Heat accumulator
 

I'd like to share something I recently found out about how my local power plant uses a new method for raising its efficiency, thereby reducing emissions. But first some background.

Most of the power plants in my climate zone are geared towards producing heat, electricity is a by-product. Of the four power plants in Warsaw only two generate electricity, the other two are shut off at the end of each heating season. The heat feeds the city's warm water network, providing heating for most of the high-rise apartment buildings during the winter, and warm water in the faucets year round. This is more efficient than local, distributed heating. One of the downsides is that it's very hard to adjust production when demand changes hour-to-hour. Most of the electricity is needed in the morning and evening, with a dip at mid-day and a minimum at night. On the other hand, heat production is needed most during the night. Some power plants have extra furnaces that get fired up when demand goes up, but this takes around 3 hours and wastes lots of fuel (almost all power plants here are coal fired, but more are building biomass furnaces). When generating heat at night, the power plant can generate lots of electricity at almost no cost, while during the day the situation is reversed.

About 1-2 years ago the largest power plant in Warsaw decided to add a heat accumulator, already in use at many Skandinavian plants. It's basically a huge thermos which gets warmed up with waste heat during the day, relieving the night load. Or it can be heated at night the help generate electricity in the day. This is a VERY simple setup: just a 47m-high cylinder, 30m in diameter, holding over 30.000 cubic meters (30mln liters) of water. No extra machinery or heat exchangers, just an extra valve. Its heat capacity is 1300MW(thermal). The cost was about US$17mln. I can't find much info on it, nothing on how much it actually saves, but I did find out from an insider that the power plant is very happy with it. So much so that other plants are planning on building something similar.

On a smaller scale, I've read that water-based household heating systems sometimes employ an insulated buffer tank to collect heat when it's available (solar) or cheap (night tariff). I belive that the rule of thumb is around twice the volume of water in the pipes and radiators of the house.

On an even smaller scale, the genII Prius has a thermos for storing engine heat.

Good idea. It sounds like it should be pretty cheap to implement (vs the savings it generates). That is a BIG thermos though. :)

I found a thread on heat buffers (heat batteries, accumulators) on a Polish builders' forum. The thread's author, Adam_mk, seems to know quite a lot, and since this is really interesting, I've translated its essence for others to read. I've only condensed the first post, with some information from later posts, but if someone wants to go through all 160-something pages (with 60 posts per page:eek: ), then they can try it through google translate. Thanks to Adam_mk for starting the thread, and thanks to many others for adding information and pictures of how they built their own buffers.

The philosophy behind a heat buffer or heat battery is to "charge" it with thermal energy whenever it is cheaply available, and use it when it is needed. The more energy that can be stored for a longer period of time, the better, so the buffer has to be large (1000-2000 liters of water, 1.5-2.5 tones of thermal mass) and very well insulated (5cm of mineral wool with a layer of aluminum foil, then 30-40cm of styrofoam).

Most commercially available hot water tanks don't have a very efficient design, the internal heat exchanger is coiled around the perimeter. The lowest coils can still give off their heat, but the upper coils sit in the heat released by the lower coils, which reduces heat exchanging dynamics. That cloud of water heated by the coils rises along the walls to the top, while cooler water falls down the center of the tank. Between these two streams of water is an area of turbulence, causing part of each stream to mix with the other. This leads to the hot water not being as hot as it could be, while the cooler areas inside the buffer are warmer than they should be.

The best remedy for this is a buffer that keeps water at different temperatures in different layers, without mixing. Consider the following design:


The coils from a heat pump or solar collector are at the bottom of the tank, and spiral down and outwards. This places them in the coolest region of the buffer, so more heat can be extracted. The water which recieves that heat can rise without washing over more coils, so those are also sitting in cooler water. The rising warm water is collected by a funnel and goes up the middle of the buffer inside of a "chimney", which delivers the hottest water straight to the top, pushing cooler water down, but with no mixing along the way. The chimney is perforated from about halfway up, with the holes becoming gradually larger towards the top. This allows warm, but not hot, water to exit the chimney at the height (level) where water of that temperature is collecting, pushing cooler levels down, but not interfering with the warmer levels above.

Water for domestic use (sink, shower, laundry) is warmed in a coil on the perimeter of the tank. Entering at the bottom, the coils move closer together as they work their way up. Notice that this design qualifies as a double-wall heat exchanger between potable water and the freon or glicol filled coils from the solar/heat pump system.

