Wrestling with water temp...

Kestrel

I've posted some of my quandry over the design water temp for my new (to be built) system.

Initially, I was planning a mixed system with panel radiators on the second floor, and in the basement, with joist space tubes under the first floor.  All was to be run with a design day water temp of 150'F - keep things simple.



Well, at I looked at the first floor heat loss - 700 sq ft and about 18000 BTUH - I really only had about 500 ft sq for tubes, and so didn't think it would work in terms of delivering sufficient BTUH.  So I thought that adding a couple radiators contributing ~8000 or so BTUH on the first floor, in addition to the tubes under the floor, would allow the floor to provide what it could, and be a comfortable temp, and I'd be able to drop the whole system temp to boot.

Well, even at 150'F, I'm seeing that radiators need to be derated by about 40-50%, and so I'm ending up with really big radiators for the second floor and basement, and wall space is at a premium, and they are obviously getting more expensive as well.



So...if I bump up my design temp for the rads to 170 or 180, I can use smaller radiators, but I'm worried about losing condensing.  Also, will I risk short-cycling?  In my discussion with Gordan, the notion of up-sizing radiators and lowering water temp, and trying to get the TRV-controlled radiators to be open most of the time would be a strategy for both comfort and avoiding the need for a buffer tank to mitigate short-cycling.  I'm in a bit of a pickle here.



Any thoughts?

kcopp

all are good.....

thoughts. What is your design day temp? How many days do you think that you will see that? I design for -10 here in New Hampshire... I think I have seen that maybe 10 times in the last 5 years. So that means that you will most likely be condensing a lot of the season... kpc

Kestrel

Specifics...

Location: Seattle.  Design Day temp is 20'F, HDD per weatherunderground is about 4500.



Heat loss, by various estimates is about 55000 BTUH for a 2100 sq ft house - 2 stories and a finished basement.  House is about 105 years old, though has been upgraded, and most of house has R13 walls, and I think about R25 in the attic.  Big heat sieves are two big single pane bay windows on first floor.



Plan is for panel rads in the 2 second floor bedrooms, and panel rads in the two new basement bedrooms and in basement family room.  Main floor is mostly open, contiguous space - and while re-doing the basement, I have access to joist space, and so will put in 3/8 pex with JoistTrak plates.



Plan was for a single Grudfos Alpha to supple two manifolds - one for  7 Rads with TRVs, one for the under floor tubes (6 loops) - all at same temp of 150'F at design.  This design was reached with consultation with both a reputable contractor and the hydronic specialist at a big supply house.  As I looked at the design however - for example, for one of the bedrooms, they're suggesting either a runtal UF-3-60, with a 190'f output of 3850 BTUH, or a Buderus N21 24x48 which has a rating of 6500 BTUH at 190'F, it's still only about 4000 at design temp, for a 200 sq ft room needing 5000 BTUH.



What I'm thinking is that I'm really going to need water hotter than 150'F for the rads for the brief times we're at design temp - though I'm thinking that the system is really going to be spending most of it's time at 40-45'F here in rainy Seattle.  I cannot have water under the floor hotter than 150'F, and it would be better lower still - so my simple design likely needs some tweaking.



So - where this is going - appropriate sized rads for hotter water at design temp,  somewhere in the 170-180 range.  An underfloor system in parallel with mixing to keep the temp lower.  I guess that raises a number of questions - can I still use a single circulator?  Can I put both a 2-way zone valve between the circ and the floor manifold (to control when the floor gets heat), and a mixing valve to drop the temperature, as well?  What order should those go in? (pump - zone - mix - manifold, or pump - mix - zone - manifold)?

Mark Eatherton

A "Condensing Purist" approach eh...

When these marvels first came out, pretty much everyone thought that they were a waste of time and money when applied to a situation where the supply temperature was greater than 150 degrees F. The truth of the matter is, if properly programmed and applied, these beauties will be in the sweet mode of condensing for better than 90% of the season.



