owed Mike a response on emitter efficeincy

jerry scharf_2

Mike,

I lost the thread but I wanted to answer you.

You asked what makes the most efficient emitter, and said you thought wide delta T would produce that. It depends.:) In this case by what you're measuring as the efficiency.

As a component, you can think of your room emitter as a form of HX. For a given inlet temperature, the HX will be most efficient with the lowest delta T, as the mean temperature is higher. It gives you the most output for the given size of emitter.

As a system component, you can have two scenarios. One is that you are load limited and the other is that you are emitter limited. If you are in the emitter limited case, you need to drive for lower delta T to get the most output possible out of the system. If you are in the load limited case the emitter efficiency is a first order wash (you integrate over the temeprature/power function of the emitter to get the output, and you control the output to the load.)

The wider delta T has some system implications. For the condensing boiler, the return temp is the guage of boiler efficiency, and wider delta T means more effieincy. It can mean better circulator pump efficiency, but there few pumps available in the US that could exploit this. There can be significant comfort issues as well, with radiant floors being the clear case. A wide delta T could cause some areas to be uncomfortable to the feet, etc. Some people might consider lowering the hottest radiator temperature to be a safety issue, but that's stretching it IMO.

As for controls, I would expect a wider delta T to relax the boiler controls and stiffen the room controls. Overshoot in the room controls would be a more significant problem with a wider delta T, though this is secondary to thermal mass. Proportional controls don't prevent overshoot from external heat sources, and the stiffer the room control the more overshoot to expect. Yet another case where wide delta T and high mass radiant don't mix well. (I'm assuming you get what I mean by relaxed and stiff like a car suspension.)

In your case, condensing boiler, standing iron and TRVs, a wide spread should help the system. In radiant floors, the comfort is a problem and lower delta Ts are usually sought. Then again, because the return temperatures can be so low with a higher efficiency radiant system, spread is much less important. With noncondensing boilers, there are sufficient limits on the stack temperature that return water temperature matters much less.

does any of this make sense or is it a late night stream of confusion?

jerry

Mike T., Swampeast MO

Thanks Jerry

Makes good sense--mainly.

"For a given inlet temperature, the HX will be most efficient with the lowest delta T, as the mean temperature is higher. It gives you the most output for the given size of emitter."

Yes, except at the complete extreme where incoming and outgoing temperature are exactly equal. In that case your your input or output is zero (or at least exactly even)! I know that's impossible unless the temp on the other side of the HX is the same as what's inside, but it does show that you MUST have some delta-t to get heat transfer to begin with... Yet, technically, the device is capable of it's highest output one nth of a degree away from where output stops!

Here's another confusing situation:

Say you have a 300' loop of suspended PEX tube heating a floor. You've supplied it with 160° water. 150° water is returning. By your measure, this is an extremely efficient system with very high output. Uh-uh! Why? Because the heat can't make it out of the tube!

Now say that's a 300' loop of pex in Thermofin or some other heavy conduction material. Same spacing, same flow, same supply temperature. BUT the return water is now 130°. Average temp is significantly lower and delta-t is significantly higher, but that system gave off much more heat!

I know that's an apple-orange comparison, but the point is that low delta-t doesn't always equate with high output--even at high flow.

"I would expect a wider delta T to relax the boiler controls and stiffen the room controls. Overshoot in the room controls would be a more significant problem with a wider delta T, though this is secondary to thermal mass."

Agree 100%. That's why you wouldn't want to use regular TRVs for a floor. BUT, the FHVs give you another layer of control--intentional restriction to flow. Because you're talking about a continuously circulating system with a finite flow ability (even with a variable speed pump such is controlled by outside temp and it doesn't change unless outside temp changes) your ability to overshoot is greatly limited because there won't be much "extra" flow ability in the system.

If you don't have a proportional pump you can engineer a reset curve that will keep MAINTENANCE flow quite consistent regardless of the weather--problem though is that your delta-t will increase with warming outside weather.

Since you have constant circulation with FHVs and the valves are modulating flow to achieve the average floor temperature just adequate to meet heat loss, the temp of the floor is going to tend to average itself out over time. Compared to a digital system which must continually either raise the floor temp or allow it to fall, I most sincerely believe that you could have some surprisingly high delta-t (and resulting low flow) without objectionable temperature differences across the floor. Again, this could be DEAD wrong.

