[sokoloff]

> I also have it so the heating turns off when I go into town and turns back on when I'm just a few train stops away so my place is nice and toasty for me getting home!

If your goal is saving energy/money, you don’t want a system capable of going from cool to toasty in 20 minutes.

Instead, you want a system that runs (much) lower water circulation temperatures (giving lower losses in the unconditioned spaces and more even room heating). That can be done to any condensing boiler by just turning down the flow target temperature.

A second layer of optimization on top of this is the addition of outdoor reset/weather compensation which will adjust that flow temperature based on the outside temperature, giving a flow temperature than can just barely restore the building to the desired setpoint temp.

With mine properly tuned, I was targeting having the thermostat act more like a high-limit and for it to call for heat between 22 and 24 hours per day while not overheating the house. That often meant flow temps in the 110°F (warm day) to 135°F (below freezing day) range. Compared to the prior winter (at a constant 160°F flow), the house used 8-15% less gas and was wildly more comfortable. (This setup does preclude using deep setback settings, which also can save money, because recovery times are necessarily long in such a scheme, unless you have an even smarter control system that can run perfectly tuned water most times but hotter water during recovery from setbacks.)

[Retric]

That’s an artifact of how heating is setup inside your home. Which is more efficient depends on where you’re dumping heat inside the home, levels of insulation, etc.

Energy moves from hot to cold linearly with temperature differences. Hypothetically, if the pipe was the same temperatures as the inside of your home all the heat transferred would be outside the envelope. The hotter the pipe the better this ratio becomes. This is true regardless of what percentage of the pipe is inside the envelope.

However, heating along the exterior of the home under windows and such then you’ll heat the exterior walls to higher temperatures than the interior thermostat thus losing more heat to the outside. Radiant heating on the other hand largely avoids this effect.

[looofooo0]

Look at ISO 7730, a lot of comfort comes from non-cold walls and their radiant heat and small difference of wall temperatures to air temperatures. So having a thoroughly heated home allows you to lower your air temperature. Apart from that modern gas and even more heat pumps greatly gain efficiency by lowering flow water temperatures.

[aesh2Xa1]

Having read that document, the ISO 7730 model itself depends on stable temperatures. However, I think the key is simply to understand thermal mass; people can be in hotter air, but feel cold due to cold surfaces (e.g., floors or furniture), which heat more slowly (or lose heat more slowly) than the air itself.

Therefore, 1st: Heating/cooling cycles from your HVAC are fighting these objects because they don't mix at the same speed as other objects (e.g., the air itself), so you end up with gradients across objects; people rate this feels unpleasant.

2nd: Mechanical equipment tends to operate more efficiently under constant load compared to constant start and stop cycles.

With #1 and #2, you can just heat constantly to increase both the uniformity of heating across objects and also the efficiency of the mechanical equipment's energy conversion.

There's a 3rd point, which, really, is just a sneaky way of reframing #1 and #2, and that is that you can also lower your setpoint and still have a subjectively superior comfort perception compared to a cyclic system. It drives home the point to say "constant 68F feels more comfortable than intermittent 72F." But it also invites the complaint about constant versus intermittent energy use, right? So I think just detailing #1 and #2 is better.

[looofooo0]

https://www.krantz.de/wp-content/uploads/2024/04/layout-spec...

[aesh2Xa1]

That spec aligns with my understanding, including the model's dependence on comfort perception. I was, initially, in disbelief about it, but changed my mind after reading thru. The texty reply was to make it more palatable for someone like me to accept. I think we agree.

One thing I missed in summary is the concept of general radiant temperature gradient. It's not only about the gradient for conduction, but for radiation (and convection). So you could probably improve my summary by talking about any gradient between different objects in the environment and their EM, which feels unpleasant (but I think it had value in its reduction of the problem, too).

[Retric]

It’s a deep rabbit hole as condensation, humidity, etc also enter the picture. Efficiently lowering temperatures for sleeping further complicates things.

