Indoor Reset

Paul Rohrs

I was able to attend the RPA-ISH Conference in Boston a few weeks back and had the opportunity to sign up and attend John Siegenthalers RPA course, "Radiant Precision". We spent several hours talking of radiant piping methods and all things hydronic.

It was interesting as we delved into "Control Strategies" of radiant systems. PID logic was talked about and Siggy said,(I'll paraphrase) that really the D (Derivative) in the PID controls are important, but are not really utilized in our radiant systems. (Not sure if I should of had that in quotations or not). He went on to say, Derivative function would be more significant in a rocket engine for course corrections or systems that need immediate response. Our typical hydronic systems are slow responding even with low-mass, fin-tube systems.

Weeks later, I am listening to Swampeast,MO - Mike T and his fascinating Viessman system with TRVs and I am thinking that outdoor temp is relevent for control stategies only for the WWSD feature. The heat loss of the structure (regardless of outdoor temp)is the only relevent factor and should be what we desire most for our heating systems. Obviously, we can send a reset temp through our system that would require PI(D) logic, but this (theoretical?) control should only focus on, and optimize the rate of heat loss in our particular structure.

Are there any manufactures with these controls out? Tekmar has outdoor reset features that can include an indoor sensor, but that is to "fine-tune" the water temp.


It seems that this control should have very fine parameters for input to let the control have a preliminary understanding of the structures needs. Max Btu's at design conditions, method of heat transfer ie, Slab-on-grade, baseboard, CI Rads, ets. (This should be based on constant flow as well)

I have been designing around injection mixing with outdoor reset for years and have had no systems with any problems to speak of. Paul Pollets RPA seminar about "European Hydronic systems" was enlightening in this regard also in that he is using the Indoor reset systems currently with his Viessman systems. Is Viessman equipment stating this is an Indoor-Reset technology, or is his system just reflecting this indoor methodology?

Viessman fascinates me and I have been talking with them to see if they are interested in distribution in my area.

I am "chomping at the bit" to stay ahead of the technology curve in hydronics, but feel I am missing something. I am looking forward to input, discussion, and how I can be designing better systems.

Regards,

PR

Mike T., Swampeast MO

You might be talking about the RS Remote Control for the Vitodens. It "measures the room temperature and initiates any necessary correction of the supply temperature".

Sure sounds like indoor feedback to me...

Something VERY important in the literature regarding that device, "Due to the 'inertia' of underfloor heating systems, the RS function must not be used for an underfloor heating circuit."

You said, "Our typical hydronic systems are slow responding even with low-mass, fin-tube systems."

When I put those two together, it seems to tell me that indoor air temperature has nothing to do with loss in the structure. It's the outside temperature that gives the boiler the ability to make a relative guess at the heat loss.

Gotta run...

J.C.A._3

Paul,

I attended a "Brookhaven Seminar" given by Alan Mercurio, of Oil Tech Talk fame in 2003. He had an attendee named John Cockerill(sp). He was showing us a control that he had conceived and produced using indoor reset as the drive.

I just found the sheet that was passed out in the class and found a web site,(www.exqheat.com). I don't know if it will help you in your thoughts but it came to mind while reading your post.

This has NOTHING to do with the RS system, but is a controller that uses the feedback from the thermostats to set the controller and cycle the boiler accordingly.

The website is still active...I checked. Give it a look and see if it is something that you believe would be helpful. Chris

larry

Certainly hydronic control is much more advanced than typical forced hot air furnace control by a long measure. But with that being said, the approach your discussing above sounds somewhat similar to what Rheem/Ruud is doing with their Mod 90 furnace. The furnace modulates from 40% to 100% of capacity and uses an ECM blower that keeps the temp rise across the HX the same regardless of modulation level. They have a proprietary thermostat that has an additional signal (V terminal) to the furnace control board that tells the furnace what modulation level is needed. I believe the thermostat is measuring the rate of change of air temperature plus cycle data, and trying to use this information to infer the needed heat production to match the loss. As the system runs and you accumulate data on multiple furnace cycles it is suppose to get smarter with regard to this calculation.

Bob Tonner

Indoor Reset ( I like the name)

Paul,

Our company has a patent pending on a technology that will not likely be out until some time in 2005 or 2006 called Dynamic Demand Profiling. Essentially, we have found a way of calculating the precise heat loss of a structure from information available in virtually every heating system, without a multitude of sensors (indoor or outdoor). Our reset formula is based upon an algorithm that calculates the "Thermal Resistance" of the structure and accurately resets the Mean Delivered Water Temperature based on the current rate of heat loss.