Is it worth it to go to the extra costs of having it in the condensing mode 10% of the time?



Personally, I think not. In fact, I think that the majority of the energy savings from these beauties is from the modulating capabilities, and not the condensing capabilities.



Why would I say something like that? Because I have an experiment going on with a condensing boiler, and what I've seen so far tells me that the majority of the condensing is occurring in the venting system, and there is no thermal gain advantages to that process. In fact, I have two condensate receivers on my condensing boiler. One for the venting system and one for the boiler, and the one for the boiler recovers one gallon of condensate for each 20 gallons that the venting system recovers. Hence, the reason for my statement.



I say design it for the reasonable (180 F at design) temperatures, and then experiment with it in real time by turning it down as low as your house needs demand.



As for high temp versus low temp, you will have to have separate pumps for those two app's, and depending upon whose boiler you need, probably one for the boiler as well.



It has been my personal experience, that whenever a conventional boiler is replaced with a mod con boiler, a minimum savings of 30% will be seen, and depending upon numerous other factors, as high as 50% reductions are possible.



ME

Kestrel

Aren't most newbies purists?

I have found such a wealth of information and insight here - thanks to all.



So, thanks for the input, Mark.  I'm more comfortable in the conclusion I was reaching - two parallel sections: rads/floor, each with own manifold and pump.  Designed to high temp at 20'F, knowing that much of the time, with ODR and normal 40-50 degree days around here that we'd be modulating way down with lower water temp for the vast majority of the heating season.



For the mixing of the floor supply with return, to temper, would you use a manual valve, or some sort of electronic one?  If I have system temp controlled by ODR, would the floor mixing the a constant ratio, or need some ongoing input to get the ratio correct at different conditions?

Kestrel

Secondary Options

Mark, thanks for your insights.

I've sketched out below secondary piping arrangements for both my original plan for a single water temp system, and two options for a dual-temp system that seeks to achieve Proportional Reset Control.



The first option retains a single ECM circulator, and places a manual 3-way mixing valve downstream of the original zone valve.  It has fewer components, but I don't know if this arrangement of serial valves will work, will allow the ECM to 'see' a pressure drop if/when the zone valve opens under the control of a T-Stat.



The second option is more complicated, in that it uses 2 pumps instead of one, but I guess is more conventional in that the lower temp section uses a pump and a single mixing valve.



All options use a supply temp sensor, and have ODR input to the boiler.  Boiler planned to use is a Lochinvar Knight WHN-085, and the whole load at design is 55000 BTUH.  The division between rads is somewhere around 10-12000 BTUH from the floor radiant, and the remainder from the TRV-controlled radiators.

Mark Eatherton

Discrepencies...

In your first system design, with all using the same temp, what you have configured will work.



In the other two examples, option 1 for 2 temps will not work unless you use a motorized 3 way valve, and I still have some doubts that it will work.



In the system with 2 pumps, the mixed down zone must always have the circulator serving the 3 way set so that the pump pumps away from the 3 way valve, otherwise you will have a short circuit, and no flow to the radiant zones.



If you use a fixed temperature non electric thermostatic mixing valve, it will essentially only kick in if the supply water temps from the source is greater than its set point, which works fine. It avoids overdriving the radiant floor zones. When the boiler is supplying water temperatures below the mixing valves set point, then there is no mixing occurring.



The ULTIMATE control logic would monitor the inside temperatures along with the outdoor temperatures, and if for whatever reason, the indoor radiant floor zones rise in temperature (solar gains, body gains, electrical gains) it would reduce the water temperatures being supplied to the radiant floor. It adds an additional level of complication to the scene, but it is really a matter of your wishes, wants, and final goals.



I have thousands of the fixed temp 3 ways out there and have had minimal issues/complaints, and they are usually related to compounded gain situations (sun shine coming through a window, for example).