That said, heat loss is through exterior surfaces so you really want to avoid spot heating of poorly insulated exterior walls. Thus the design of baseboard heaters can make a larger impact than you’d think.

[looofooo0]

Baseboard heaters need very high temperatures. I would not recommend installing this anywhere. Having big Typ 33 heaters for temperatures below 45°C will greatly increase efficiency of your heating system. Otherwise, a split air con is also an efficient way of heating.

[sokoloff]

Baseboard heaters are often sized such that very high temps are needed (because that's what cheapest/lowest labor/least space used), but they don't have to be sized that way. In the attic bedroom, we have baseboards around the entire perimeter on two walls and same in the bath. I run the attic zone on the same water temp (outdoor reset controlled to be quite low) as the rest of the house (mostly large cast iron rads, one cast iron convector). Good insulation and air sealing in the attic means that the attic zone calls way less than the downstairs.

[looofooo0]

Which is not exactly efficient. 40C or less is desirable.

[DamonHD]

I moved all my radiators away from under windows (and upgraded the windows to triple glazing) to avoid maximising the temperature differential and energy loss through the wall under the windows, while eliminating the cool drafts that the under-window radiator placement was intended to counter.

[Benanov]

Radiators were originally designed to heat more than needed, so you could open the windows.

In New York, at least - the standards were never changed to accomodate for closed windows in 1920. Snopes has a rundown. https://www.snopes.com/fact-check/apartment-radiator-pandemi...

[organsnyder]

My house (built in 1916) was insanely over-provisioned. When we upgraded to a modulating-condensing boiler, we halved the BTUs and are still able to easily keep the house heated to any desired temperature even on the coldest winter days.

[sudahtigabulan]

Do you mean you moved them to another wall, or just increased the gap?

(not a native speaker here)

[DamonHD]

Moved them to another wall.

Eg: https://www.earth.org.uk/note-on-superinsulating-bedroom.htm...

[HPsquared]

What do you do for ventilation?

[DamonHD]

The windows all still open, but in winter we have (nearly) enough MHRV (Mechanical Heat Recovery Ventilation) not to need to ventilate directly, eg see:

https://www.earth.org.uk/MHRV-mechanical-heat-recovery-venti...

[DamonHD]

I'm really confused about (not complaining, just not understanding) the downvotes.

These are statements of neutral fact, and the whole process is described in some detail on my site, for each room that we retrofitted.

I don't understand if I have caused offense or something: apologies if somehow so!

[joncrocks]

I'm not sure, but I think that the reason that radiators are placed near windows (at least historically) was to avoid hot/cold spots in rooms.

By placing the radiator near the place that is likely the coldest place in the room, you ensure that the room is an even in temperature as possible. Rather than to counteract 'cool draughts'. I think.

So perhaps people thought that your initial comment was wrong/misleading.

But if you have triple glazing and this mitigates the heat loss, then the coldest wall of your room may no longer be the one with a window, so you may well be doing the right thing for your room(s).

[DamonHD]

Even if the coldest wall is still the exterior one (it should be, thermodynamically), best maintaining comfort in the room need no longer be by pumping heat out through that wall (or window) to reduce thermal gradients in the rest of the room. Those residual gradients (and, eg, cold drafts down those cooler exterior walls) can be small enough to not need fixing any more.

[looofooo0]

Agree, switching on and off is the worst way of heating. If you look into ISO 7730, then a lot of comfort comes from non cold walls: https://de.wikipedia.org/wiki/ISO_7730 This means that in a reasonable insulated home, your best bet for comfort is to just keep the temperature constant and low like 20°C. This also allows you to lower your water temperatures which improves efficiency of your heat pump or boiler.

[static_motion]

Depending on the region, "reasonably insulated home" really is the factor that makes this not so viable for a lot of people. In my Mediterranean-adjacent climate country, most homes are just not well insulated at all, and having heat running 24/7 during winter is extremely costly and inefficient even if the heating is on a low setting.

