Delta T pump, circ controlled zones, with a Prestige Excellence

flitchplate

I called Triangle Tube but they were vague about it. Is a Delta T self adjusting variable speed circ compatible -- i.e. does not conflict and /or able to add value to performance of my TT Prestige Excellence?

Paul48

I just bought a new bumper for my truck....will it fit? Can you see why TT was vague? They would have to spend hours on the phone with you. Assuming you would be able to recognize and answer their questions regarding the setup of the system. There are plenty of folks that will help you here. We need a lot more information about the system though.Is the boiler piped P/S? What is the heat loss of the structure? Which size boiler? What is(are) the head for the zone(s) in question? What is the heat loss, by zone, for the zones in question?

SWEI

Maybe. I find them to be unpredictable and generally end up switching back to a fixed speed.

The challenge comes from the fact that both the boiler and the pump use a control loop to modulate, and the PID coefficients are not the same.

Paul48

SWEI........When will the boiler manufacturers and pump manufacturers get together? My response to the pump manufacturers is......."Still wrong, but much better" is not the goal here. That is not a direct quote, but it does answer a statement from a trainer in a video. It may involve cooperation between manufacturers, but the industry is ready, and the consumers deserve it.

hot_rod

it seem all that is missing is a 0-10V enabled circs, maybe a few smaller "residential" sizes. Those do exist currently, generally in a 230V 50hz Euro version

I'm pretty sure you will see some small 0-10 circs around the first of the year.

I have tried the 50 Hz Euro versions here in my shop, connected to 240V and one of the solar control outputs, works fine. I think many of the boilers have the signal output, not sure about settings.

Here is some good info on motors and switching from 50- 60 Hz.

hot_rod

hot rod said:

it seem all that is missing is a 0-10V enabled circs, maybe a few smaller "residential" sizes. Those do exist currently, generally in a 230V 50hz Euro version

I'm pretty sure you will see some small 0-10 circs around the first of the year.

I have tried the 50 Hz Euro versions here in my shop, connected to 240V and one of the solar control outputs, works fine. I think many of the boilers have the signal output, not sure about settings.

Here is some good info on motors and switching from 50- 60 Hz.

Credit attachments above to Pedrollo Pump

Steve Thompson (Taco)

We've been making circs who's speed is controller by a 0-10Vdc external signal for over a decade - just not ECM types. 00 Series VDT information attached.

Haven't been hugely popular as very few boiler manufactures could provide the control signal. That is until now...

Paul48

Steve......Gravity conversion, 7.5 gpm, 2 ft/hd(at design). You want to pump through a mod/con. Can any of the 00- VV's go low enough with that low a head?

flitchplate

"The challenge comes from the fact that both the boiler and the pump use a control loop to modulate, and the PID coefficients are not the same."

This is the same issue suggested to me by another wet head: that the PID control of the Prestige Excellence will be messed with by the Delta T pump.

Right now I have 3 speed Grundfos pumps but have been recommended by a contractor to consider the VT2218; Tacos new low flow pump replacing the Bumble Bee.

http://flopro.taco-hvac.com/media/Viridian_VT2218_100-114.pdf

I wanted to upgrade from the split capacitor constant rate pump to an ECM and variable speed model. But not a Delta P for the same reason: possible conflict with my boiler's controls.

Taco tech folks assure me the Viridian will work on the secondary (circ controlled zones) and not interfere with the boiler working on the primary circuit.

The only reason to go to the Viridain (as opposed to the B&G Vario, for example) is the Delta T feature.

BigRob

My experience with delta T control is from last winter heating a 12 unit condo building with 12 zone valves and radiant flooring. The existing pump was undersized and so I picked up the 0011. The distribution piping is unknown- most likely a series loop. There were times the pump would modulate down to about 50% and stay there when set to DT=10 to 15+deg. At least one condo would start getting cold. My thinking is the flow took the path of least resistance and my condo wasn't on that path. It seemed like the low speed was just not high enough to draw water from all open zones. As a result, some loops would stall and cool. I was fine with turning the DT set point to 5 to 7deg and having essentially a fixed speed pump. That warmed things up. I'll give things another try this winter and see what happens. If you can set a minimum speed in the vt2218 that would be neat. The head is still too low for my needs. A settable low speed would be great for boiler pumps, too. Not sure why there is no 0-10V input. It's so easy and useful.