That table I posted with many zones with extreme delta-t may well be pushing things too far. There were at least two spaces that REALLY needed supplemental heat in cold weather. The odd thing however is that other spaces with whopping delta-t were designed for carpet and/or large, heavy area rugs--I believe that such would tend to lessen temp difference across the floor.

don_52

a related... sorta

Gents,

some number of months ago someone wrote about
oversizing their panel rads, the result being
to stay in the "condensing range" longer.

i.e. size for 140 @ design instead of 180.
i know this is different than were Mike is
going, but somehow related.

i found a source that carries panel rads in a
wider selection of sizes ( i can e you the docs )

i think that ( space permitting ) that you could
monkey around with the size of the rad to get the
desired outlet temp.

i discussed this will a "Wally" offline, oversizing
rads and using the pump to reach a "similar" effect,
the numbers seem to support this. ( been a while )

perhaps a combination of the two or three ( rads, pump
and trv/fhv ) items may work.

just a thought...?

regards, don

don_52

a related... sorta

oops, messed up

don_52

a related... sorta

Gents,

some number of months ago someone wrote about
oversizing their panel rads, the result being
to stay in the "condensing range" longer.

i.e. size for 120/130 @ design instead of 180.

i know this is different than where Mike is
going, but somehow related.

i found a source that carries panel rads in a
wider selection of sizes ( i can e you the docs )
can you imagine an 8"h x 120" long rad?

i think that ( space permitting ) that you could
monkey around with the size of the rad to get the
desired outlet temp.

i discussed this will a "Wally" offline, oversizing
rads and using the pump to reach a "similar" effect,
the numbers seem to support this. ( been a while )

perhaps a combination of the two or three ( rads, pump
and trv/fhv ) items may work.

just a thought...? constant circ, go into oversized rad at lower temp, net effect is the same for the rad,
(since it's bigger)

then with the lower outlet drop right into your floor
loop, cutting out all the other buffers, valves etc.?

maybe? but i could be wrong.

regards, don

Mike T., Swampeast MO

Same Direction I'm Going

If budget and space allow it seems hard to oversize a radiator--that is as long as your system can deal with the resulting low return temperatures.

You have to forgive me for always thinking in terms of proportional control and essentially forgetting that wall thermostats exist. I'd gladly give up every fancy thing I've built or restored in this house to save my TRVs!

Only by observing and measuring the effect of TRVs on my rads have I been able to come up with these ideas that some say are "ahead of the curve".

If you want to fault me for not thinking the "American" way when it comes to hot water heating, then you're certainly welcome. I just ask that you remember that I essentially knew NOTHING about hydronic systems 10 years ago and was determined to turn an old gravity system into the best and most efficienct thing it could possibly be. Believe me--I had some WILD ideas before I knew that TRVs existed. Once I read about them, essentially every problem I was having regarding the design disappeared.

Once I changed the supply temperature with outdoor weather and kept constant circulation, things that many would consider impossible suddenly became VERY possible.

I found that if I could work with a given supply temperature at design and devise a radiant floor transfer mechanism to achieve the desired surface temperature that all I had to do was use an appropriate reset curve and just keep those radiant floors circulating constantly without any additional form of control! Super-simple radiant was born. When I questioned the ability to do such here many years ago, people said, "your numbers seems good, bit I have no idea if it will really work that way."

The most amazing thing I found with that super-simple radiant is that it seems to work regardless of the heat loss of the space! For a given panel design at a given supply temp, the actual surface temp seems to be a near constant. If the space is loosing a lot of heat, the delta-t increases. If the space is loosing very little heat, the delta-t nearly disappears. Again, it was that concept of MAINTENANCE working in my favor! The system became self-regulating! The only thing I know of in this universe that has that ability is radiation! Why? Because radiation is ALWAYS a two-way concept--it is impossible for radiation to occur in a one-way fashion! That push-pull between the heating device and the space itself becomes self-regulating when you're close to temperature maintenance!!! If you're only familiar with digital systems, you're not going to be able to find this and you probably won't even believe that it's possible. The closest you can come in a digital system is with tube-in-slab. Surely you've noticed the remarkable forgiveness in tube layout, spacing and control when you're doing tube-in-slab. At the opposite extreme surely you've noticed the remarkable unforgiveness of suspended tube.

To be continued...maybe.

don_52

I've learned from you...

again some months ago, i was researching
the constant circ. with panel rads, trv's full outdoor reset vs other methods, well i ran across some thread and i believe" you made the case" for it.

how's it go? "elegance is simplicity by design".

in logic and nodal analysis they call this method
"folding" meaning doing away with all the intermediary
steps, i'm convinced that by thinking in a thermal -
hydraulic way we could simplify most any design, do either
more or the same with less "stuff".

and you've done this, and it appears your about to take
that to a new level, i for one find it most interesting.

to be a fellow heretic here's a link to
a uk controls outfit;

http://www.seachange.co.uk/

look under "technical data", waaaay different from here.

regards, don

here's another goody;

http://www.weather-display.com/

will work with your gear plus is expandable
with one-wire stuff found at;

http://www.aagelectronica.com/aag/index.html

have fun

jerry scharf_2

This may be a mess

but I'll try anyway. I'm going to quote your message and respond in line. It seems like the easiest way. Let's hope the forum software keeps this somewhat readable.

> Makes good sense--mainly.

>

> "For a given inlet

> temperature, the HX will be most efficient with

> the lowest delta T, as the mean temperature is

> higher. It gives you the most output for the

> given size of emitter."