[looofooo0]

These home will most likely have a split air con, which will be the most efficient way of heating them. Also there is so much other room for improvement like drafty windows and doors etc.

[walrus01]

In many warmer climates, the mini split air conditioners sold are cooling capability only. This is much cheaper to purchase for a 12,000 up to 24,000 btu/h unit than one which is also capable of heating the interior.

[looofooo0]

Not true, I can buy a Chinese 4kw Model for 650€ which can heat till -15degree. Doubt that you can save much.

[walrus01]

Go price air conditioners in the uae or Kuwait or similar, the cooling only models are very much a thing that exists on the market.

[Syzygies]

Yes!

I used to divide my time between a concrete hulk of a NYC apartment building, and a California home insulated to notoriously poor California standards. I was plenty warm in New York winters just from my neighbors' heat nearly all of the time. In California, there was a narrow window (think "Apollo 13 re-entry") between too cold and too warm.

Then we modernized ceiling fans, and I hit on running them in "winter mode" drawing hot air up to flow back down the walls. Bingo! I love that ISO 7730 confirms this.

[frereubu]

I've read that it's always more efficient to turn heating off when you're not home and then turn it back on when you return. Is the reason for it being on 22-24 hours here that it takes a very long time to get back to the desired temperature, meaning you'd actually be cold for quite a while as it returned to the desired temperature?

[leoedin]

The hidden factor here is that condensing boilers and heat pumps have non-linear efficiency vs flow temperature curves. Heat pumps in particular show high increase in coefficient of performance (CoP) as flow temperature drops.

The other variable is how well controlled your heating is. A lower flow temperature means less overshoot of the target set point - and as losses scale linearly with temperature delta, that can mean higher energy losses (depending on the characteristics of the controller of course).

Whether or not you care about losses in unheated spaces depends on your system topology. Personally, all my heating pipes are within the thermal envelope of my house, so flow temperature has no bearing on those losses at all.

If you had a resistive electric boiler, flow temperature would have absolutely no effect on efficiency. You'd be completely right, that running heating only when you needed it would be more energy efficient.

[looofooo0]

You missing ISO7730, it is a system for humans and not air temperature control. (tl;dr heating your home 24/7 allows you to lower air temperature for the same comfort. )

[frereubu]

Thanks for the explanation.

[rsynnott]

> I've read that it's always more efficient to turn heating off when you're not home and then turn it back on when you return.

50 years ago this was _always_ the case, but condensing boilers and especially heat pumps muddy the waters a little. Condensing boilers can be close to 100% efficient (vs ~70-80% for ye olde gas boilers), but generally only at a fairly specific operating temperature, which may be lower than you'd need to get a rapid rise in temperature. Heatpumps are >100% efficient (that is for every joule of electricity you put in they move more than one joule), but are even more fussy about operating temperature.

The answer now is going to be a solid 'it depends', based on behaviour of the heating system, outside temp, desired inside temp, insulation...

[lucb1e]

> Condensing boilers can be close to 100% efficient (vs ~70-80% for ye olde gas boilers)

So you save up to 30% of the gas while heating your home nearly 24 hours a day, instead of saving 67% of the gas by using it only for the ~8 hours that you're home and not under a duvet?

The math might work out for those who work from home, but I mean in the standard case with an hour's commute (round-trip), an 8-hour work day, and a 30-minute lunch break (9.5h gone, 7h sleep -> 7.5h during which the apartment should be warm if you run no errands). Of course, you'd schedule it to start before you get home, but it can also stop a bit before going to bed

I've been hearing both arguments for years and while it's exceedingly convenient to believe the condensing boiler story and just heat 24/7 to always come home in luxurious warmth, nobody ever does the math. You're one of the few people who even mention what the alleged savings are in the first place

