The Evapotuner: efficient pwater boiler

[Fradow]

After failing to adapt the Saltuner to CSSGs (Saltuners write-up link) and enabling Spaced Out!, it was time to delve into Polluted Water boiling, since Regolith doesn’t fall from the sky and acquiring renewable filtration medium isn’t as easy.

This time, I’m happy to report a success!

While it’s not as good as an improvement over other methods, the Evapotuner is a very competitive build with advantages over other methods.

TL;DR
- For 5kg/s pwater input, use an average of 600W for 30°C (PWV) or 700W for -10°C (CSG).
- Not tested above 5kg/s (there is some margin, but it won’t support 10kg/s).
- No Steel, Plastic or Aluminum is required, though some 125°C+ liquid (like Crude Oil) is required.
- you get a water output of about 10°C (for 30°C input) to 15°C (for -10°C input) above the input pwater temperature. A preheater (built-in) is required to avoid solidifying water with negative pwater temperature.ure.
- Switching to Aluminum radiant pipes brings down the power cost to around 500W for a -10°C input.
- An alternate design (thanks to @Saturnus & @ghkbrew) can reach an average of 400W for 30°C with Steel and Aluminum.
- A version with twice the heat exchanger size and Aluminum pipes can also reach an average of 400W for -10°C. No Steel needed.

Note: all power costs assume a pwater coolant. It's easy to work out the improvement it would make by using Super Coolant or Nuclear Waste, but a bit irrelevant as it doesn't change the overall build and makes it late-game.

Summary of other methods
- Water Sieve + Rock Crusher: requires something to crush (Salt for example, which you get for free if you desalinate a CSSG) and Dupe Labor. The power use for both is around 500W (according to wiki numbers, which assume a 0 operating Dupe). That’s the main measuring stick.
- Water Sieve + another Sand source: other renewable Sand sources require a lot of infrastructure and thus are not really comparable. For example, an Ethanol loop to produce pdirt => Pokeshell => sand, mining a Gilded Asteroid Field for Regolith, or getting Regolith from a Regolith Asteroid.
- Using an external heat source (Core heat/volcanoes…) and Steam Turbines: there is an associated opportunity cost to that. You could run a Petroleum Boiler instead for way more power. Even if you planned to use an external heat source to produce power in the first place, that power plant will be less efficient if you introduce colder pwater, though that can be mitigated with a good counterflow heat exchanger. Note that I consider that a very good solution, but it’s hard to compare.
- Melting dirt: just don’t. It’s going to be terribly inefficient because Sand has a lower SHC than Dirt. Plus you lose half the mass to digging. You are better off boiling pwater and using STs.
- Using a tricked Tepidizer: that’s an exploit that I really hope Klei will fix one day. I personally don’t want to rely on an exploit that could be fixed at any time for my regular survival game.

When comparing the Evapotuner with Water Sieve + Rock Crusher, you essentially pay 100W to 200W more and have a more complicated build to remove the need for Dupe labor and Salt.

Building materials
- All Insulated Tiles are Ceramic in the testing build. It should work nicely with Igneous Rock as well. If you only have a bit of Ceramic, use it in the left part, where there is the most heat differential.
- Radiant Pipes are Copper in the testing build. Iron and Gold will work just as well. Cobalt would slightly improve performance, and Aluminum greatly improve performance.
- the AT can be Gold Amalgam or Steel. Both work to about the same efficiency, I was not able to measure any meaningful difference.
- Auto-Sweeper and Conveyor Loader can be built with +50°C Overheat temperature materials like Gold, Copper or Iron.
- the pool is out of Crude Oil. It could be swapped with another 125°C+ vaporization point liquid, though generally Crude Oil is the earliest you can acquire.

Priming
Priming the Evapotuner is super easy: dial the pre-heater to the max (or use an external pre-heater/already hot pwater) to speed it up, and just launch it. Dial it back down afterwards, to a setting you feel is safe.
It starts from a vacuum or PO2 atmosphere without any issue. On the other hand, I'm pretty sure it won't work if you have a mixed gas atmosphere, and it definitely won't work with other liquids in it.
Because priming works on its own, it should not have any issue with start/stop, though I didn't test it. The only issue is lowered efficiency when getting back to operating temperature, though the trapped water mass should help with that.

Power used
For a constant 5kg/s with -10°C/30°C input pwater (CSG and PWV default temperatures):
- The Aquatuner is at 42%/36% uptime, for about 432W/504W. Unsurprisingly, most of the power is used here.
- The Liquid Pump is at 50% uptime, for about 120W (which confirms the build doesn’t delete mass). Since we are comparing the build to a Water Sieve, which outputs piped water, it’s only fair to count it.
- The cooling disposal Tepidizer is at 6%/5% uptime, for about 57.6W/50W.
- When dialed at -5°C, the pre-heater Tepidizer is at 2% uptime, for about 19.2W for the -10°C input. This is not a very safe setting, I recommend dialing it up a bit higher for survival, so expect to spend a bit more. Obviously, 30°C input doesn’t need pre-heating.

