I've noticed several hydronic radiant installations in the area using wall (air sensing) thermostats to control the zones. I'd allways thought that the better approach is to measure and control the slab temperature with a sensor in the slab.

What type of zone sensors / controls do you install, and why? (aside from outdoor resets) I'm trying to plan out our high-mass whole house system and need to choose which route to go.


Thanks!

both.

it is impossible to control a slab with floor sensor only. people do it of course, but it's a very poor method of control. Your slab must vary in temperature to change its output depending on how much heat you need. a floor sensor is a fixed setpoint. that means you'd have to change it all the time to modulate your slab's output... and it does need to be modulated.

air sensing, with a floor sensor and some sort of reset is ideal. truly ideal is indoor feedback, especially if you need to juice up the response time without having problems with overshoots, but that is a more expensive system. If the HOUSE itself is high mass as well as the floor, then basic reset, air sensing and floor sensing should be fine.

I naturally agree with Rob's statements but do not share his enthusiasm for floor sensors, finding standard ambient thermostats more than adequate for the vast majority of residential applications. This is more especially true for the novice, as the floor sensor set up is beyond the skill sets and experience of most.

If you want to go to Tekmar school, you too can get your room temperature control down to the 1/10 of one degree over-kill. If a fellow really needs indoor sensors, outdoor sensors, slab sensors (I admit some do) then he should seek professional help.

Two wires and a parameter setting are not really beyond the skill sets of most, IMHO. And "adequate" seems to depend a lot on education level. I did projects for awhile where some areas had floor sensors and some did not. the clients noticed where they were missing floor sensing, and added it later. that's why I am an advocate. But you don't need full on teknet 4 for basic floor sensing either. a 508 is not networked or even all that hard to work with. ten minutes with the user manual and you can have it up and running. not exactly rocket science.

The rest of the tekmar system may or may not be "overkill" on many systems. But that depends on the system. A basic outdoor reset with a floor sensing shouldn't challenge anyone who can program a VCR or Tivo. If you can't do that, then you shouldn't be installing a modern heating system without help for sure... even running your boiler requires at least that much these days.

In high mass applications with solar gain especially, floor sensing can mean the difference between a significant undershoot coming out of solar gain time periods and not having a significant undershoot. I wouldn't even call it optional in those cases... such as here.

You're right as rain, aggressive control stategies do create a lot of work for me...and you are still my hydronic hero!

MA

Thanks for the infor so far guys, I asked in the original question in a very general sense, so lets narrow it down to my specific scenario:

- 3000 sqft total living area
- 3 zone heating control per floor, one pump per zone
- Basement to ceiling ICF (Quadlock plus panels, R-30)
- Each floor will be polished concrete, Insuldeck with 3" slab

- Little to no solar gains from windows (large overhanges)
- Total design load of 38,000 btu with HRV
- Design Temps: -10°C / 29°C
- Degree Days: 3100

I was planning on using a dedicated, direct vent, tank water heater for the system due to the low demand. We also have a fireplace on the main floor that will put out 42,000btu if needed.


So what would be your ideal control package for this type of scenario?

I would not use a water heater for that load. I would use a mod/con boiler. with your degree day count it's getting more marginal, but you're still probably talking about 150+ therms/year difference.

I would not zone by pump. I would run the whole system on one pump straight through a low pressure drop mod/con like the prestige solo though at your max load, maybe morgan's Ultra shenanigans would work just as well!

I would use the basic reset on the boiler as long as you don't care about turning the thermostats up and down a lot for rapid response (which you shouldn't). With a good reset curve, floor sensing is still nice, but certainly less necessary. Using a PID logic thermostat though is still a pretty good idea to manage all that mass... as "advanced" as that might be, it's not that advanced.

And I would redesign the house for more wintertime passive solar gain... can't beat free heat ;)

All this assumes "best value" and not "I like the tech" is the goal.

I should have also included, our gas costs are $1.031/therm ($9.772/GJ), and electrical is $0.0615/KWh.
What about something like the Takagi T-K3? I could go to a pair of those for the radiant and DHW.

One feature that has me hesistant to go with a boiler is that we are doing a heated shower surround on a seperate zone (so actually a 4 zone system). This will short cycle a flash heater with such a small flow and heating demand.