We have a condensing boiler, chosen by my landlord so I'm no expert but I looked into it because we pay the bills in the end. The device's manual lists the efficiency as 88% ƞ4 at 60°C return water temperature, called high-temperature operation, and 98% ƞ4 at 30°C return temperature. It also gets tested yearly by a professional (Schornsteinfeger I think they call it here) and produces two efficiency measurements. Just looked up the record again: the mechanic handwrote "min" and "max" with them, so I presume that the "max" one is where the system operates at maximum capacity (minimum efficiency, then?), where the efficiency is 98%. At the "min" setting, the efficiency is shown as 106% (iirc some older measurement techniques don't include the condensation efficiency gain in the percentage, that's how it goes above 100%, or so I read when I looked it up a few years ago). For that difference, please correct me if I'm overlooking something but using a low heat for 24h/day makes no mathematical sense to me

[rsynnott]

Yeah, this only works with quite well-insulated houses, where they're very little heat loss (the system will then spend most of its time off _anyway_, as it has reached desired temperature).

[looofooo0]

ISO 7730 to the help. Just keep low overall temperatures and heat 24/7 in a reasonably insulated home.

[sokoloff]

I work entirely remote so, other than travel, there are not many long periods when the house is unoccupied.

I target the long run time to maximize efficiency. A 160°F pipe will lose more heat to the part of the building that I don’t want to heat as well as more heat to the wall right behind the radiators. It also results in the house going micro too-hot, too-cold, too-hot, too-cold as it cycles. Mine is constantly trickling in just enough heat to replace the heat lost instead of cycling between adding way more than needed then none for a while.

Another large effect is that low return water temperatures into the boiler allow for greater condensation of exhaust gas energy to be used in the building instead of sent outside. Walking by my house on a cold day, you’ll see minimal steam plume during operation. All that steam I see my neighbors emitting is energy they paid for and delivered to the outside… (They paid a lot for a boiler with a 95% or 98% sticker and run it at 80% efficiency.)

https://kw-engineering.com/how-to-optimize-condensing-boiler...

[amluto]

> Another large effect is that low return water temperatures into the boiler allow for greater condensation of exhaust gas energy to be used in the building instead of sent outside.

Correct.

> Walking by my house on a cold day, you’ll see minimal steam plume during operation. All that steam I see my neighbors emitting is energy they paid for and delivered to the outside… (They paid a lot for a boiler with a 95% or 98% sticker and run it at 80% efficiency.)

Please check your assumptions.

A boiler operating in condensing mode will produce a trickle of liquid condensate (that may well be drained somewhere that you can’t see [0]), teeny tiny drops of condensate suspended in gas (colloquially “steam”, but it’s more like fog), and some residual water vapor mixed with the exhaust gasses. You can see the “steam”, but you cannot see that residual vapor except to the extent that it condenses further as the exhaust stream cools after it exits, much as you can see some of your own exhaled water vapor on a cold day as it condenses outside your nose or mouth. The exhaust gas is saturated: it has maximum humidity and is at its own dewpoint, so there is a lot of visible fog. The droplets that form inside the boiler and escape with the flue gas do not represent wasted heat: their heat of fusion has been captured.

A boiler operating in non-condensing mode will produce no liquid condensate, and its exhaust will be well above its own dewpoint. It will contain far more water vapor than a condensing boiler, but you cannot see that vapor except insofar as the flue gas has a different index of refraction than the surrounding air and distorts the background a bit. Depending on weather, a bit of it may condense later. All of it is wasted energy.

[0] This liquid condensate is nasty stuff: it’s basically carbonated distilled water plus some impurities but not usefully buffered, and it’s rather acidic. It will quickly corrode many metals, including copper and many common copper alloys, non-stainless steel, galvanized steel, etc. Non-condensing furnaces and boilers are generally carefully engineered to avoid condensation, because the condensation would damage them. If your plumber is unaware of the degree to which boiler condensate is corrosive and uses copper pipes or metallic fittings (push-to-fit in the style commonly sold as “Sharkbite”), the system will fail. Use plastic pipes (PVC or PEX) and plastic (or maybe stainless steel) fittings such as ordinary solvent-cement PVC fittings, “engineered plastic” PEX fittings, or push-to-connect fittings with plastic wetted surfaces. John Guest makes these, and there is also the somewhat bizarre ProLock brand, which seems to be some sort of joint offering from John Guest and Sharkbite.