That’s a total of 700.8W for -10°C and 600W for 30°C, very competitive numbers (I didn’t count the Dirt shipping, which accounts for less than 5W).

Note that results will vary with a different input temperature and a different flow rate.

It’s possible to have better numbers by using a wider heat exchanger or better radiant pipe materials, but you are quickly going to see diminishing returns.

Details tidbits
The top-most part needs to be that high because a water droplet tends to form on the lip. It can be one cell lower if a deodorizer is put just left of it to take the PO2 out so that the additional cell to store PO2 is not needed.

The Mesh Tile below the vent is necessary to create an EZ Bead pump, which pushes the Steam up and minimizes heat transfer between the counterflow heat exchanger and the boiling chamber. A Liquid/gas-bypass could have been used too, but there have been reports of it misbehaving on Spaced Out, plus it makes the initialization a tad more complex.

The heat bridge with incoming pwater top right is necessary to force condensation when there is a Steam buildup. A Steam buildup may happen after a while and totally destroy performance because the Steam temperature equalizes over the whole exchanger, which means it’s lower at the left and higher at the right. This means pwater comes in the boiling chamber colder, requiring more heating power, and Steam at the tail end won’t condense on its own anymore, compounding the buildup.

An alternate boiling chamber design
The EZ-bead pump I use to minimize heat exchange with Steam as a flaw: it allows Polluted Oxygen creation on startup. While it’s not a big issue in snaking layers, it could introduce issues with other heat exchanger forms (staircases and waterfalls).
This can be solved with an alternate boiling chamber design created by @Saturnus that doesn’t allow off-gassing:

While it allows other heat exchanger designs, it has 2 flaws: when tested in the same heat exchanger as the Evapotuner, it has a worse efficiency (about 930W vs 700W), and since it runs above 125°C, you have to use Steel.

It’s nonetheless a very good alternative, as it allows to create the best pwater boiler for both the power efficiency and number of pipe sections metrics, at 400W for a 30°C input with Aluminum pipes (tested by @ghkbrew, shown right), which is better than the Evapotuner on both the pipe sections and power usage metrics.
On the other hand, you need to be able to fit it somewhere, and it might perform worse with an irregular flow due to low trapped mass.

Heat exchanger notes
After some tests and theorizing about heat exchangers, my conclusions so far on the different designs are that:
- snaking layers are going to be very good at dealing with an irregular flow and allow using an EZ-bead pump, making it my favorite. Plus I just like the squarish form. On the other hand, they lose on raw efficiency for a regular flow and require more materials and cells used. It also has a lower efficiency when there is water in the Steam chamber (up to 100W loss) that I discovered very late, but probably isn’t worth trying to fix.
- staircases with the alternate boiling chamber are the best for people who want the best numbers (power efficiency and space efficiency) and regulate the flow to make it perform well. And you have to like the shape (I don’t). A super interesting thing about them is that the area used by the Steam auto-scale, which is not the case with other forms.
- waterfalls suffer from low trapped mass and a fixed steam chamber space and just won’t deal well with an irregular flow. @Blah did some tests, and the announced results (600W for -10°C input for the same amount of radiant sections as the Evapotuner) don’t seem worth it to me.
- exchanging heat in the EZ-bead pump isn’t great. After a trial using some state of the art conclusions from @Yobbo, a somewhat tall EZ-bead pump with a thinner but taller heat exchanger that has about the same size as the Evapotuner had a way worse efficiency. It also struggles a lot to initialize. It's not fully explored yet though, there might be some issues with the quick test build I did.

It’s also worth noting that the longer the heat exchanger, the longer it takes to warm up, during which it operates at a lower efficiency. While it’s hard to properly test, I speculate that a longer heat exchanger will have an overall lower efficiency if you dump all your geyser input into it without regulation.

Conclusion
Hit me up if you find something that could significantly improve the efficiency!

Which boiling chamber design and heat exchanger design is in the end a matter of preference. For practical purposes, I recommend my Evapotuner version, as the squarish form makes it easy to fit anywhere, it has a decent efficiency, and can be built without any special materials.

 

[Jann5s]

Amazing write-up, I think I'll build this. What's an EZ-bead pump?

[Fradow]

There is a link in the write-up, but here it is again:

Basically, it uses falling beads to pump gases upward. Here, it pumps the Steam away from the boiling chamber to minimize the time it's in contact with the boiling chamber. Preventing that heat transfer to outgoing Steam improves the efficiency a bit.

[Saturnus]

The good old days of making water by boiling polluted water.