When heat pump or solar heat is not available, the buffer can be heated ("charged") with a furnace, stove or fireplace - cool water is taken from the bottom and hot water returned at the top. This allows heat from a fireplace/stove to be saved for later, and allows the furnace to always run at its maximum efficiency, instead of having to turn on often for short runs at a lower temperature.

Somewhere at the bottom of the tank there should be 2-3 electric heating elements (1500 watts each) to charge the tank when other heat sources are not available and when electricity at night is cheap.

Using hydronic floor heatering, with its large surface area and low temperatures, will allow much more heat to be used from the buffer, as opposed to normal radiators which require higher temperatures.

One more secret:


All of the inputs and outputs are curved in such a way that water that is leaving/returning to the tank causes the whole mass of water inside to delicately turn in one direction. This buffers the entering/exiting water's kinetic energy without causing turbulence and mixing between thermal levels.

For this design to work well, it should be taller (at least 6ft/2m) than wider (27-35in/70-90cm), so that each thermal level has room to store energy, while the area between levels is minimal.

A heat battery with 1-2 tons of water is a good start, but can later be expanded if required. Here is a link to the Polish thread on how to connect two or more water tanks for maximum efficiency. Of course, using one large tank would be even more efficient (less total surface area = less heat loss), but space or financial restrictions may not allow that.

If you are trying to store the most heat in the smallest possible space, then you can look at using phase change materials to increase capacity. Paraffin wax is a by-product of the oil refining process and many refinaries have more of it than they can get rid of. This material is very good, since its melting point of around 55-60°C is the temperature that you'll usually have in your heat buffer, and its volume doesn't change enough to be much of a problem. The paraffin tank should have lots of small diameter tubes (more tubes = more surface area = better heat exchange) for water to deliver and recieve heat. Here is a link to the Polish thread.

Heat buffers of the DIY variety are the best: they are individually tailored to each house's energy needs and space limitations, and they are much cheaper than a commercial alternative. People who have made one (info in the thread linked at the beginning of the post) said it costs around 5000-6000PLN (around US$2000), while a commercial tank costs almost twice as much and is less efficient. A homemade heat battery is usually adapted from a large commercial tank, or from a pipe segment with custom endcaps welded (bolted) on. The metal can be normal steel, around 4mm thick. Commercial units are often made from stainless steel, but this only raises the cost and makes it harder to weld, while the "stainlessness" is never needed: you won't have a nice shiny cylinder to look at since it'll be under at least 35-45cm of insulation, while on the inside the surface will get covered with a layer of scale after the first time it's heated to 80°C, after that the water will contain no gasses or minerals, so there shouldn't be any corrosion.

Check out AC_Hacker's post for some great links to similar buffers.

Here are pictures of what the hp/solar coils should look like:


It is actually two coils, plumbed in parallel, one on top of the other. The spirals are 15mm copper tubes, the common input/output is 22mm.

Here is the coil fixed to the bottom cap. Notice the heating element inside the coil. Two more elements will be installed (arrows).


The chimney (before perforating the top half):


The heat exchanger for domestic hot water. Notice how the coils get closer to each other near the top.


The inside of a (partailly completed) heat buffer. Notice the how the inputs/outputs will cause the water to rotate. The direction of rotation should be tailored to the Coriolis effect in your part of the globe.

The heat buffer is a great place to dump excess heat from air conditioning or ventilation, to use later as hot water or heating. If more than one heat source is used (not counting furnace, stove, fireplace, etc.), then more than one set of coils at the bottom of the tank may be needed. The most efficient way to place them is one under the other, but with a second funnel between them, going to a second, smaller chimney. The cooler of the two sources (heat pump) should use the outer chimney, while the hotter (solar) should use the inner chimney to send its heat to a higher level.


Electric heating elements can not charge the buffer using cheaper night tariffs, but can also be a dump load for wind, hydro, PV, etc.

It's good practice to put 4-8 temperature sensors inside the buffer at different levels to know how much energy is stored. Only one sensor is not enough, since heat is stored in different temperature layers, so it's possible to have a thin layer of 80°C at the very top while most of the rest will be around 50°C, or to have 80°C in the top 70% of the tank. In the latter case the buffer has much more stored energy.

Very interesting!

I have a 1000 liter tank (~300 USG) which is made for this already, from the Austrian Okofen (US dealer is located at oekofen-usa dot com). The one I have has two heating circuits, a solar heat exchanger and hot water is flow-through in a stainless steel pipe, not "sump water" (always fresh water, both hot and cold).