My suggestion would be to down load the Knight I&O manual, and follow their lead. They have a LOT of experience in this field.



You are headed in the right direction, but are violating some basic requirements (pumping away from the valve instead of towards it) that will cause you to pull your hair out over time...



As it pertains to controlling the radiant floor, if you a need for electric control interface (i.e. internet based controls) then you will have to use electric zone valves. If this is not a major consideration, I would suggest you consider the use of non electric thermostatic controls, even on the radiant floor circuit (Oventrop, Danfoss, Honeywell etc). That control is proportional, and works fantastically with the DCECM pumps you will be using. It will give you the ultimate in control and comfort delivery systems.



As I pointed out, the only disadvantage is the lack of ability to interface the newer PC based control logics into the system without creating a whole new set of problems.



I am working on that as we speak, because my home in the mountains is a mix of electric, and non electric controls. I am using a control logic called ENV from Climate Automation Systems. I can make it stand on its head and spit wooden nickels if I want too :-) And it sends me an email if it "sees" any issues, that I get to describe.



HTH



ME

NRT_Rob

but wait!

by the time you add mixing and a second pump, you've got extra money to put into distribution.



What about radiant ceiling? Radiant Wall? different radiators? combo floor/radiant? add plates to the joist radiant? improve your insulation? increase flow to trouble zones?



lots of ways to address heat output. I almost never advocate for two temp systems... one will almost always do. but there are exceptions of course.



even if the difference between condensing and non-condensing is, say, 3% or 5% that can be real money on the table. Certainly there is a BIG difference between a 180 and a 150 design temp in terms of "time out of condensing". at 150 you're probably condensing more than 95% of the time. at 180, you lose it for at least half the winter.



how much money is on the table, per percent efficiency, per winter? maybe not huge, but if just a couple of rooms are the problem, the cost of avoiding it may not be huge either.

STEVEusaPA

quick question for Kestrel

sorry because it's only for personal reasons, but what program are you drawing with? 

Kestrel

Good thoughts....

These are the constraints that are driving my thinking...

I'm remodeling basement and have joist space open beneath the first floor - which is most all hardwood and a big open floorplan.  Square footage is about 700, but with obstacles and cabinets and such, I really only have about 500 sq ft for tubes.  Yes, I'm installing JoistTrak extruded plates.  Heat loss for that floor is about 18-20000 at design (partly driven by two large, old, beautiful but inefficient sing pane bay windows.)



The initial plan was for joist-space tubes only on that floor, and radiators on the second floor and in the new basement.  Plan was for simple, single 150'F temp at design, with help of hydronic professional and the supply house.  As I looked more closely, I didn't think I could get 18000BTUH/500SqFt=36 BTUH/SQFT.  Also, I'm pretty sure that the radiators, once derated for 150'F water by about 30%, would not deliver necessary heat.



So...I'm thinking, to keep radiators of managable size I would need hotter water.  However, I can't really go higher than 150'F for the under-floor, so I'm kind of stuck with a 2 temp system, if I want to keep the under-floor as part of the plan.

So...I'm thinking that at design I'd use 180'F for the rads, and 140-150 for the floor, and also add a couple of radiators to the first floor - maybe 10000 from the under-floor, and 8000 from the rads - that puts me at 10000/500=20 BTUH/SqFt, a more realistic number.



If this could be done with one temp, I'd love it, but the radiators in the rest of the house end up really big, and wall space in this smallish house (2100 sqft) is at a premium.  I'm game for any great suggestions - like mixing floor and radiant, which is where I'm heading.

Kestrel

drawing software

I used MS PowerPoint.

I though the drawing was kind of rough, but got the job done.  I've got a new iMac - about 6 months old - and it's the version of office that they were shipping with it at the time.

Jason_13

Modulation

Mark,

I have argued for a few years that higher temps we not condensing in the boiler but in the vent, which I might add was stoned and skinned verbally for that thought early on.