Modulating the water temperature in theory will increase occupant comfort, but we see no evidence that it will incease system efficiency. We have, however, found that moduating the Differential based on the system's capacity to recover (BTU input vs BTU loss) is an effective way of increasing efficency through fewer but longer burn cycles (not a new idea).

The problem is that wide swings in delivered water temperatures (Differential) works against the occupant comfort objective of reset, so a comprimise must be made.

I don't know if this is what you were thinking about, but it is something that we are working on.

Bob Tonner
President
InfinityLab Inc.

www.infinitylab.com

Greg Swob

Are we perhaps talking about the Beckett Heat Manager? It is not an outdoor controller, measures supply/return temps, then calculates a temp set point for the boiler to operate at. Greg

jerry scharf_2

some observations on the controls

This turned into a bit of a stream of conciousness post on the general nature of controls and those used for heating buildings.

Siggy is right, there's little value to the D component of a PID controller for a slow function like heating a room. Even if the heating system has zero mass, the room has lots. The proportional is the most common (i.e. outdoor reset,) and there are some interesting things that can be done with the integrating part (this looks at something analogous to the total heat added during a period.)

There some great work from 50 years ago called information theory. It looks at what is the maximum theoretical information given a set of inputs. It was the key to all of digital coding, and people may have heard of the Nyquist limit, which is one useful case of this work.

I bring this up because I believe there are a number of things that are talked about, but rarely are the inputs dicussed. The average boiler has a couple on-off switches as input, one the supply temeprature aquastat and the room thermostat. One can make certain extrapolations based on the time intervals seen on the two switches, but it's hard to be intelligent.

The next step getting smarter is the smart boiler. It makes no assumptions about the room heat control, but adds several sensors and controls inside the boiler. This allows the boiler to make better decisions about what it should do, but is still guessing about the nature of the demand. All the modulating boilers fit into this class, as well as others.

The next step is to add better sensors in the builing, and have that make better demand estimation. Tecmar and Caleffi are two systems that take somewhat different approaches to this. Unfortunately, there are very few cases where this intelligence integrates smoothly with the boiler intelligence mentioned above.

The next level is to collect information elsewhere and make other subtle changes to the indoor climate. Should you make the rooms a degree warmer if it's 35 and raining outside vs. 35 and sunny? How much passive solar input will come into a given building area and when so we can turn down the high mass emitters in advance of this change?

Once you've got the information, then you need to look at the places and types of controls (control points.) If you have proportional points vs on-off, things change. If you have variable pressure or constant will impact the changes to the system as demands shift. Do you have control of the supply water temperature or is that done beyond your control?

For me, all this comes before the control strategy. If you're a boiler manufacturer, you can't assume too much information back from the building or understanding of the piping/builing controls. This tends to make the engineers work really hard trying to guess at things and run the boiler accordingly. There are aome that do an excellent job of this, but it's always harder and less effective than having the information to make the right decision.

Looking at the interaction between outdoor reset and setback recovery is one classic case. If the same controller is handling both the room temps and boiler temp, this can be easily handled.

jerry

Mike T., Swampeast MO

Bob,

The Vitodens appears capable of what you say, yet it has no indoor temperature sensor. Only the outdoor sensor and whatever sensors are built in to the boiler itself. (Still not sure if it senses return temp or not by the way.)

Cloudy/rainy weather finally went away and days are significantly warmer with nights significantly cooler. Still not cold and the boiler is having to learn how to most effectively run in digital operation--remember it can't modulate down to nothing. Low end is about 22 mbh.

This is an old gravity system so there's LOTS of water in the mains. Also, the radiant floors are just connected to the mains with no form of control. Even with all of the TRVs satisfied, there is still a small amount of flow and small heating load from the radiant areas. Therefore there is always some circulation in both main pairs.

The old boiler (a simple standing pilot cast iron cube) was connected to mechanical outdoor reset and WWSD. Circulator constantly running except when stopped by WWSD. Once the house was maintaining heat, the boiler firing time became remarkably consistent--about 2 minutes, 20 seconds. Only the interval between firings would change with outside temp.

There is no low-loss header for the Vitodens and I speculated that it would turn the mains into a buffer. Guess what? It is.

While the Vitodens is still learning, it's a precocious kid that learns fast.