The solar heat exchanger is at the bottom, but it has a "chimney", an internal vertical tube connected, with holes in it. The hot water in the buffer water will exit from this chimney at exactly the level where it does most good, as the density of the water changes slightly with temperature. If there is plenty solar heat, it goes to the top to give a good temperature for the hot water supply, leaving the bottom less heated for times where there is less heat from the solar panels.

I have a question about storing the extra heat generated during the day. What is the best commonly available medium to use? I have a basement and crawl space to put something in but nothing large. Would a 4ftx4ftx4ft concrete block, very well insulated and buried in the ground, with PEX tubing running through it work? Can't really have a
1-2k gallon tank anywhere. I understand the efficiencies probably wont be good but I want to make a system as cheaply as possible using common materials.

> What is the best commonly available medium to use?

Best can have many meanings.

You might want to find out what "specific heat" means, it might help you to answer your question.

-AC_Hacker

I've read about storing heat in phase-change materials.

One idea to keep automotive engine coolant warm was a phase change salt, see this thread over at EcoModder for more info.

Another idea, on a larger scale, is to use wax (paraffin). Basically, you need a well insulated container with lots of small tubes crisscrossing, then you fill that up with the right type of wax. Paraffin is a by-product of oil refining, I read somewhere that many refineries would be glad to get rid of their extra paraffin. Make sure you get the type with melting point around the temperature you want to keep your heat stored. Once everything is working, water flowing through the tank will be warmed up - you'll see it's exit temperature gradually decline, then flatten out for while (melting temperature of the wax), then decline again. The rate at which heat is tranfered from wax to water (or water to wax when charging) depends on how close together the tubes are, how much surface area they have, etc.

Parafin tanks work great when paired with a heat buffer.

I can't remember the exact numbers, but I think that a paraffin tank can store something like 30%-80% more heat than a water tank of the same size. If anyone is interested I can try to find more info.

This topic was recently brought up on another group I participate in. We were discussing using a similar setup to heat a tank of paraffin using engine coolant, and then using the tank to maintain the temperature of the living space in an RV overnight without using fuel.

The only major downside brought up about paraffin was that it's flammable. With that (and expansion) well-managed, it seems like a workable idea to me.

I believe that commercial/industrial grade paraffin has a higher flash point so that will help with the flammability issue. If the wax is away from a flame there should not be any issue with it igniting.

Before I go any further does any one have do's and do not of putting a wax fire out?

Don't put water on it. Ask me how I know.

The best place to start is the heating system itself. The optimum is a large mass, like a thick stone or concrete floor with built in liquid (hydrating) tubing for heating. It takes ages to heat it up, but also ages to cool down. So you can heat it when you have the sun, and let it cool down at night. The more the mass, the better!

A small area, high temp radiator based system, as well as air based systems, are less efficient if you have a variable heat supply like you have from solar power. Turn your aircon/HVAC off, and your house will heat/cool quite quickly.

A second advantage is that underfloor heating (large mass and area) requires a lower temperature. This is not as important when burning fossils, but essential in solar heat and heatpump installations. The more your heating system can cool the solar heat panels, the more energy you can extract.

Also, if you store heat other places than the floor (for instance in the clay/earth/sand under your house), a large mass - low temperature heat storage loses less energy than a small mass - high temperature storage.

In the project I am devising, I am using several ways of storing heat. Long term storage is under the house (which is clay down to about 7 feet). It is being insulated on the sides, which means most of the energy loss is actually up through the floor and into the house - which is no loss at all. Storing heat during daytime, I have a total capacity of just over 1000 USG (including the main heat buffer).

During summer, when production is high and demand is low, I can absorb as much heat during day as possible, into the clay, as well as the 1000 USG water capacity. During nighttime, the water tanks can keep transferring heat into the clay in anticipation of the next day - to prepare them to be able to absorb more heat. I intend to sync weather prediction data into the system, so if the following day is overcast/cold/windy, I will keep the heat in the water tanks for use during that day and possible the next ones.

I am sorry that I cannot give many more details - or proof - yet, as the system has not yet been finished. When it is up and running, I will supply both proof as well as more detailed information of how it is working. Including some graphical presentations which hopefully will be easier to understand than my mumbling above!

Water and PCM are amongst the better heat storage mediums normally. But for large storage capacities, they become expensive and it is not so easy to build a house on top of a 25000 USG tank of water. Although clay, stone and sand generally only hold about half the heat of water, they are usually there, under your house, already. All you need to do is shoot pipes into it, in a way where you can store and possibly extract heat.