I am not truly convinced that modulation is a great big money saver. I wouldn't argue that it can save money but when comparing two boilers, one cast iron and the other a mod/con, both properly sized, piped, ODR, sidewall vented etc, that the cost of one over the other is not more than 4-5%. This thought patter changed recently now that there is cast iron boilers that can return 100f - 110f water temperatures. That raises the system efficiency which is a big factor in fuel costs. When we consider life cycle costs, the cost of one is about the same as the cost of the other. There is something to be said for thermal mass, a little more water volume(both which act similar to small buffer tank), and pre-purge pump. -Also the cast iron boiler efficiency is affected less by increased water temperature. 

With all that said if the unit is going to condense all the time due to operating water temps below 120f the mod/con will gain the btu's from latent heat which I think really sets them apart.

Kestrel

Please clarify, Sensei...

Mark,

Thank you for the insights.

A few questions on what you wrote...



In the system with 2 pumps, the mixed down zone must always have the

circulator serving the 3 way set so that the pump pumps away from the 3

way valve, otherwise you will have a short circuit, and no flow to the

radiant zones.

Doh - fixed that in new schematic below - correct?



If you use a fixed temperature non electric thermostatic mixing valve,

it will essentially only kick in if the supply water temps from the

source is greater than its set point, which works fine. It avoids

overdriving the radiant floor zones. When the boiler is supplying water

temperatures below the mixing valves set point, then there is no mixing

occurring.

Do those sorts of valve mix down to a specific temperature?  Mix down to that point and no further?

Are there non-electric valves that maintain a constant proportion?  Such that if I have the ODR and the higher supply temperature on a curve, that the valve will generate a similar curve, though with lower max temp, and shallower slope?

As I think about this - is there a difference in comfort between heat from the floor and heat from radiators?  If so, then once my heating needs fall with the shoulder seasons, I'll need the supplemental radiators less, and can employ the floor for a greater portion of the heat delivery - yes?  During those periods I can turn down the TRVs and let the mixing valve be open more to the primary supply temp.



...sun shine coming through a window...

If only, if only, here in Seattle.  Adding insult, our south-facing windows are entirely shaded by the house next door.  Oh, to move to Napa...



if you a need for electric control interface (i.e. internet based controls)

No, I was thinking not, though it does sound cool...



If this is not a major consideration, I would suggest you consider the

use of non electric thermostatic controls, even on the radiant floor

circuit (Oventrop, Danfoss, Honeywell etc). That control is

proportional, and works fantastically with the DCECM pumps you will be

using.

By that sort of control, do you mean some sort of proportioning valve?  In the schematic below, would you make pump 'B' also an ECM?  If so, what would control its flow rate - analogous to the TRVs on the radiators?  What sort of non electric thermostatic control do you mean?  In the location of valve 'A' on the schematic?  Or on the individual loops of the manifold?  I am certainly planning on using that sort of system on the radiators - TRVs on each rad, with a Grudfos Alpha (or Wilo Stratos - any opinion?).  Again, how would you implement a non-electric control on the floor radiant circuit?

LarryC

Why not decrease the tube spacing to increase heat transfer?

Non heating pro here, but am I missing something?  Kestrel states he can not get enough heat to the first floor at a water temperature of 150 F.  What if he increases the amount of tubing under floor by changing the tube spacing?  Instead of tubing on 16" centers, why not twice as much tubing on 8" centers.  Would that be analogous to installing an oversized radiator?  As long as the tubing lengths are not excessive and the system could supply enough water, wouldn't he get 150% more heat thru the floor for the same supply temperature?



Along the same lines, he could just double up the tubing density in front of the bay windows to increase the heat delivered in that area.  Or add a suitable radiator in front of the bay window to add supplemental heat to that location.

Mark Eatherton

Physical limitations...

Larry, Yes, you are correct that higher tube density = higher emitting surface at a lower operating temperature, however, any time one is in contact with the emitting surface, that surface should not exceed 85 degrees F, which limits the output of the floor to around 30 btu/sq ft/hr.



