Jump to content

Water plumbing heat experiment


Recommended Posts

3 hours ago, Saturnus said:

It'd be especially hard to make the small vacuums that fills the heavi-watts wire runs-throughs

This is actually pretty easy if you use the fact that Dupes can diagonally construct/deconstruct tiles. Build the wire bridges, and then 2 farm tiles where the heavy-watt wire has to connect them. Surround the farm tiles with the normal tiles, deconstruct the farm tiles. You now have a vacuum inside the little heavy-watt bridge room and all it took was 200 dirt. :) Although the spot where you used the vacuum above the clean water pump to have a heavy-watt bridge does look hard to create.

Overall, I agree this build would definitely take some patience to set up, but now we have a good idea of what is possible when we push the tools we have to their limit. :)

Link to comment
Share on other sites

@Saturnus I tried out your save and let it run for a while, after a cycle or two (morning of cycle 544) the steam stopped condensing and pressure kept building up to > 450 kg with continued running. I turned off the polluted water input but there's so much steam inside the boiler now that it doesn't get below the threshold for condensing anymore. I think your design might have just barely gone over this heat exchanger's throughput limit.

Link to comment
Share on other sites

1 hour ago, Sevio said:

@Saturnus I tried out your save and let it run for a while, after a cycle or two (morning of cycle 544) the steam stopped condensing and pressure kept building up to > 450 kg with continued running. I turned off the polluted water input but there's so much steam inside the boiler now that it doesn't get below the threshold for condensing anymore. I think your design might have just barely gone over this heat exchanger's throughput limit.

Yeah. I noticed it myself. It was late last night. I should have let it run longer. It's probably due to try to run it at 5555g/s for a little too long. Solution is simple. Paint in 5kg 105C in the condenser end and it back to normal. Just don't try to push it above 5400g/s and it should be fine to run. At least it did for 40 odd cycles for me.

Making the condenser a bit bigger might make it run slightly higher.

Stabilized version. (Changed in above post as well). 

First_Law_Take_Two.sav

Link to comment
Share on other sites

2 hours ago, Sevio said:

@Saturnus I think your design might have just barely gone over this heat exchanger's throughput limit.

I'm not so sure. Well, you may be right if you want to keep the compactness. You have too much steam for the first heat exchanger but you can use the second heat exchanger with the cold liquids below. It may reduce the first heat exchanger effectiveness though.

You can also do like in my design and run the aquatuner loop unisolated through the steam but this also reduce the efficiency.

You are near the theoretical maximum. The heat exchanger cannot heat the input above 100°C and the aquatuner thus always needs to add at least 21°C which he cannot do faster than 6.6 kg/s

Link to comment
Share on other sites

55 minutes ago, Cilya said:

You are near the theoretical maximum. The heat exchanger cannot heat the input above 100°C and the aquatuner thus always needs to add at least 21°C which he cannot do faster than 6.6 kg/s

I've seen it spike above 100C input temp when it's run too hard but then as noted the condensing stops and steam builds up. To have a realistic goal the 98-99C it peaks at when running 5400g/s is probably the highest you can run safely. 

Link to comment
Share on other sites

Well, I'm stumped right now. @Saturnus Based on your latest design I've merged yours and mine into a new boiler build in the top left corner, but I can't get this one to run stable even at 5000 g/s. The only change that I've really made to how the heat is handled in the boiler itself is that I increased the size of the heat exchanger by 2 sculptures.

Aquatuner-boiler-mk4-1.thumb.jpg.ac71c9355897cd2670cc9b2daae55eeb.jpg

I was able to cut out one power transformer since if we're running heavy-watt wire the aquatuner won't need a dedicated one (given this save's power network, the third one is still there, just lower down but it does gate the entire build's power now)

I found out through testing that tiles or airlock doors are better at conducting heat between fluids than a thin layer of polluted water, so I separated the two fluids completely in the bottom tank. Another test I conducted showed that with a layer of water above and below, if you have a horizontal temperature gradient airlock doors are better at equalizing the gradient than tiles, so I put airlock doors in the middle of the tank.