As for PCM specifically, the high grade professional stable material is so expensive that it is only worth while if it changes phase on a daily or at least weekly basis. 3/4 of my 1000+ USG water storage capacity is prepared for insertion of PCM modules, should I later choose to do so. They would most likely be set around 120-140F or so - this means I can store more heat on a good solar day compare to a plain water tank. But I need to see first how many days a year the water tanks reach max capacity. I might even consider a mix, maybe a set of ~ 100F and some at ~ 140F to get different "steps" of capacity increase for different situations.

See also "annual geo solar heating" and similar, for more information on this matter.

Outstanding post, many underlying principles explained.

Lots more info at your suggested link (above).

-AC_Hacker

Thanks!

This is what made the foundation of the seasonal storage function of my system:

Greener Shelter

The system mentioned above is driven by convecting air, driven merely by the flow created as the density of hotter and colder air starts circulating.

My system is waterborne (hydronic). But note that I do not circulate the same liquid all over the system. Basically, there are three main "blood vessels" in my system:

Heating system, connected to a 100-house central heating gas fired (mandatory subscription) system. This circulates in the 2750 USG main water buffer, as well as into underfloor heating and radiators. It is pressurized as these systems normally are.

External storage system: Also water based, but a separate system. The water pressure is much lower here, just the normal pressure of about a 10 feet water height. This goes into the three 300 USG water tanks, into the underhouse long term seasonal storage, and around the foundation for another reason I think I explained before.

The last one transports the heat between the two circuits mentioned above, and the solar heating panels. I can direct and redirect as I wish, to transfer between either of the two, or all three of them, using several valves. This one is glycol based, as it will freeze at times during the winter.

All three circuits are connected using two heat exchangers. One built into the main water buffer connects the two water based strings, and the external storage string is connected via an external separate heat exchanger.

Check my previous posts, I have posted diagrams earlier - simplified ones.

Here you can see a schematic (system is modified somewhat, but it shows the principle) http://ecorenovator.org/forum/renova...old-house.html

Other relevant posts:

http://ecorenovator.org/forum/solar-...ing-house.html

http://ecorenovator.org/forum/conser...mulator-2.html

How do you know?

Going back to the first post in this thread (see quote below) I found out that a second power plant (about 200km away) has recently started to use a heat accumulator. This one is smaller (38m tall with 21m diameter) with a volume of only 12,000 cubic meters of water, but I found that its temperature range is between 54°C and 98°C, which means that it holds about 2210 GJ of useful heat energy (if my calculations are correct). From the info I could find, the cost of this accumulator was ~$5mln.

EDIT: I just read that the power plant in Cracov also recently built a heat accumulator. This one has 20,000 cubic meters of water.

So, I bought a used heat buffer.
It is 160cm tall and 60cm diameter, giving roughly 500 liters, of which 120 liters is an internal metal tank for domestic hot water. I searched for a plumber who would redo my old heating system to install the tank, and only the seventh(!) one was brave enough to take the job.
https://ecorenovator.org/forum/attac...1&d=1646945682
He hooked it up, we filled it with rainwater, and... it turned out to have a leak in the internal DHW tank. NO!!!!

So I called the guy I purchased it from, he was terribly sorry, he was sure that his workers had pressure-tested it before he sold it. He promised to deliver an almost identical model, but with copper internal tank, which does not have the corrosion/leakage problems. It took many months before he found one, but finally it arrived in February:)
https://ecorenovator.org/forum/attac...1&d=1646945682

Not waiting for the plumber, I tried my luck with the connections and all seem to be good.
https://ecorenovator.org/forum/attac...1&d=1646945682

I put a 10cm layer of mineral wool and plastic wrap as insulation.
https://ecorenovator.org/forum/attac...1&d=1646945682

I filled it with rainwater and no leaks!
https://ecorenovator.org/forum/attac...1&d=1646945682

This week I have been slowly heating it to higher temperatures (I am approaching 60°C) and checking for issues with leaks, pressure, temperature differences between top, middle and bottom, circulation pump settings, etc.
https://ecorenovator.org/forum/attac...1&d=1646945682

I have not yet connected the DHW (still testing solely for household heating), but one problem that I have already identified is that I am losing heat to convection. I have a one-way valve before the boiler, so that loop is OK, but the pipe leaving the tank to the house's radiators is quite warm, and the radiators never completely cool down, even when the circulation pump has been off for hours. I was not expecting this, a plumber once told me that a stopped pump will block convection. Apparently not.