Adding a good extruded plate is the same as throwing another pass or two in the joist bay.



It is not a lineal function. THe water gets cooler as it works its way through the circuit, so the first foot puts out a lot more energy than the last foot.



I'd prefer to do what Rob said, which is to look to other surfaces, like the ceiling. It can produce higher BTU/Sq Ft/hr because there is no human contact. I have rad ceilings in my mountain home, and have to tell you it is one of the most comfortable situations you've ever seen. Interesting thing is, as it gets colder outside, the ceiling makes the floor warmer to the point that you'd think we had radiant floors instead of ceilings.



His other alternative would be to use a below floor convective system like UltraFin. This would keep all emitters on the same playing field as it pertains to supply water temperatures. I'd throw extra UF's in, and also pipe the grid parallel reverse return to insure the most efficient operation, but it does work. The floor is still limited, but the operating temperature can be higher without incurring wood damage.



ME

Mark Eatherton

Jason...

I hear you about cast iron. I was well into low temp cast iron boilers (Buderi) long before I was in to modcons. My back appreciated the switch :-)



As it pertains to condensing cast iron or near condensing cast iron boilers, the jury is still out in my court. I want to see exactly how they hold up, because historically, they haven't held up well. I have this problem with placing something that is KNOWN to lose parent materials (cast iron and cast aluminum) during the course of normal operation.



My gut and previous experience tells me that modulation is good for about 2/3's of the savings, and condensing for about 1/3 of the savings, obviously dependent upon the net application.



Modulation on atmospherically vented appliances is a disaster looking for a place to happen. True full modulation, like we see in mod cons doesn't even fall into the same category. No excess air to destroy the efficiency numbers...



ME

Jason_13

My thick head

I always play the devils advocate, that is what helps me keep an open mind about subjects but at the same time I am very opinionated.

I find it very hard to believe that modulation is 2/3's of the fuel savings. That would mean that mod/cons would be 2/3's less expensive to operate than cast iron boilers assuming the mod/cons never condensed.

When we modulate a mod/con we have less flue gasses in the boiler even though they are moving slower. Don't we lose thermal transfer as some of the flue gasses do not touch the water side of the heat exchanger. The flow through the boiler should at that point be more laminar than turbulent. When the boiler has more flue gasses in the boiler trying to get through the flue gasses they become turbulent so the thermal transfer increases.

I have read some engineering data from two manufacturers (on the QT) that showed the difference in savings between mod/cons and cast iron all else being equal is between 3% & 5% dependent on system water volume.

I have two jobs last year where with the tax credits in place, the owners removed cast iron boiler a few years old and installed mod/cons. Both were properly sized and there was no significant fuel savings when compared to btu's/per degree days. The cast iron boilers had ODR installed and the mod/con water temps were lower than the cast iron boilers were. The only difference was the cast iron boilers had pre-purge on the pumps. This was accomplished by using a 2 stage ODR control with stage one being boiler pump and stage two firing the boiler. The original installs with the cast iron boilers, smaller than the one removed in both cases, was the largest amount of savings. Job one was 49% and the second was 43% again using btu's per degree days.

One last thing the boilers mentioned in my original post we non-condensing cast iron boilers. I too have heard but personally have not experienced  problems with cast iron condensing boilers as I have never installed any.

I know a guy that has installed many many Revolution cast iron boilers that can bring back 55f return water temperature and cut fuel bills 35 - 50% consistently. He uses 2 - stage ODR as explained above. This boiler has been on the market since 1998 I believe, no condensation issues.

Sorry for the long post just curious as we all want to do what is right and what is best for the customer and the world. I can't agree that mod/cons are that much more green when you take into account higher pollutants to manufacture and will need replaced more often.

Kestrel

Great thoughts...

As for the radiant floor - we're pretty committed at this point - half of the JoistTrak is already installed, and I've got almost all of the protruding nails ground off at this point - dirty, spark-y work.

