of the house. I calculated it to 44,000, but I'll bet it's lower than that in reality...



And again, the Delta-T pump does not set or control the supply water temperature - that may be the element you're missing - it doesn't even care what the supply water temperature is. It only cares about the Supply-Return difference. BTU requirements change - if the Delta-T is maintained, the system flow rate will change accordingly.



I think, Gordan, we may be at the point where we shake hands and agree to disagree...

John Barba

HI HR

I've seen those plots as well...and one thing I can't quite figure out, and what no one has been able to adequately explain to me...is that heat output droops with a fixed speed circulator as well and, according to the charts, would't "work."



Not sure what that means, but it must mean something...

hot_rod

true

but the fixed speed circs are not being marketed as as "smart" circ. If the installer is to pay 2 or 3 times as much for an intelligent circ, then it needs to be able to perform magic. Well near magic.



If in the case of ODR controlled systems the smarter circ cannot offer better low temperature performance, under some conditions, then leave that out and focus on what it can add or offer, and the Bee seems to have plenty to offer.



So "Bee" it.

John Barba

Tru dat!

Just because the pump is "smart," it doesn't mean the person using it doesn't need to be!

CMadatMe

This is Great

There are many lessons to be learned on this thread. One of the them is how two different mfg's can come together and provide excellent thoughts and insight in a great debate. Thanks HR and JB. Can't wait to get to ISH and see everything we want here and can't have.

hot_rod

Speaking words of wisdom

let it bee, let it bee