So, what to do with the convection?

1. Add some electromechanism to automatically open the hot valve when the thermostat turns the pump on, and close it otherwise,
2. Redo the plumbing to add a heat trap between the buffer and the radiators
3. Someone proposed adding a one-way valve with a spring after the pump.

I am not sure whether the spring in the one-way valve is strong enough to stop the flow, but maybe I could try both options 2)and 3)? They are more passive than option 1) so less to fail.

Any suggestions?

A simple check valve has a metal disk. The weight of that disk is probably enough to stop convection flow. I suggest just adding a check valve after the pump. I would not use a check valve with a spring because residential hydronic circulating pumps are low head pumps - they create very low pressure. The pressure needed to push the spring open will reduce flow, and could completely stop the flow.

And you are correct. A stopped centrifugal pump will not stop water flow. The water flows right through the pump.

Had the same issue when I built my flat plate Solar Hot Water panels and mounted them on the roof, on a cold night the quick lowering of water temperature in the panels would initiate reverse thermo-syphoning out of the hot water cylinder tank.

Fixed by inserting an all brass water swing valve in the line to the panels, place it slightly angled so the weight of the internal valve flap closes fully when the pump has turned off.

Cheers
Mike

I had a spring-loaded check valve in my spare parts, so I installed it today.
And you were right: The spring is too much for the circulation pump. Even on the highest of the 3 speeds it made strange sounds and hardly moved any water - it barely pulled any hot water out of the tank, after 30-40 minutes the radiators on the ground floor were still cold.:(

So, I took out the check valve, but did add a heat trap (looping up about 20-30cm above the hot water exit), so I'll see how that works. Its placement is not ideal (after the thermal mixing valve, should be before it), but in the summer I might clean that up and add a non-spring check valve if I decide to add the old tank in parallel.

Update.
I read about heat traps and concocted something like this:
https://ecorenovator.org/forum/attac...1&d=1651087611
The heat trap goes up ~35cm, then down ~20cm, through the pump and up to the radiators. The heat trap is above the mixing valve. Could this be the reason why it doesn't work? The heat still siphons up through the pump and to the radiators:(
Here is a simple diagram:
https://ecorenovator.org/forum/attac...1&d=1651087611

During the summer I plan to rework this so that the heat trap is before the mixing valve. Since the low ceiling limits the height if the heat trap goes up above the tank, then can it go down, like this?
https://ecorenovator.org/forum/attac...1&d=1651087611

If this will work, how low should it go?

I made the heat trap which dips almost all the way to the bottom of the tank:
https://ecorenovator.org/forum/attac...1&d=1664128136
Unfortunately, heat is still being sucked gravitationally from the tank! Only the top of the trap should have been hot, further down it should have been cooler.
But after opening the valves, within a few minutes the whole heat trap and mixing valve were hot, with the heat creeping through the circulation pump up into the house:(

I'm thinking about that flap valve, as Mike suggested. But must it be mounted so the flow is horizontal, or can it be vertical (so that the weight of the brass flap will hold back the push of gravity-induced flow, like position #2 below)?
https://ecorenovator.org/forum/attac...2&d=1664128221

Earlier I tried a one-way valve with a disk in it, in position #1, but its spring was too strong and grossly hindered the pump.
After that experience I am afraid that the weight of the brass flap will have a similar effect.

Try putting a heat loop on your return line. If it works, it works.

Swing flapper valve must be operated with the valve disk vertical, eg whole valve body is horizontal, for low pressure pumps.
To ensure the flapper disk seats correctly I mounted mine at a slight upwards angle from horizontal, my magnetic drive DC pump only draws 1 amp @12 volts so is low power, has plenty of flow through the valve.

Mike

Update:
Last weekend I mounted the valve vertically (position #2 a few posts up), so that the weight of the flap holds back the flow when the pump is off. In fact, I even added a few washers to the flap to make it a bit heavier.

It appears that there is enough flow when the pump is on, even when it is at the lowest of 3 speeds. But unfortunately when the pump is off the there is still gravity-induced flow :( It might be less than without either the heat trap nor the flap valve, but the whole plumbing stays warm even 20 hours after the pump goes off.

I wonder how much is it that the hot water wants to escape upwards (which the heat trap by itself should already stop), and how much it is the cold, dense water from the house's radiators pushing into the bottom of the tank and forcing the warm water out of the top?

I dunno what else to try:(:( Adding another small washer or 2 to the valve's flap?