Save file attached below.

The First Law.sav

Link to comment
Share on other sites

@Sevio There's a few things. The water lock at the output of the condenser is actually needed or the granite tiles at the end doesn't fill their purpose of transferring heat to the water coming from the tepidizer. I've also found that when you make the tepidizder channel longer then remove the mesh tile and build the sculpture directly in the channel. This last bit isn't in my updated save file but it does help a bit. So replace the tile that forms the airlock, and make the tepidizer channel at least two tiles longer so you can drop the statue into the channel and get rid of the mesh tile. And the heavi-watt wire in the output channel should not be in direct contact with the water tank. The purpose is to help transfer the heat from the water that drips down to the granite tiles at the end of the condenser that's why I run the heavi-watt wire through the tepidizer channel to avoid it coming in contact with the water tank below.

Oh, and I'd probably prefer to leave the separator that automatically filter polluted and clean water should you have over runs.

The granite tiles, the heavi-watt wire, and the statue combines to heat the water from the tepidizer about 0.4-0.5C before it enters the condenser. But you need as much distance from the tepidizer to the inlet as you can get or the tepidizer doesn't run optimally, ie. it shuts down prematurely. At least two tiles from the inlet I think is optimal.

I also thing placing the vent in the bottom right corner is the best. I've played around with it a lot and it seems like the best placement where incoming water runs across at least two aquatuner tiles before boiling in the top left corner.

Link to comment
Share on other sites

2 hours ago, Saturnus said:

And the heavi-watt wire in the output channel should not be in direct contact with the water tank. The purpose is to help transfer the heat from the water that drips down to the granite tiles at the end of the condenser that's why I run the heavi-watt wire through the tepidizer channel to avoid it coming in contact with the water tank below.

I'm not entirely sure I follow, electric wires don't conduct heat to each other, only to the element in the tile they occupy. Is it just to have a reason for/take advantage of a wire bridge between the tepidizer's upward channel and the clean water's drop shaft?

In any case, I did the sculpture change as you recommended, as well as increasing the size of the tepidizer basin. I rerouted the heavi-watt wire through the tepidizer channel, added 2 extra wire bridges on top of that and replaced the tiles between tepidizer up channel and water drop shaft with granite. (they were abyssalite before) The water drop shaft is still water locked even though it doesn't look like it so I added a painting with the space left over for extra steam condensing.

Aquatuner-boiler-mk4-2.thumb.jpg.9109c76014cc712cb8134154fe49d697.jpg

I think I may have found the root cause of the instability though. I had my Aquatuner switch set to above 10 kg, which caused it to evaporate the whole second layer of polluted water. Once that happens, I think it refills first with (colder) water straight above the liquid vent, spreads as a thin film across the basin and destroys a bunch of heat in the tiles to the right of it as per the heat destruction bug with colder water on top of warmer water. I Increased the switch to 80 kg which seems to prevent it from happening.

These changes helped get 5000 g/s just barely stable again, but I still can't seem to get the input water heated consistently to 97 C or above anymore and when I try to increase it to 5200 g/s it destabilizes. Maybe there is such a thing as too large a heat exchanger where the steam doesn't get enough time to heat the polluted water near the pump?

 

The First Law.sav

Link to comment
Share on other sites

@Sevio I'm not sure what issues you have. And the condenser is probably the correct length with ten statues.

You can have a look at mine probably last mod on this build. I've spent way too much time trying to take out or put in tiles to see how it affect airflow in the condenser to get the absolute most out of it. It chucks along happily at 5350g/s feed and 2200g/s loop valves. Raise it a bit further to 5400g/s and it gets unstable but it takes many cycles before it does.

After about 200 cycles my granite tiles at the bottom started giving in so I changed them all for manual airlocks. There's a feed line preheater pipe that raises input temp about 0.6C. Could probably optimize that a bit and save a few watts.

I experimented a lot with tepidizer placement. This is the optimal.