Also, in terms of other locations for emitters, like the ceiling - I don't think we're prepared to take apart the ceilings in the rest of the house - that's just too big a job - we extensively remodeled the main and second floors about 8 years ago (wish I had changed the heating plant at that time!) and we're not going to undo that, I think.

Right now the floor of the first floor is exposed as I work on the basement - gotta build a man-cave for the teenager!  And so, as our old gas-fired forced air system was at the end of its life, and I wanted the ductwork out of the headroom, we decided to go radiant/hydronic, but as I say, we're a bit limited in overall scope.



My issue is that we're running into physical limitation of both the output of a joist-space tube system, even with extruded plates, and the size of the radiators.  I posted my calculations and worries about using the floor as the only heat emitter for the first floor - I just don't think those numbers add up.  As for the radiators...

For example, for a 160 sq ft bedroom with a 4000 BTUH design heat loss, I have had proposed to me by various installers a Runtal UF-3-48 or 2 12x24 Buderus N21's  - which when I went back and looked at their output at 150 degrees, were 2492 and 2287, respectively.  If I raise the ave Temp to 180, they're at 3720 and 3177, if I'm reading the manufacturers' charts correctly

Again, if I'm understanding correctly, both the Runtal and Buderus radiators (and I assume others as well) put out about 800-1000 BTUH/SqFt at 180, and about 67% of that at 150 - so I'd need about 6 SqFt of radiator for that room at the lower water temp - which is starting to take up a lot of wall space



Given the physical limitations, I'd love to use the lowest water temperature possible, and the smallest boiler that is appropriate - and have a system that is efficient, not short-cycling and comfortable.  That's why I think I'm moving away from the TT110 Solo and will be using a Knight WHN085.  I'm still debating the buffer tank - my discussion with Gordan last week has led me away from this, but comment if you think that that is a good or bad idea.

I'd also love, Mark, if you have time, to address my specific questions several posts above where I quoted your comments - I don't want that to get lost in the lengthening thread!



Thanks to all - I'm really looking forward to posting pictures of a beautiful system here when I get is all worked out and installed!

NRT_Rob

yes

I pretty much ran through the options in my post:



What about radiant ceiling? Radiant Wall? different radiators? combo floor/radiant? add plates to the joist radiant? improve your insulation? increase flow to trouble zones?



there may be others, but those all apply as options available to get water temps down.



if you have to do two temp you have to, but generally I would consider the impact on fuel usage and initial cost and *usually* one or more of those options becomes more cost effective.

NRT_Rob

I think

a big part of the "modulating" benefit is really a "sealed combustion" benefit. not drawing atmospherically from the surrounding space is a big deal.



if I had to throw a dart I'd say OF THE SAVINGS it's probably 20% condensing, 30% modulating, 50% sealed combustion.



I think that's a big part of the reason why savings always beats the AFUE increase significantly as well. AFUE should catch modulation improvement. but it does not catch the down side of atmospheric makeup air draw pulling cold air into the envelope.



just idle speculation though.

NRT_Rob

I don't think mark is TOO far off

we routinely see 30% to 50% reduction in energy costs going atmospheric cast iron to sealed combustion mod/con. big, big savings. If you add all the features the mod/cons have built in to the cast iron, you can narrow that of course. But then you're at a more similar price point too.

Jean-David Beyer

Don't we lose thermal transfer

It seems to me that a gas and air mixture is going to burn at about the same temperature irrespective of the amount of modulation (within reason). The total amount of heat will change, but not the temperature. When a modulating boiler is running at a low temperature of water supplied to the load, the water in the heat exchanger is colder than when it is firing at a higher rate to produce high temperature water to the load. So at low modulation rate, the gradiant across the heat exchanger from the fire side to the water side is actually greater than when firing to produce maximum water temperature. So we gain thermal transfer with modulation.