I think you have to make sure your loops are really well balanced with delta T. We don't have balance valves. I'm thinking proportional pressure might be the right way to go in my situation. The only option I know of that would work for me is the 6.5" magna3 120v. Anyway, I like delta T. It makes a lot of sense if the tool has more settings.

Steve Thompson (Taco)

Big Rob - similar to the Bee the VT 2218 does have a manual speed setting (lower min speeds than the Bee), along with reverse temperature reactance and temp setpoint features.

Regarding your situation, I think the only solution for, shall we say "lack of properly designed/installed" systems is to overpump to overcome the worst case situations. In short, if the system sucks, neither properly dialed in Delta P or Delta T will solve the issue. Need a BAP (big a*& pump). Challenge is drawing the line between a "really well balanced" vs crappy system.

I can't expose much about the 0-10Vdc on The Wall, other than we are aware of this need...

Merry Christmas to all - drop by our booth at AHR/ASHRAE in January (and of course the famous Taco event!).

SWEI

Paul48 said:

Steve......Gravity conversion, 7.5 gpm, 2 ft/hd(at design). You want to pump through a mod/con. Can any of the 00- VV's go low enough with that low a head?

The VV controls combine traditional TRIAC phase cut dimming with a PWM type pulse mode to give large turndowns. It's still important to size properly, given the physics of centrifugal pumps.

BigRob

Steve Thompson (Taco) said:

Big Rob - similar to the Bee the VT 2218 does have a manual speed setting (lower min speeds than the Bee), along with reverse temperature reactance and temp setpoint features.

Regarding your situation, I think the only solution for, shall we say "lack of properly designed/installed" systems is to overpump to overcome the worst case situations. In short, if the system sucks, neither properly dialed in Delta P or Delta T will solve the issue. Need a BAP (big a*& pump). Challenge is drawing the line between a "really well balanced" vs crappy system.

I can't expose much about the 0-10Vdc on The Wall, other than we are aware of this need...

Merry Christmas to all - drop by our booth at AHR/ASHRAE in January (and of course the famous Taco event!).

I hear you

Steve Thompson (Taco) said:

Big Rob - similar to the Bee the VT 2218 does have a manual speed setting (lower min speeds than the Bee), along with reverse temperature reactance and temp setpoint features.

Regarding your situation, I think the only solution for, shall we say "lack of properly designed/installed" systems is to overpump to overcome the worst case situations. In short, if the system sucks, neither properly dialed in Delta P or Delta T will solve the issue. Need a BAP (big a*& pump). Challenge is drawing the line between a "really well balanced" vs crappy system.

I can't expose much about the 0-10Vdc on The Wall, other than we are aware of this need...

Merry Christmas to all - drop by our booth at AHR/ASHRAE in January (and of course the famous Taco event!).

To be clear, can you adjust the minimum speed while in delta T mode beyond the current high and low setting? I would think a range of 4 to 5 speeds would be great. Actually, a low medium and high would probably be fine. It's such any easy thing to do. I don't see why you wouldn't. It would be helpful in my case and a lot of other retrofits to have more flexibility. Plus, it allows the installer to feel like things are final and there is a solution. Just my 2 cents here. Go USA.

Paul48

SWEI........Was that a "no" ?

SWEI

That was a YES.

Steve Thompson (Taco)

Sorry for the confusion...

When in delta T mode the circ picks it's own speed. I might be missing something but don't understand the reason to limit speed and use the delta T feature are the same time. I was referring to manual speed selection, disabling the temp control function.

Even on a very small system, if the temp differential is satisfied, as long as where the system curve meets the circ pre-programmed minimum speed all is good.

The VT 2218 will go as low as 1.5' @ 4 GPM or 3' @ 2 GPM as an example so small zones aren't an issue.

Please help me better understand your need - we might be able to add something to the software code (no promises though).

SWEI

From what I have seen, the problem appears to be low loop gain (rate of increase) on the upside. Slowing down slowly is OK, but when the boiler ramps up its firing rate, the circ takes too long to react, causing the boiler to hit its limit(s) and short cycle, sometimes even shutting down entirely.