When outside conditions are such that lowest modulation is still too much heat, operation is really interesting:

After not firing for quite a period, the burner will come on at full fire. You can watch the actual supply temp rise rapidly. Within a few seconds, the burner shuts down. This repeats every few minutes for a few times. Then, the burner will come on and ramp down immediately. Actual supply temp is at the reset setpoint and it STAYS there for a very long time--easily 20 minutes! At the end of the cycle, the supply temp starts rising quite rapidly (after having stayed steady) and the burner shuts down.

Then the burner stays off for a long period and supply temp begins a very gradual decline.

My only explanation is that the boiler is learning to charge the buffer (e.g. all of that water in the mains) in a manner to allow the longest possible burn time with minimal target supply temperature overshoot or undershoot.

Mike T., Swampeast MO

Another Level

Between the "smart" boiler and "more and better sensors".

It's the boiler that learns how to adapt to the system to which it is connected. Ideally it learns how to keep everything (boiler and system) operating at peak efficiency. Such is relatively simple if the system has constant and proportional flow. It gets much more difficult with a barrage of digital inputs.

In that regard, modulating boilers are different. I believe that the Vitodens is the only one capable of learning in the way.

Fortunately it does NOT lock you out from being able to control the supply temperature--it learns to adapt to the temperature curve you set.

As I mentioned elsewhere, my Vitodens is learning to use the volume of the system as its buffer when demand is lower than low-end modulation. I suspected this would happen--in fact I reasoned that it must happen were this boiler to be "perfect". So far, it does appear perfect.

Interesting that the control unit displays not only current outside temperature, but "damped" outside temperature. (The damped temp is blank on mine.) I suspect that in Germany MANY more options are available and that you can install a 2nd outdoor temp sensor on the "hot side" of the structure instead of just the "cold side". The unit will then use the difference between the two to establish a damping factor that makes a corresponding adjustment in the supply target.

Anyone know if such a sensor is available HERE and can be installed in a system of type 000:000???

This is going to sound like a bizarre idea, but here goes anyway:

Say you have a proportional system with very low heat loss--not much greater than the low-end modulation rate of the boiler.

One approach to this problem would be to drive the system through a buffer and allow a wide variance in the buffer tank temperature between "charges" with heat.

As an alternative, why not install a well-insulated hot water tank in series with the system load. The burner WILL find all of that extra water. Even with extremely low loads, you should still get nice, long firing cycles.

hr

Danfoss has

an indoor reset control. It was built to help overcome solar gain problems. I tried on on my office which has a long windowed south facing wall. Mine was tied to a 3 way Esbe valve with a motor actuator.

hot rod

To Learn More About This Professional, Click Here to Visit Their Ad in "Find A Professional"

jerry scharf_2

> Between the "smart" boiler and "more and better

> sensors".

>

> It's the boiler that _I_learns_/I_

> how to adapt to the system to which it is

> connected. Ideally it learns how to keep

> everything (boiler and system) operating at peak

> efficiency. Such is relatively simple if the

> system has constant and proportional flow. It

> gets much more difficult with a barrage of

> digital inputs.

>

> In that regard, modulating

> boilers are different. I believe that the

> Vitodens is the only one capable of learning in

> the way.

>

> Fortunately it does NOT lock you out

> from being able to control the supply

> temperature--it learns to adapt to the

> temperature curve you set.

>

> As I mentioned

> elsewhere, my Vitodens is learning to use the

> volume of the system as its buffer when demand is

> lower than low-end modulation. I suspected this

> would happen--in fact I reasoned that it

> _I_must_/I_ happen were this boiler to be

> "perfect". So far, it does appear

> perfect.

>

> Interesting that the control unit

> displays not only current outside temperature,

> but "damped" outside temperature. (The damped

> temp is blank on mine.) I suspect that in

> Germany MANY more options are available and that

> you can install a 2nd outdoor temp sensor on the

> "hot side" of the structure instead of just the

> "cold side". The unit will then use the

> difference between the two to establish a damping

> factor that makes a corresponding adjustment in

> the supply target.

>

> Anyone know if such a

> sensor is available HERE and can be installed in

> a system of type 000:000???


The vitodens is a top of the line in the smart boiler. All the things you say fall under the smart boiler in my way of breaking things up. Your sorting my do this differently.

I've not played with a Caleffi iradiant system, but from the sounds of it it has similar learning controls on the building side. If I wasn't doing the controls as a lab, it would be really fun to put the iradiant control in and try to make it dance to my tune.

>

> This is going to

> sound like a bizarre idea, but here goes

> anyway:

>

> Say you have a proportional system

> with very low heat loss--not much greater than

> the low-end modulation rate of the boiler.