[sokoloff]

I’m imagining that what I see in my neighbor’s exhaust is the subsequent condensation as their exhaust gas cools to where the dew point is met and visible moisture becomes apparent.

I can see a clear difference between running my own boiler at 25°F OAT (lots of “steam”) versus 40°F OAT (almost none) while I see my cross-street neighbor showing large plumes on both. I’m not sure if I mistyped above or I’m actually thinking about it wrong, but I don’t think my observations are incorrect.

Having that water condense outside the building (giving up heat to the neighborhood) is less efficient than having that water give up its heat into the incoming (return) water.

[Retric]

> A 160°F pipe will lose more heat to the part of the building that I don’t want to heat as well as more heat to the wall right behind the radiators.

You’ve got the first part of that backwards, it’s the walls near your radiators that are your problem and need more insulation.

[sokoloff]

Indeed the building is 100 years old and impractical to retrofit insulation in any cost-efficient way (structural brick, lathe and plaster walls with about 1” of air space in the original parts of the building).

[eru]

Btw, a heated blanket would be a lot more efficient, as it warms just your butt.

[sevensor]

Heated blankets are ok, but you have to arrange them just so, and then you can’t move without fussing with the cord. It’s the last resort after layering warm clothes before bumping up the thermostat.

[pests]

I use a heated blanket as a bottom layer sometimes. Lets you move around and do whatever you want with the blanket fixed in place and the cord not in the way. I have larger heated blanket that has independent power/settings for each half. I turn one half on max and leave the other off and can roll and find the perfect direct heat and if using another regular blanket on top all that is captured too.

[frereubu]

Thanks for the explanation.

[eek2121]

Unsure if what you posted is true because I don’t know about water based systems, however this has been proven false for heat pumps.

Specifically, setting a fixed temp vs turning things up/down/off when you are leaving and reversing it before you get home. There was little difference either way. The amount of electricity consumed was similar to both.

I wish I had a link, they even tested cases where efficiency was lost heating things up. This includes “emergency heating”.

[sz4kerto]

> If your goal is saving energy/money, you don’t want a system capable of going from cool to toasty in 20 minutes.

Depends. As explained in a sibling comments, I have some rooms that have combined UFH and radiators, and if the desired temp is more than 1 celsius away from the current temp, then both are driven, otherwise it's just the UFH.

[sokoloff]

Indeed “depends” is almost always the answer.

So long as you can get the boiler return water temps low enough, you can operate the boiler in its high efficiency range.

Most dual-temp setups are set for the highest temp and mixed-down to provide the lower temp for under-floor. That’s cheapest in terms of equipment and install but cannot be as efficient as a system that mixes down when both loads call but also lowers flow temp (thereby lowering return temp) when no high-temp rads are calling.

[quickthrowman]

This is a great post which describes how most commercial boilers are controlled. I’m looking at a sequence of operations for a boiler project I did recently and the hot water supply setpoint for -20F outdoor air is 145F and for +45F the hot water supply setpoint is 120F.

Most home boilers lack an outdoor air reference temp sensor but all commercial boilers have them.

Also, condensing boilers are amazing, the size difference alone vs an old tube boiler is wild, very small in comparison.

[dheera]

Capable doesn't mean it always puts out that amount of heat.

A well-designed system would have good insulation, can dump 10000W watts of heat out and bring the room from cool to toasty in 5 minutes, and then scale back and maintain the temperature after that by putting out 500W after that.

This also tends to be more efficient in practice because if you know it only takes 5 minutes to heat up you are less likely to want to leave it on when you're not around.

[organsnyder]

But it can't heat up the walls and other surfaces in that amount of time. The building will feel colder for the same air temperature setting until those objects have had time to warm up as well.