So would you setup a manifold with zone valves instead? I'm open to either system, and since it's not much flow a single pump works too. I'd been thinking a set of four Taco 007's for the current layout.
Maybee using variable speed pump for the whole system instead?


Unfortuantly with this lot it's not an option. We may include soem skylights, but thats about it for south facing glass.


I do like the tech, but this is a house for my wife and I to live in for the next 50-60 years, so reliability is much more important. Hence the appeal of a standard off the shelf hot water tank.

Thanks for your input!

I'm told one good thing about the slab sensor is that if an outside door is opened during the winter, since the slab is still nice and warm, the heating system doesn't overreact to a brief drop in air temperature.

I've just had my hydronic system designed with an electric boiler--for passive solar during the day, and radiant heat at night at low time of day rates (4.2 c/KWh). A wall thermostat for both floors/zones was specified, and that seems like a good design for me. If the system instead was told to "charge" the slab at night to a certain temperature with a slab thermostat, the house could end up being too hot or cold the next day.

Right now I'm guessing I'll turn down the thermostat in the morning (when electricity rates go up), turn it up in the late afternoon, and turn it up further at bedtime to "charge" the slab for several hours overnight. I'm guessing this might need more tubing in the slab to allow it to pick up heat more quickly at night.

Why a PAIR of TK3s?  A single, with the zone control & heat exchanger loop set up for DHW priority should handle a load like this just fine.  Unless you're a hot-water hawg, a smaller (more appropriately sized) mod con + an indirect would probably come in at comparable money (or lower) than TWO TK3s.

If for the DHW end you go with a "reverse-indirect" (eg. TurboMax, Everhot EA series, ErgoMax) set up as a buffer for the heating system you can't possibly short-cycle the burner, whether a boiler or a Takagi, or whether for micro-zones or DHW draws.  Hand washing draws- anything under 2 gallons/draw @2gpm on a TK3 are typically under 50% efficiency- half the heat is lost up the flue during flue purges, and at low-flow low-fire it's at it's lowest combustion efficiency durign the draw.  (This is the very definition of "short cycle"!)  But  maintaing a buffer tank or indirect you'll hit 82-85%, minus a very small standby loss (as compared to a standard tank HW heater.) Maintaining an indirect at 130-140F+ for DHW with a mod-con cuts into condensing efficiecy limiting it to about 87-88%, but reduces cycling losses- the actual difference may/may-not be negligble.  But it'll INCREASE the efficiency of a tankless by minimizing the short cycles, and boosting the burn rate into a higher combustion efficeincy zone.  If you go with a mod-con you may prefer a different topology to max out on space-heating efficiency by being able to run at lower than DHW temps much of the season, but with a tankless, a buffer-centric zone with no short-cycling on DHW draws will be more efficient, and puts less wear & tear on the tankless.

A single ECM pump + zone valves will likely end up using about as much electricity as 1-2 Taco-07s.

An indirect tank will outlive any standard gas-fired tank by a factor of at least two.

If you go with an indirect fired tank NOT set up with priority controls, it'll probably still do fine at the space heating loads you're talking.  But if you can set up the main shower drain with a drainwater heat recovery heat exchanger that pre-heats the cold feed to both the indirect and the shower (or the whole house), it's like adding 25-30KBTU to the burner while taking a shower (but not for tub-fills, since you need simultaneous water & drains flows to get the heat exchange.)  GFX/PowerPipe/Retherm are the easiest to find in the US.  Bigger is better (both length and diameter) for improving the performance, but even a 36" x 4=incher is usually good for ~50% heat recovery at shower flow rates.  They're extremely low maintenance, and use no fuel- if you have at least 36" of vertical drain downstream of the shower to tap into, it's a no-brainer, if you're looking at the long-term view, or even the short term view, if it means you can now go with a 60KBTU/h boiler instead of a 100KBTU/h boiler.

