low tem radiant in same systtem as high temp system 1

[EnergyProfessional]

Could you review and comment on my design idea?

I have an existing building with hydronic heat that requires about 140-150°F supply temp. I also add some unit heaters that I design for 140°F. Overall that part may have a dT of 25°F (that design is a bit screwed up, and we add perimeter heat - so that dT is not necessarily accurate n real life). Will require ~100 gpm. Boiler setpoint probably will be 90-140°F depending on OAT and valve positions.

The new building addition will get radiant infloor heating. Currently I plan on designing for 120°F and 20°F dT. It seems many of the loops can et by with lower supply temp. Will also require ~100 gpm.

I attached two options. both have in common:
- condensing boilers will get minimum flow via a bypass valve (each boiler has 25 gpm min flow). Flowmeter will modulate bypass valve to either provide 25 or 50 gpm depending on how many boilers are on.
- they will operate in parallel whenever possible for better efficiency (turndown 1:20) and have each an isolation valve.
- objective is low return temp for better efficiency and also reliability and less complexity
- radiant floor shall be protected from getting more than 150°F to prevent concrete cracking. Under normal operation this would not happen, but have to account for someone setting something wrong or other issues forcing the boiler to provide 180°F etc.
- this needs to work when radiant has no load, or when the high-temp system has no load, or anything in between.
- all zone valves will be 2-way.
- For radiant floor it doesn't matter if the supply temp drops or rises for an hour. If I only get 110 instead of 120Ç° for a while, it isn't really important.

Everything will be fed by a new boiler plant (2 boilers).
Option 1 creates a separate loop with pumps for the radiant heating. Advantage is it could lower return temp more. disadvantage is it has 2 more pumps and at any given time I need radiant heating, 2 pumps need to operate. It also could get too little hot water (only bypass flow) when there is no flow through the high-temp system.

Option 2 creates a scenario where when boiler temp is low, the radiant loop gets straight boiler water. As the temp rises, the 3-way valve will blend boiler return water in to lower water temp to radiant floor. I may adjust control system to keep it warm (maybe 130°F, or also could set it lower, or make that dependent on OAT...). If I'm in a pinch I also could send 140°F water to the radiant floor.
The disadvantage is my boiler return temp will be a bit higher (about 107°F instead of 100°F). Advantage is I need fewer pumps.
Option 2 has the hydraulic property to switch between low and high flow. It could be enhanced to use 3 smaller pumps to improve low-flow efficiency operating just one pump (they will be VFD). One pump is standby, but I think even if 2 pumps fail there could still be decent operation left.

Maybe in theory Option 1 could have more efficient boiler operation, but Option 2 seems to have better pump efficiency. Preliminary pump sizing showed me either way pumps would be 3 hp size (even different models)... but when operating in parallel at design flow two pumps in parallel use less than one pump (per Grundfos sizing tool)

Obviously in real operation at medium load many valves will be throttled, dT should be larger and overall return temp lower.

Option 3 resolves the problem with higher boiler return temp. but hydraulically it may be more challenging as it switches between high flow - lower dP and low flow - higher dP. But my 3-pump idea may help. Option 3 basically switches the radiant flow to be in series to the high-temp system if needed. If for some reason there is no load on high-temp system, I would not have colder water to blend in and added a potential circulation pump (this is only for when boiler goes rogue and provides too hot water). not sure this pump really help. I could accomplish radiant overheating protection by monitoring radiant return temp and throttling the radiant valves. but this could give me a very high dT system.

For option 2 and 3 it seems I need a checkvalve.

I like to know if you have suggestions to improvements or opinions on either scenario. My mechanics like option 2 better for fewer parts and the possibility of three pumps. ( I realize I could have three pumps in option 1 as well, but 6 pumps in total seems unrealistic for space and budget reasons)

 

[Drazen]

i don't understand which principles and standards you use to design radiant heating.

maximum allowed floor temperature is 29 degree C in occupied regions, which equals 84 F, and that imposes supply temperature.

dp is what you are trying to control to not have too much imbalance between circuits, and for dT, you get what you get as there are other things to be controlled. if you would try to fix dT, your options to design circuit geometry and composition to satisfy heat load needs would be largely reduced.

design principles are different from those when designing radiators.

for the mentioned reasons you have to control supply temperature for radiant heating, otherwise you would have control mess.

 

[EnergyProfessional]

Drazen: I believe what you referring to is the floor surface temp (according to all resources max is ~87°F, a little lower on wood). but to get surface of 87, water supply has to be higher. Max temp for water supply in concrete is 150°f (according to uponor design manual). so i try to keep it below 140°F under all circumstances (design 120°f on design day... probably much lower rest of the year)

Actually i noticed my option 3 has a hydraulic error... it actually will require the pump(s) that I made optional.

my basic idea is to use hot water in the high temp loop. Once that water cooled down, it (or parts of it) goes through the radiant loop to cool down further o have a large system dT. Low return temp keep the boiler more efficient in condensing mode.

 

[Drazen]

87f is allowed only for brief passing, which applies to border areas only.

you have to calculate supply temperature by designing geometry and hydraulic setup taking care of max. allowed temperature and your demands imposed by heat losses. once you get it by calculation you have to apply system which will maintain it if you want controlled situation in your environment.

if you allow it to vary, you won't have controlled system. if it goes over calculated one, your floor will adversely affect health of users.

 

[EnergyProfessional]

Maybe I should have clarified, this is a vehicle garage (55°F) and some shop areas (65°F) and I used LoopCAD for the layout. Each zone will have a control valve, so it won't get too warm. Most zones get away with less than 120°F supply temp. and when the sysetm operates in the real world i may adjust that setpoint. i also may be able to adjust control valves. for example is zone is too warm, can program the control valve to no open more than 75%. thsi being a garage that normally would get gas-fired unit heaters, I think comfort will be improved.