Harvey Ramer

SWEI said:

From what I have seen, the problem appears to be low loop gain (rate of increase) on the upside. Slowing down slowly is OK, but when the boiler ramps up its firing rate, the circ takes too long to react, causing the boiler to hit its limit(s) and short cycle, sometimes even shutting down entirely.

I have come to the very same conclusion Swei.

I would also like to expand on it a little bit. Different boilers have differing durations of startup sequences. Most, if not all will go to high fire till the flame is stabilized and then modulate down to target setpoint. That would make it necessary for the Bee to perhaps run at full speed for a bit longer upon initialization.
Also as Swei mentioned, it becomes necessary for the Bee to react very quickly to a widing delta-t. I think the first step in this would be to put the sensors in immersion wells instead of strap on. A strap on sensor on a 1" copper pipe can take up to 8sec to respond to a fluid temperature change. It only gets worse as the pipe gets bigger or the material of the pipe changes. Such as steel piping, It takes a lot longer to respond then copper.

SWEI

If we had any means of dictating speed to the smaller ECM circs, we could use them in a lot more applications. While very few boilers currently have 0-10V outputs for the purpose, most of them do have ModBus capabilities. Using that, we can read the internal sensors in real time and craft a strategy to control the circ in a manner which does not conflict with the internal controls.

NYplumber

I have a Bee on a WHN boiler as the boiler circulator. Works just fine, with one limitation. Where the hx head loss intersects the Bees curve, even at the slowest speed the DT is too small and flow is too high.

A dt pump is just fine, just be sure the lowest flow rate is above what the hx requires or you will end up with a leaking boiler.

Lastly, before going to the extreme of a dt boiler circ, get the rest of the system tweaked to perfection. This process will take months of fiddling.

hot_rod

could it be in the lag time of the sensors? Different types of thermistors respond differently. Here is a link to some good reading on sensor types and how they can be engineered to specific needs.
http://www.ussensor.com/technical-info/thermistor-terminology

BigRob

Steve Thompson (Taco) said:

Sorry for the confusion...

When in delta T mode the circ picks it's own speed. I might be missing something but don't understand the reason to limit speed and use the delta T feature are the same time. I was referring to manual speed selection, disabling the temp control function.

Even on a very small system, if the temp differential is satisfied, as long as where the system curve meets the circ pre-programmed minimum speed all is good.

The VT 2218 will go as low as 1.5' @ 4 GPM or 3' @ 2 GPM as an example so small zones aren't an issue.

Please help me better understand your need - we might be able to add something to the software code (no promises though).

No worries. This is a good discussion - want to make sure I'm using the product correctly. I believe I have a situation where one zone has high flow resistance and one zone has low flow resistance. When each zone is on separately, no problem. When they are both on, the low flow resistance circuit dominates. Because of this, I believe, as the pump throttles down the low resistance zone is around its target DT, but the high resistance zone is getting only a little flow. So, now the pump is at minimum speed and one zone is fine, but there is no head to get the column of water flowing in the high resistance zone, and as a result it's out of control. If the pump is stimulated (ooh la la) and the speed increased to full, then the fiction in the low resistance circuit increases and a greater proportion of water flows to the high resistance circuit. I understand distribution piping diameters could have been selected to load match each zone, although it's unclear if that is the case here. Probably not. Maybe you guys could come up with a "retrofit" mode that has a more aggressive control strategy. My comment earlier about a programmable low speed value, I believe, would help the above situation. You might sacrifice some DT range, but maybe that is better than the perception things aren't working. You have three buttons on the new pump. I would have max/min speed in %, a % setting for control aggressiveness, and a random stimulate function (yeah baby) that pulses to a user settable speed % so to pump doesn't get stuck on low speed and take forever to climb out. This is my current thinking. I reserve the right to be proven wrong.

BigRob

SWEI said:

From what I have seen, the problem appears to be low loop gain (rate of increase) on the upside. Slowing down slowly is OK, but when the boiler ramps up its firing rate, the circ takes too long to react, causing the boiler to hit its limit(s) and short cycle, sometimes even shutting down entirely.