First_Law_Take_Two.sav

2017-09-09 (4).png

Link to comment
Share on other sites

@Saturnus Thanks for posting that save, I think I've managed to make it to 5350 stable now. (as you said 5400 works for a bit but goes unstable)

I matched the way you did your heavi-watt wire above the Aquatuner and moved the tepidizer one space away from the vent. I'm not sure if it's placed optimally since my channel is slightly longer. I'm actually thinking that the horizontal wire bridge you used at the top layer of the Aquatuner basin was what did the trick, it seems to help equalize and preserve the top layer's temperature better after each evaporation cycle.

Link to comment
Share on other sites

@Sevio I think we're just asking too much. With an average feed temperature of about 97.2C the theoretical maximum throughput is 5522g/s. At 5350g/s we're at 96.9% of that. 5400g/s is 97.8%. The difference might seem small but it actually hides that we have to reduce losses but no less than 41% to get that extra 50g/s. Not impossible but probably not worth the time.

The only thing I can think of to get us there would be to encase the boiler in granite tiles, or at least have a bottom plate of it. surrounded by vacuum or abyssalite tiles. The granite might help stabilize the temperature once it's heated up.

Link to comment
Share on other sites

@Saturnus Yeah, you're right that this is about as good as it's going to get. Even 5000 g/s is still a great result for a standalone boiler. :)

Now I'm finally ready to continue or restart a survival game. :)  Building this in survival will be quite the task, I think the hardest part is going to be the vacuum in the water drop shaft and the cooling tank, and keeping CO2 out of the boiler area while building it. It will also need an access airlock for dupes to remove dirt every so often.

Link to comment
Share on other sites

While I was trying to compare the theoretical efficiencies of different design, I started to draw an abstract version of the exchanger you are working on. I think it would be nice to have some kind of symbols to describe parts of a system without going into the details. For instance, abstracting away the details about materials used and tricks to handle liquids and gas and keep only the information about heat production and exchange. An exterior person can understand the general principle without knowing how the heat exchanger are built and then only look on how to build these exchanger efficiently. Here is what I came to for your boiler.

 

boiler.png.0064d3b998fb796e1890a1764adef8c3.png

While it is simplistic, I can compute a maximum potential efficiency (assuming perfect exchangers) of this design from this and compare it to others designs.

Link to comment
Share on other sites

4 hours ago, Sevio said:

@Saturnus Yeah, you're right that this is about as good as it's going to get. Even 5000 g/s is still a great result for a standalone boiler. :)

Appears I was wrong. I had one last stab at it to make 5400g/s. I got it and more. Rock solid 5500g/s. And uses less power too.

It was the original idea about preheating the incoming water that saved power by dumping the incoming water directly into the heat exchanger thereby making the tepidizer do less work as incoming water is heated 9.9C by the heat exchanger. And it was adding 10kg of hydrogen to the feed pump chamber, and some wire bridges to catch the steam flow, heating up the hydrogen that made it possible to raise the feed water temperature by 0.8C to 98C on average.

It can possibly be turned up a bit higher but would definitely need an extra statue in that case as the steam pressure would start to build up if it was run higher than 5500g/s with the present set up. It's also very close to the maximum as feed water temperatures now occasionally goes above 100C.

First_Law_Take_Two.sav

Link to comment
Share on other sites

@Saturnus Very nicely done! Your method of preheating the incoming water inside the cooling tank is very effective, not only heating the input water but cooling the output water as well.

Was polluted oxygen outgassing not a problem for you when you added the hydrogen to the feed pump chamber? I've added 2 kg per tile to the feed pump chamber but immediately it was displaced into a single tile by a few grams of polluted oxygen. Same thing happened for me when I added 10 kg per tile. I've also tried with Methane and Carbon Dioxide but polluted oxygen outgassing still happened. I suppose I could embrace it and just let polluted oxygen build up there forever, but that would make the boiler react ever slower as time went on.