Obviously the whole heating load calc is theoretical, they will open overhead doors different than i assumed, infiltration is hard to predict, and 99.9% of the time we don't have design OAT.

Back to my original question, regardless of how I came up with 120°F (or 110, or 100..)... my goal is to efficiently have this low-temp system and the high temp sytem (140-150°F) in the same boiler system. I like to have as little pumping power and has high dT (low return temp to boiler) as possible. With the feweset number of parts than can break or fail in a way that causes problem.

 

[Drazen]

i am not familiar with garage heating at all, the whole concept is somewhat mysterious to me, but there is likely good reasoning for that in very cold climates. What theory is targeting at? If garage is open much of time radiant heating does make some sense as it is least bound to air temperature, but i cannot imagine energy efficiency justification.

anyhow, 87F is allowed maximum for short passage areas like garage is. for shops, anything over 84F will destroy occupants' blood circulation of feet over time.

anyhow, if you put surface temperature sensor to switch off control valve, you will avoid excessive surface temperature, but if you don't design system to get desired output with acceptable surface temperature, you will never reach output, control valve will always close before capacity is reached.

that is why supply temperature has to be carefully calculated and separately controlled if you want meaningful control.

if you wan lowest possibly return temperature, go for lowest possible flow. you have number of parameters to play with - supply temperature, piping distance, piping diameter and flow. the other paremeters are not negotiable - maximum surface temperature, and, eventually heat output. dP is what you arbitrary select, to ensure efficient use of flow adjustment valves.

sometimes i need to go through dozen of scenarios until reaching the best combination. that is where use of good software really helps.

 

[EnergyProfessional]

The ceilings are quite high, and I hope to get away with lower setpoint than using air-heaters. Much of the load are overhead doors etc., so lower setpoint should save energy. We also don't want to clutter up the ceiling with unit heaters. there will be mezzanines with vehicles parked under, and a crane etc. We also like less maintenance.

Regarding overheaddoors (OHD), All I read indicates recoery rate should be high since the warm slab automatically increases heat capacity as air temp drops. I'm a little undecided if i should include the OHD load fully in radiant layout since that only is a temporary load and the slab itself doesn't cool down quickly. OHD load is a delayed load as far as the water int eh slab goes.

Regarding in-slab sensors, i thought about those. but wonder where to locate them so that they are replaceable and woudl they be on top of slab in between tubes, or right above tubes? i was planning to monitor return water temp and control valve so that a specific return water temp is not exceeded. One could assume if return temp is 95°F, then slab should not be more than 85°f or something like that. It seems more reliable to measure return water temp since that is an average for the entire slab, and not just for a single specific slab location.

 

[Drazen]

radiant floor will hardly react to occasional door opening. air temperature can hardly be controlled, yet feeling of comfort due to radiation will remain.

i would be interested to learn what are real targets of such garage heating, my first assumption is that the main goal is to retain some feel of comfort, not to strictly maintain indoor temperature. that is my guess, of course. knowing exactly what common practice wants would help in better setting up system. in churches fixed temperature is also not easy to maintain by radiant heating, and often design goal is to reach certain temperature before ceremony begins, accepting that some temperature variations are inevitable during ceremony.

sensors should be mounted as close to floor surface as possible. temperature in line of pipes is much different.

for larger surfaces some pattern like double meander ensure uniform temperature over floor. generally, during start-up, floor temperature should be measured by outside instrument, and sensor re-calibrated until tripping would take place at exact temperature.

for smaller system, i usually specify one sensor only, putting it on circle with smallest pipe distance, i. e. largest "density" of pipe pattern. of course, previous calculations have to be conducted correctly.

 

[EnergyProfessional]

Main goal is to keep air temp low while maintaining comfort. There also are high ceilings (up to 30'). If I can just lower temp by a few degrees and avoid to much stratification I can save a lot.
We also don't want to clutter up the space below mezzanine ad have a crane. So not having unit heaters will be nice for that.

We do have 2 fire stations with infloor radiant heating in garage and it is comfortable. in other garages we have gas or hydronic unit heaters, also some ceiling gas-fired radiant heating. None of that is really good. We also like just two boilers to maintain (which we have anyway) instead of a dozen of gas-fired unit heaters. those gas-fired units also need wall and roof penetrations. So there is a lot that makes radiant floor nice.

Based on your advice I re-worked my radiant layout. I think I can get by with 110°F supply temp (of course, I will be able to adjust if needed). I also plan to make that setpoint dependent on OAT and actual load (so when it is moderate outside, I may only supply 90°F). So that should help with control.

I have a hard time imagining the detail for an in-slab sensor. I imagine I could install a metal conduit piece into the slab when it gets poured and then fish the sensor in. But wonder f that is great to retrieve the senor if I need to replace it? If there was some sort of radiant sensor (kike an IR camera)....

Assuming I want to supply 110°F, what is the best way to mix that down? I attached a revised piping option. In option 3 I realized I need to add more pumps, but I tried to get away without the mixing valve, by just controlling the speed of those pumps, like injector pumps. not sure this really is a good idea.

 

[Drazen]

mounting sensor in sleeve is logical though not all manufacturers offer sleeve in set.

if sensor would fail in future it is always possible to mount new sensor at supply pipe before reaching slab and than adjust is according to measured surface temperature. though i would prefer sleeve wherever possible.

injection system is completely new to me, appeared as small variable speed pumps became very sophisticated. i would not know pros and contras compared to classic three-way valve. i believe, though, that one more pump is missing in your setup, the one for main secondary circulation in radiant circuit.