I do not want to go in the direction of electrovalves that open when the pump activates, as I want to keep this low-tech.

Mike, after some reading and thought I believe this would work if I was fighting reverse flow. In my case I have no reason to believe that the flow reverses.
In fact, it keeps flowing even when the pump stops, so I need to eliminate that without hindering flow when the pump is on.

Perhaps your swing valve is not sealing fully when closed, there must be enough getting past it to keep the water moving.

When the sun goes down and its a clear night, the water in the panels can rapidly cool, with cold water moving back and in turn pulling hot water back out of the cylinder.

This summer I reworked the heat trap plumbing on the hot side, extending it first barely above the top of the tank (can't go higher due to low ceiling) then down almost to the cold input:
https://ecorenovator.org/forum/attac...5&d=1697951112
Hoping this shape, plus the weighted flap of the swing valve, would be enough to stop heat escaping when the pump is not working.

Now that heating season has started I see it didn't stop, the pump and plumbing are still warm/hot, even after a few hours :(

Only closing the manual valve next to the pump stops the flow, I am afraid that the only solution here might be a mechanical gizmo to open/close the valve whenever the pump is/not powered. I was hoping to avoid that, as I wanted this as simple and passive as possible.

Makes you wonder how long a heat trap is required, perhaps duplicate what you have and place two next to each other; or add another flapper check valve in the cold line entry into the tank.

Simple solution seems to be getting more complicated. Because you have a powered pump, a simple solenoid valve that is activated whenever the pump is running may be the easiest to implement.

Wondering about the following: In all diagrams of heat traps, the water at the top of the trap is as hot as in the top of the tank, and its temperature goes down as you get lower, so that in the part that starts ascending again the water is cold.
Now, if my pump just turned off, then of course the whole heat trap will be filled with hot water, but it will cool down to the distribution described above.
But what if:
a) The heat trap is insulated exceptionally well, so the cooling is very slow?
b) The hot water in the parts of the heat trap farther out is trying to climb to the rest of the house, pulling more hot water out of the top of the tank. At the same time, cooler (=heavier) water from the house above wants to go down, into the bottom of the tnk, pushing the hot water out of the top.
It looks like once this is set in motion, there will be gravitational circulation for a long time - until the tank cools down. And this indeed is what I have observed - the heat trap only hinders gravitational circulation when it is "cooled off", but not when it is fully warm :(

I read about some solenoid valves, and they all seem to have 1 hour time limit for being powered on, apparently to prevent the electrics from overheating (some have an optional additional heat radiator extending this to 2h). This does not really suit my set up, as in the deep of winter the system may be on for 2-4h at a time.

So I also read about motorized ball valves. These seem to need additional electrical controls, as they do not close by themselves after being powered off.
Also, they take appr. 15 seconds to open, not sure how the pump would handle the additional load and pressure while the valve takes its sweet time to open? Might not be an issue, as it would on be a few seconds.
What would be an issue is when the valve fails to open at all after the pump starts up. That would ruin it, unless I invest in additional protection.

You might check Belimo brand valves, you can get them configured so power on is open and power off close by a spring. Also available with extra benefit of modulation to control flow (that requires control circuit) Very durable and cost effective.
We used them for years on centrifuge systems in machine shop settings in order to prevent over pressure in the feed system. I have a couple open/close to prevent floods if water is sensed on my basement floor.

As a test, what happens if you direct a fan over the pipe rising from the low bend to cool it down? Maybe why it's not working is because the pipe never manages to cool down enough for the trap to work.

Well, technically, I'd be losing more heat then, at least until it cooled enough to stop the flow ;)

In fact, it is insulated pretty well, so a fan wouldn't do much.

I read here that good insulation of the loop/bend is one of the requirements for a trap to do its thing.

I also read there that a similar heat trap on the return might also be needed, so I have prepared all materials (including a second additionally weighted flap valve), just waiting for a day off with warmer weather, to turn off heating for a few hours.

I added the heat trap and flap valve to the return from the house:
https://ecorenovator.org/forum/attac...7&d=1702369586

The additional resistance is evident from the sound/load of the circulation pump. When the pump is in 2nd speed I can hear the flap (with additional weight) bobbing up and down.

Tonight I left the pump off for the whole night to see whether heat is still escaping, and unfortnately it is - supply plumbing was warm to hot (30-40C next ot the pump, basement temp is <10C) :(

So the dual heat traps + weighted flaps add resistance when the pump is on, but do not hinder the loss of heat when it is off.
So much for keeping this system passive.