I then had the "brilliant" idea of putting liquid water as a top layer inside the feed pump chamber. Of course the pump sucked it up immediately and put it in the boiler. Oops! But it left a little bit of water in the rightmost tile, so instead of considering it a failure, I  put hydrogen in the left tile and now polluted oxygen outgassing cannot happen! I have no idea how you would build this in survival though...

aquatuner-feed-pump.jpg.60048b677747f9c035de98686b1ac20b.jpg

I've actually been experimenting with Cold mode on my end, for that I moved the thermo switch right next to the tepidizer so I could precisely control the temperature it delivers to the main heat exchanger. This in turn delivers a much more stable temperature to the cooling tank, allowing one to maximize the cooling potential of the boiler with sustained throughput.

tepidizer-finecontrol.jpg.8cf58cbcd141963fe5572b849ea08815.jpg

I've currently got it running stable with 10 kg/s for the feedback valve (can't use feedback when the cooling tank is very cold), 3100 g/s for input valve and 65 C for the tepidizer. It produces cool 10 C water and keeps the area around the tank at about 9-11 C. The ability to run in cold mode is also why I can't send the input water through the cooling tank for preheating.

Edit: Lowest temperature that I would consider safe is about 60 C on the tepidizer and 2800 g/s on the input valve, but that kind of throughput might need some careful ramping up to be stable. The aquatuner's output goes to -15.7 C polluted and clean water exiting the system is -0.2 C right now.

@Cilya Looking forward to seeing what you come up with as theoretical maximum efficiencies!

Link to comment
Share on other sites

I tested an earlier version of this in my current game. The biggest problem I had with it was germs. When the polluted water coming from tepidizer came in contact with the purified water it transferred the germs into the pure water. This could be averted with a heating tank before this to make sure water is fully tepidizer heated. This kills the germs in it.

Now that you are testing the cold version the 65 C limit for the tepidizer wouldnt kill the germs in the polluted water and it would transfer to the pure water. I know the germs will die in pure water, but it takes ages(half time 20 cycles, probably lowered a bit by the 10C). How much can you do in cold mode(20C-25C) with also killing the germs?

Link to comment
Share on other sites

@Jackblac Thanks for asking, I hadn't actually considered that germy polluted water can transfer germs to clean water above it. So I edited in 1 million food poisoning germs per tile into my input polluted water tank (where they should thrive) and let it run for a bit. The germs do indeed spread to the clean water on top of the heat exchanger and get into the cooling tank.

My cooling tank has completely separated spaces for the polluted water cooling medium and the clean water, so at least it doesn't get furter infected there, but the germs do make it all the way into the output water.

My current settings are 59 C for the tepidizer and 2700 g/s for the input valve, which is outputting water near 0 C, so letting the clean water sit for a while would let them die off a bit faster than normal, but that requires switching output tanks. I think it's going to require a complete rework of the boiler to keep the polluted and clean water separated. I only hope it doesn't completely ruin the efficiency if I do that.

Link to comment
Share on other sites

Germs are a real problem when you are thinking about lavatories and showers so thats why the purification should kill the germs during it aswell. I am at cycle 918 atm and promised my dupes they are getting lavatories and showers by cycle 1k.

5 minutes ago, Sevio said:

My current settings are 59 C for the tepidizer and 2700 g/s for the input valve, which is outputting water near 0 C

Right now with a basic pure water circulating purifier setup I can get 2600g/s and 22C water. That is just enough for me to cool my base a bit. I will be testing out a version of this with a seperate preheater (germ safety and more automation imo) and polluted water as the cooling liquid at some point.

Link to comment
Share on other sites

If I'm not mistaken, germs are killed in clean water. So you could let your water rest in a tank while the germs die slowly. Alternating between two water tank could be done with a state machine which is probably not hard to build here. I'll give it a try.

 

edit: Nope ! It doesn't die in clean water.

Link to comment
Share on other sites

@Cilya @Jackblac I'm happy to report that the heat exchanger still works with completely separated flows, so the germs issue is no more. The boiler roof had to be raised up by 1 tile to make this possible. A preview:

Aquatuner-boiler-mk4-3.thumb.jpg.f7aa55889c5edb49263c4b5f08352b29.jpg

I'm currently running it at 2400 g/s while it ramps up, as a lot of heat was lost doing these renovations. I also had to raise up the right wall of the Aquatuner basin by 1 tile to prevent condensed water from flowing back into the aquatuner during startup and destroying tons of heat. This may have been a major flaw in the previous designs too, but not noticed because we ran them at high throughput constantly.