Should you decide to go with the TK3 instead of a boiler you need to be aware than it's combustion efficiency at lowest fire (11KBTU) it likely to be well below it's mid-modulation efficiency.  I don't KNOW that it's a problem with the TK3, but I have 2nd hand reports that running another tankless at under ~20kbtu/h for long periods in conjunction with an air handler coil (140F out, ~115F return) resulted in rapidly destructive condensation on the copper heat exchanger. (The TK3 is also a copper HX.)  Setting it up for either higher minimum fire or return temps of 125F+ would probably keep both the efficiency higher and the self-destruction risk lower.  (I'm still looking into this. Woulda thunk that something designed to handle 40F water in from the street with 110F output could manage just fine forever with 115F return water.)  Tankless HW heaters in mid-modulation typically run ~83-85% combustion efficiencies (with jacket losses, more like 82%), but at high-fire run a bit higher.  I'm not sure how low the TK3 gets at its lowest fire, but at some point the fire is slow low that there's insufficient turbulence on the heat exchanger for good heat exchange and the stack temp goes up. It wouldn't surprise me that a 200k burner like the TK3 is only getting ~75% efficienecy at 20KBTU/h in (10% of full-on), and sub-70% at the 11K.  (Just a WAG- don't have 3rd pardy data, haven't tested it myself, but it's not much different than any other copper water-tube boiler.))

A boiler with a stainless heat exchanger will probably last longer (at 3-5x the cost of a tankless), but they all have maintenance issues- not nearly as "set & forget" as higher temp heating systems and sub-83% combustion efficiency boilers.  A ~50KBTU 80-82%AFUE cast-iron beast could run DHW-temp central reverse-indirect/buffer system too, as long as you plumbed it with enough boiler-bypass to keep the return water to the boiler above 130F to protect IT from destructive condensation.  And with drainwater heat recovery it's likely that 50-60K would be all you ever need.  (Was your heat load calculated, or measured?  Manual-J etc have built- in margin- your actual heat load could be as much as 25% lower, in which case even an 80% AFUE 60K boiler would be oversized.) Odds are good that if you set up the nominal output of the TK3 to ~30-35K and run it with a buffer it'll beat the cast-iron beast with some margin. The cycle losses of the lower-mass burner are lower, as are the jacket & flue standby losses, and since it'll be modulating somewhat the burn times/duty cycle are longer- long enough to make a total as-operated efficiency difference.

The system currently being installed in my house (as yet untested & tweaked- no actual performance data to report):

*~30KBTU design day heat load

*5 zones of staple up- (I need 130-140F water much of the season) w/ single ECM pump + zone valves

*ErgoMax indirect-DHW/buffer (central buffer to system, no priority controls)

*Takagi KD20 as heat source, running circulating boiler water, not potable water (and if it craps out in 3 years I'll just find something better- it was cheap :-)  Most in the field have lasted longer- don't know it's actual expected lifespan in heating apps, or in this system topology.)

*4"x 48" PowerPipe drainwater heat recovery (cheaper than marital counseling, should I happen to be the first one to shower on a -7F winter morning. :-) )

I'm sure by March I'll know a lot more about how well this actually works, eh? ;-)  I intend to measure it's combustion efficiency at a couple of burner output points to avoid falling off the low-fire efficiency cliff without going to a ridiculously low duty cycle.  (I may wing it and measure just flue-gas temps rather than a full-on gas component analysis, since that's easier to do and will likely be "good enough" for my purposes.)

I wrote:

"Maintaining an indirect at 130-140F+ for DHW with a mod-con cuts into condensing efficiecy limiting it to about 87-88%, but reduces cycling losses"

Seems I may have underestimated how much you can get out of a mod-con running a central buffer tank type indirect. At 140F you'll be sub-90%, but if maintaining that tank at 130F (with additional dip-tube maintained stratification during heating calls for even lower return temps to the boiler) you can hit the mid-90s pretty easily with a mod-con running near it's low-fire sweet spot:

http://hpac.com/images/Fig1.gif

IIRC Peerless has a set of controls for it's mod-cons designed to optimize buffer tank maintenance burns, which may be where I go after the KD20 is toast. (I fully expect the rest of the system to outlive the Takagi by more than a decade, maybe three.)