>

> Yes, except at the

> complete extreme where incoming and outgoing

> temperature are exactly equal. In that case your

> your input or output is zero (or at least exactly

> even)! I know that's impossible unless the temp

> on the other side of the HX is the same as what's

> inside, but it does show that you MUST have some

> delta-t to get heat transfer to begin with...

> Yet, technically, the device is capable of it's

> highest output one nth of a degree away from

> where output stops!

>

> Here's another confusing

> situation:

>

> Say you have a 300' loop of

> suspended PEX tube heating a floor. You've

> supplied it with 160° water. 150° water is

> returning. By your measure, this is an extremely

> efficient system with very high output. Uh-uh!

> Why? Because the heat can't make it out of the

> tube!


yes it's confusing. I'm not saying that a HX that has low drop is necessarily good, just that low drop makes a given HX work best. Slow the flow in that suspended tube down so that the return temp is 120, the output has dropped from the already paultry levels.

>

> Now say that's a 300' loop of pex in

> Thermofin or some other heavy conduction

> material. Same spacing, same flow, same supply

> temperature. BUT the return water is now 130°.

> Average temp is significantly lower and delta-t

> is significantly higher, but that system gave off

> much more heat!


Each emitter will have it's own curve and it's own ability to deliver heat. For either of the cases, flow will control delta T and higher flow will increase mean emitter temperature and thus output.

>

> I know that's an apple-orange

> comparison, but the point is that low delta-t

> doesn't always equate with high output--even at

> high flow.

>


never said it did. just said that as a component the output of any given emitter will always increase as the flow of constant temp water does.

> "I would expect a wider delta T to

> relax the boiler controls and stiffen the room

> controls. Overshoot in the room controls would be

> a more significant problem with a wider delta T,

> though this is secondary to thermal

> mass."

>

> Agree 100%. That's why you wouldn't

> want to use regular TRVs for a floor. BUT, the

> FHVs give you another layer of

> control--intentional restriction to flow.

> Because you're talking about a continuously

> circulating system with a finite flow ability

> (even with a variable speed pump such is

> controlled by outside temp and it doesn't change

> unless outside temp changes) your ability to

> overshoot is greatly limited because there won't

> be much "extra" flow ability in the system.

>


My point is slightly different. What does a higher vs lower gradient of temperature do to the ability to control the system?

For the boiler, wider delta T helps, since the N degrees of hysteresis in the boiler control has less impact on the room control side.

For the room side, since you are already reducing the flow to get wider delta T, you have less control autority to adapt to changes, especially external changes that have significant duration and happen at faster speeds that the damping of the emitter and room combined. The sun starting to shine through a large window is a good example of this. It's also the problem with high mass systems and feedback only systems. You really want a more predictive function to start the shift before the change occurs...

Now I don't exactly know how much control authority is needed, it would vary by each installation and each moment. I just notice that running wide delta Ts implies that the flow is fairly slow in the normal case. This reduces the ability of the valve to make changes.

There's nothing like the simulation, wind tunnel and test flight programs that make airplace controls so well understood. As an interesting aside, it was Alexander Graham Bell (the telephone guy) that did some of the earliest pioneering work on effective flight controls. he was the one that realized that the controls had to appear to "do" what the pilot wanted, regardless of what actually needed to happen with the plane to make it so.

> If

> you don't have a proportional pump you can

> engineer a reset curve that will keep MAINTENANCE

> flow quite consistent regardless of the

> weather--problem though is that your delta-t will

> increase with warming outside weather.

>


from where I sit, the circulator flow is just another knob. I like knobs. For me constant pump head makes controls easier, so I tent to use that. My current plans are that I adjust the boiler temperature and zone flows to have the highest demand zone running around 65% of max flow. Everything else goes down from there. Once I get it working and some experiece of having the thing work at all, we can try some of the ideas you are suggesting and see what the results are.

> Since

> you have constant circulation with FHVs and the

> valves are modulating flow to achieve the average

> floor temperature just adequate to meet heat

> loss, the temp of the floor is going to tend to

> average itself out over time. Compared to a

> digital system which must continually either

> raise the floor temp or allow it to fall, I most

> sincerely believe that you could have some

> surprisingly high delta-t (and resulting low

> flow) without objectionable temperature

> differences across the floor. Again, this could

> be DEAD wrong.

>


My view of controls says there are any number of cases where it will indeed have problems. I can create a number of scenarios, both with correct design and incorrect design that will cause problems. Just because you get away with it doesn't mean it works. A control strategy needs to handle the outlying cases as well...

> That table I posted with many

> zones with extreme delta-t may well be pushing

> things too far. There were at least two spaces

> that REALLY needed supplemental heat in cold

> weather. The odd thing however is that other

> spaces with whopping delta-t were designed for

> carpet and/or large, heavy area rugs--I believe

> that such would tend to lessen temp difference

> across the floor.


two parts to the equation; the R values of covering and the demand limited flow rate. As long as you have excess emitter capacity, the restriction in water flow needed to get the required power will be a major factor in the delta T. (just to be clear about terms, power is energy per time and is the unit of KW or BTUH.)

That's enough for now.

jerry