Although the steam still condenses very well, I'm not sure yet if it can reach the same amount of throughput as before in either hot or cold mode, it will take some testing to verify.

Link to comment
Share on other sites

Well... I've a build a three tank, three states machine. There is always

  1. One tank filling
  2. One tank where the water stays still waiting for the germs to disapear
  3. One tank emptying

Once the tank filling is full, the three roles are permuted so we are now emptying the tank which have been left alone. The state is materialized by 9 kg of water sitting in the small enclosed area below tanks : the 9kg are in the are below the tank being filled.

20170910123739_1.thumb.jpg.a1ed9fbc71522d950ad506db32f2be49.jpg

It's cool. But the germs are now dying inside the pipes. So...  The whole thing is not very useful. :)20170910123744_1.thumb.jpg.c9c84a62aa3d6106fb8a69e1e99bfc59.jpg

Link to comment
Share on other sites

3 minutes ago, Sevio said:

@Cilya @Jackblac I'm happy to report that the heat exchanger still works with completely separated flows

I like it better with separate flows. That's what I wanted to try. I also want to pump water/steam from all the steam geyser into it. I'm not concerned about space. I always had a lot of free space in my games. And I started a 512x512 map thanks to Onion Patcher, so I really have a LOT of free space.

Link to comment
Share on other sites

@Sevio The polluted oxygen build up is actually intended. You can see I also added 10kg of that. It continues to rise. The two gases continue to switch places, and although it does affect efficiency the net gain is still positive so I just let the pressure rise. The higher the pressure the better actually as higher mass means more stable temperature. I also tried the water in right tile trick but after a long time all hell broke lose as it turned to steam and sudden you has 3 gasses fighting for space there as a small amount of polluted oxygen would gas off when that happened.

My intended use for this would be as a sewage treatment plant. That's why I inherently don't like the cold mode option. And that's also the reason I want to use directly feeding into the polluted water with a stop valve like it is now, so it can be passively fed by lavs/showers/algae purifiers etc, and I can have an overflow buffer tank before it*. In that case you'd have to separate the input from the output water and and leave out the separator in the bottom right corner but it's an easy modification and doesn't affect efficiency much.

As for efficiency. The aquatuner sets a hard limit. The limiting factor is how high water temperature you can reach before the boiler.

*an overflow buffer is just a pipe ending in a vent. Forward bypass the vent and water will flow to the vent only if the pipe is block. Then you just add a pump to pump the water into the feed line again after the blockage is cleared.

 

Link to comment
Share on other sites

@Cilya Clever, using tiny tanks with water as a hysteresis! Might be a few other applications this technique would be useful for. :)

My heat exchanger was suffering from the problem of not transferring enough heat to the polluted water, so I went all out with granite liquid bridges and gas bridges. That seems to have helped, took a bit of time but it has stabilized now and I can run it at 2800 g/s with a tepidizer temp of 59 C, output water is 1.4 C but will probably go down a bit further after the environment cools down a bit more.

The downside is that if the boiler is run in hot mode, all the pipe bridges will break and will leave a permanent "Building broken (31)" message on your screen. With separated flows, I think the heat exchanger is also too weak to reach the previous throughput in hot mode. I can run it at about 4500 g/s stable now, higher might be possible but I haven't had time to fully test its capabilities.

Link to comment
Share on other sites

I have a very hard time with the setup of a boiler. There are huge heat destruction, and I didn't find why yet. Some steam condensate and fall back to the tepidizer. But it goes down to 90°C and cool the whole thing down. The only things below 100°C are a heavy-watt-wire and a ladder. It's like they are at 90°C so they help steam to condensate and the produced water cools them to 90°C.

I've reduce the heat loss just by removing the ladder. I can't get rid of the heavy-watt wire though :)

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

Please be aware that the content of this thread may be outdated and no longer applicable.

×
  • Create New...