I just had my ultra boiler combustion tested at full fire at full temp at 89.9% efficiency after two years of service. that's not performance efficiency of course but it does demostrate that the typical efficiency range, steady state assumed, is just about 90% plus in all conditions.

round numbers of course. as always I feel compelled to note that will all the additional cost and complexity of your system, you will pick up only a couple of percent over a tank heater in the first place and it will never, ever pay back the difference in cost/labor. You will have much more DHW production though. I would use your configuration for fairly DHW situations with low heating loads, only.

And, why haven't you got your distribution temperature down, naughty dana!

Dana my sagacious friend;

I think you have buffer tanks and indirect (or companion) water heaters confused. An indirect-fired water heater has a "built-in" heat exchanger which transfers heat from the boiler water (medium) to the potable (drinking) water. As you correctly point out, when coupled with an ModCon boiler, the efficiency (and performance) are currently unmatched. However, while splitting hairs on one should point out is "thermal" efficiency. Condensing boilers are not static animals.

Here in MN it is very common for my boilers to condense for the majority of a significant DHW call. the tank is set to 140F but the incoming water (especially in winter) is far below setpoint.

A buffer tank by contrast, simply adds mass to a low-mass (most ModCons) boiler making itact more like a cast iron dinosaur for short load zones (system or architectural design flaws) oversize boilers (designer/installer design flaw) or handicapping cast-iron minded contractors.

IMHO

hehehee
--
Morgan M. Audetat
www.badgerboilerservice.com

yes, constant circ systems, low mass heating systems, or multi-zone systems are "system or architectural design flaws" ;)

nearly all systems benefit from cycling control. A good white paper our friend dana shared with me shows it takes minutes for mod/cons to actually hit their efficiency numbers on a call for heat, and even the smallest mod/cons are below min mod a significant portion of the heating season.

Cycling control is electronic, or mass-based. The mass can be in the emitter, or in a buffer of some kind. But it's a good idea in many, many cases. cycling efficiency is a very, very long way from steady state efficiency.

Wow, I didn't realize this thread had come back to life! Thanks for all the extra input, I'm trying to catch up on it. I'll try to respond to a couple of the questions:

Dana: That is a calculated heat loss. With the HRV, it will likely be closer to 26KBTU design load. The house still at a design / structural stage, but is R-30 ICF's from basement to the roof.

Rob: At what point does the cheaper tank style water heater stop making sense? With our load, it's going to take a while to use enough fuel to justify a mod/con.


One point that I think may sway the decision on this system is the floors are very high mass. The basement is a 9" thick slab, and main / upstairs are 10" Insuldeck. Theres a total of 370,000 lbs of concrete in just the floor slabs alone.

do a simple payback or ROI calculation and decide for yourself. but the trick is, you have to decide what fuel cost to base the calculation on.

That's the thing, most recomendations have been for the more sophisticated equipment. Our rates are natuaral gas at $1.031/therm ($9.772/GJ), and electrical at $0.0615/KWh.

The system would be very basic if I just used two direct vent tank water heaters (DHW and Heating seperate), and kept it simple. Just keep thinking I'm missing something in terms of efficiency...

if you have the installed cost of the water heaters, vs the installed cost of the boilers, you can make the comparison. that is what you need, plus the degree day/efficiency estimate and guesstimate as to the fuel costs you want to base the comparison on, to make the decision. Generally speaking in a real heating climate your load would justify a mod/con boiler. That said, your natural gas appears to be pretty cheap, might be different for you.

note that some codes do not allow for dedicated tank heaters for heating purposes as well. you would do one for both, separated with a heat exchanger. That's likely a more efficient solution anyway.

Looks like @ .0615 KW electric boiler is a good choice for the heating load. Those are favorable rates viewed nation wide. Electric modulating boilers with out door reset are a good choice here. These are 100% efficient, simple in operation, run like clocks, wall mounted and no stack issues, and considerably less parts that can fail. We can provide pricing and a pre assembled option if you are interested.
Mod cons are a great choice but they require a more sophisticated install, service and venting. With your heat load in the 30,000 BTU range combined with the low electric rate I would prefer an electric boiler.
Dan

Electric?   What if the rates are $.12/kw up here in norcal?  Can you specify some brands that I might research if I choose to not go with an L propane gas heater?