I have seen this, too. I agree. The 0011VDT I have could react faster, even when set to the most sensitive setting.

BigRob

Harvey Ramer said:

SWEI said:

From what I have seen, the problem appears to be low loop gain (rate of increase) on the upside. Slowing down slowly is OK, but when the boiler ramps up its firing rate, the circ takes too long to react, causing the boiler to hit its limit(s) and short cycle, sometimes even shutting down entirely.

I have come to the very same conclusion Swei.

I would also like to expand on it a little bit. Different boilers have differing durations of startup sequences. Most, if not all will go to high fire till the flame is stabilized and then modulate down to target setpoint. That would make it necessary for the Bee to perhaps run at full speed for a bit longer upon initialization.
Also as Swei mentioned, it becomes necessary for the Bee to react very quickly to a widing delta-t. I think the first step in this would be to put the sensors in immersion wells instead of strap on. A strap on sensor on a 1" copper pipe can take up to 8sec to respond to a fluid temperature change. It only gets worse as the pipe gets bigger or the material of the pipe changes. Such as steel piping, It takes a lot longer to respond then copper.

This would be another great user configurable setting.

Paul48

I've spoken against zoning mod/cons, but how about an interface that would accept 6 t-stat inputs, and you tell the mod/con what each is. Still not a big fan of zoning, but if it is going to be, why not. I know that some of the mod/cons have a couple curves, why not more, as an option?

SWEI

hot rod said:

could it be in the lag time of the sensors? Different types of thermistors respond differently. Here is a link to some good reading on sensor types and how they can be engineered to specific needs.
http://www.ussensor.com/technical-info/thermistor-terminology

It looks to me like those time constants are mostly about internal heating from current flow.

How the thermistor is potted will make a difference, but I'm pretty sure a thermowell would solve most of the issues with inputs for the pump controller. Unfortunately, adding thermowells to an already crowded near-boiler pipe job is not always practical.

I'd still like some kind of configurable loop gain (even a low-medium-high would probably do it) to accommodate different system characteristics. An external speed input would still be best.

Steve Thompson (Taco)

I read min speed is still too much on the Bee. Have you tried to introduce artificial frictions loss by slightly throttling down the valve. Circs are dumb - they don't know if they are pumping against friction loss or a partially closed valve.

Reality - I've seen this before with the Bee on setpoint moving fluid through a cooling coil - just wouldn't go slow enough. That's why the min speed on the new VT 2218 is lower than the Bee.

NYplumber

null
Steve, I had that idea only after it was installed. I have other boilers with that artificial friction employed with standard pumps. Works quiet well.

SWEI

The issues I am seeing are from too low a flow and/or too slow an increase off that low on boiler circs.

NYplumber

Swei, the Bee starts off in high speed for three minutes if I recall. Once the bee finds its dt target, the boiler is forced to follow suit based off inlet and outlet temps.

On smaller residential systems, variable speed boiler pumps arent cost effective. Better to size the fixed speed pump to the correct dt for the coldest day of the year.

Note: my findings were on fire tube low head loss designed boilers.

SWEI

NYplumber said:

the Bee starts off in high speed for three minutes if I recall. Once the bee finds its dt target, the boiler is forced to follow suit based off inlet and outlet temps.

In theory, yes, but when the minimum speed of the Bee is too low for the boiler, the boiler cycles on and off its high limit. When that happens, the Bee often seems to get "stuck" in its lowest speed slot and doesn't modulate back up.

On smaller residential systems, variable speed boiler pumps arent cost effective. Better to size the fixed speed pump to the correct dt for the coldest day of the year.

Agreed.

my findings were on fire tube low head loss designed boilers.

As are mine. I'm pretty sure this would be solvable if we had the ability to adjust the loop gain on the circ.

NYplumber

Swei & others, do psc circs with dt function (ie, 007dt) fall into this trap?

SWEI

Haven't tried those, given that the VVF models are about the same price.

The internally -controlled Taco models have a near-zero minimum flow, so I'm not sure how they would work out in ΔT mode.

We don't anywhere near that kind of turndown for gas-fired boilers, so we just use an external TRIAC module and set the DIP switches to make the minimum and maximum match the needs of the boiler.