>

> One

> approach to this problem would be to drive the

> system through a buffer and allow a wide variance

> in the buffer tank temperature between "charges"

> with heat.

>

> As an alternative, why not install

> a well-insulated hot water tank in series with

> the system load. The burner WILL find all of

> that extra water. Even with extremely low loads,

> you should still get nice, long firing cycles.




I have just the case you speak of. Much of the time my load will be at or below 15MBH, which is the bottom of just about any condensing/modulating boiler out there. I was planning to do plan B, but have decided to wait until I have a season under my belt and see how much the system needs it. I've been assured that 15 minute light to light on the burner is fine for my boiler (munchkin T50.) By the way, I went with the munchkin partly because it doesn't try to be as smart as the Vitodens. It's really hard to create controls when the boiler has a mind of it's own. I hate things that are smarter than I am and won't tell me.

My concern with your plan A is what happens to the return water temperature as the buffer gets close to fully heated. Don't want that to impact the boiler efficiency.

jerry

Mike T., Swampeast MO

Glad that tank in series idea wasn't so kooky after all One person asking me about FHVs has a system with only a 30 mbh design load and I've been watching and thinking...

Can't disagree that the Vitodens locks you into doing things its way. BIG reason that I chose it over others is that the way it works is exactly what I was looking for. The seemingly perfect ratings didn't hurt either.

How about driving two variable speed hydro-air handlers directly from the built-in circulator (provided their head loss isn't too great of course) and without the low-loss header? Then later adding FHVs for in-floor heat downstairs into a separate mixed loop with its own variable-speed circulator? Still no low-loss header because the controls he's using for the air handler can effectively lock them into reduced capacity when the floor is handling most of the load? Can also lock out the air handler based outside temp to ensure the floors carry all their possible load.

Constant, variable driving force for the HX in the air handler with variable speed in the delivery. Sounds sweet to me. How 'bout to you?

I don't much care for forced air, but that should be about the closest thing to a truly proportional air system around.

hr

Your plan B, Mike

is beginning to sound like a Voyager or Polarsis with a modulation burner. I'd like to get my hands on a Voyager with a Munchkin burner and brain. Then you have the 50 gallon buffer built into the heat generator! Of course th VS pump Viessman uses would also be a nice option. Heck i would even deal with the 230V issue that comes with that pump!

Until they modulate the condensors to near zero, they will always be challanged with micro loads. Short of buying a couple small output, and low modulating rate, boilers, the buffer tank still has a place in the systems we design and sell customers for ultimate comfort i.e. micro zoned, multi loaded radiant/ hybrids.

Sounds like the modulating TRV approch of yours,,and Viessmanns is real close but not 100% there.

If 100% is your "I won't rest till I get there" goal. thanks for the time and energy pursuing it!

hot rod

To Learn More About This Professional, Click Here to Visit Their Ad in "Find A Professional"

Mike T., Swampeast MO

Just Wish

They'd give some details on the actual operation of the brain--in particular how it deals with situations where it's forced to go into digital operation. Also, some details about how it handles WWSD would be very valuable.

It appears to allow a substantial drop in system temperature when it's near what I consider WWSD, but I have a sneaking suspicion that if it's say 20° outside and the load is still very low that it won't allow so much drop.

Would hate to make the tank too big and create much of a parasitic load in non-conditioned space, but at the same time want to ensure the boiler finds enough load to keep burn time reasonably long...

Let's see:

Say I want to reserve 11,000 BTUs with a 7° temperature variance. So, I can put 7 BTUs into each pound of water. 11,000 / 7 = 1571 pounds of water. 188 gallons--tank TOO big and/or temp variance too small.

If I work the other way:

Say a 60 gallon tank weighing 500 pounds. With a 10° differential I can reserve 5,000 BTUs.

If the load on the system is 10,000 BTU/hr and the burner can deliver a minimum of 22,000 BTUs in an hour and IF (the BIG if) the boiler allows a 10° variance at whatever outside temp this loss corresponds to, I think the burner would fire for about 40 minutes, then go off for about 30.

Sound logical? 10° supply temp variance over 70 minutes doesn't sound too bad...

Send me those losses and other data, Mike! Make sure you include city--I'll look up weather data.

Mike T., Swampeast MO

Voyager or Polaris with a Modulating Burner

I'm truly not joking when I say that's what this old gravity conversion system with Vitodens is looking like!

That's why I kept saying that it seemed UTTERLY stupid to put a buffer in a system with so much content in the mains.