Back when we didn't have refined metal or radiant pipes we measured efficiency by how much throughput we could have with an AT running 100%. This one was at 5500g/s. Quite impressive considering the compact size and obvious limitations.

Then later I discovered that you could trick the tepidizer, not by switching it but by carefully letting water run past it, so that one end was cold while the other end was hot because at the time the tepidizer was limited by the average temperaure across it (unless you resorted to switching).

So I made this "supercharged" one which boiled 10kg/s at 66% AT uptime by super efficient pre-heating with the tepidizer.

If anyone want a trip down memory lane this thread to the second one will take you there. It also links to the previous one.

 

[degr]

pretty far away from something efficient, because of short snake, but at least it use cold water from AT to chill output water, instead of AT vs Tepidizer battle.

 

[tuxii]

What a beautiful creation.

Any idea how much of a performance hit running this is?

Lately, I have tried and reduce piping in my builds simply to help the game run better.  I wish I had a way to quantify the performance impact of one segment of material flowing by pipe or rail.  I'd rather use more power and save frames.

[Fradow]

No clue about game performance. This isn't something that's a concern for builds I do. According to Peter Han, most of the performance hit comes from Dupes intelligence (and he's not the only one), so this build should not be something that greatly impact your game performance.

In fact, if you were previously using Dupes to supply a Water Sieve, it should on the other hand improve your game performance by allowing you to have an inaccessible build that doesn't require any dupe operation.

But feel free to reduce the heat exchanger size, and use the more space efficient staircase design!

[tuxii]

Stating that Duplicant intelligence is bigger hit to performance does not make the performance hit of piping irrelevant, regardless who says it.  I've had dozens of test maps with no/few dupes and it does not take that much active piping to significantly impact performance on my above-average computer on empty test maps.

I think the performance hit of piping is far understated.

[Spoonwood]

9 hours ago, Fradow said:

since Regolith doesn’t fall from the sky and acquiring renewable filtration medium isn’t as easy.

Use chlorine from a chlorine vent to feed dasha saltvines and crush salt endlessly (it's a sand-positive process).  Edit: You said this one.

Another idea lies in deodorizering polluted oxygen from say a polluted oxygen vent, or from offgassing polluted water, vent the excess oxygen to space, use the clay plus excess coal to make ceramic, and then crush the ceramic into sand (that is also a sand positive process).

[Gurgel]

Here is an alternative that seems to be missing:

Aquatuner + Tepidizer as heat source + Steam Turbine as condenser + heat exchanger

It is limited to 2kg/sec with one steam turbine, but you can use multiple ones with the rest the same. Will not be competitive on performance, but is interesting for comparison and dead simple. 

 

Stats (estimated, Super Coolant in the heater, requires Steel):

I currently have that running, with about 4.5C temperature increase for an 42 element AL/Al pipe heat exchanger at 1.6kg/s. Electricity use for an 18C input feed at 1.6kg/s is around 350W, and recovery via turbine is around 240W (warning, WAG), so a total of estimated 110W, input pump not included, output pump not needed. Scaled to 5kg/s that is 340W.

As an estimate without Super Coolant, the consumption goes up in the heater, probably to around 680W, and hence 440W total.  Scaled to 5kg/s that is 1.36kW. Pretty bad in comparison

[Fradow]

It seems I indeed forgot to explain why some solutions are suboptimal (but still valid if you want to have a simpler build or use less space).

First, as far as Super Coolant is concerned: I just didn't bother testing it. It would make the build very late game, and at that point power isn't a concern anyway. Since the absolute best anyone got is 21% AT uptime with a -10°C input, by extrapolating about half that power use would be about 300W for 5kg/s (since you still got the pump and tepidizer fixed costs). You gain a bit more in variants with more AT uptime, for example it saves 250W on the 700W at -10°C variant, bringing it down to 450W.

I did mention using a tricked tepidizer: it's an exploit, I personally won't use it in regular games.

Using a tepidizer pre-heater vs a heat exchanger: while it massively save on size, it also means a hotter output and more power used. As an example, if you were to heat a CSG output to 30°C, you'd save about 100W on the AT side, but that would cost 180W of tepidizer power. It's only a useful move if you really need hot water in the first place (for example for pinchas).

Using Steam Turbines to condense Steam is more power intensive as well, since that means you need more AT uptime to compensate, and when using Polluted Water coolant, they need twice the power you get back more STs. That was already something pointed out during the Saltuners design.

 

As I said in the main post, my opinion is that a great option would be geothermal or volcano based STs with a heat-exchanger before: it's not the optimal use of heat, but it's simple to setup, and permit to convert heat into electricity.

But that's not what I was focused on in this build ^^

[tuxii]

@Fradow

Do you think you could make your CSSG-taming Saltuner work if you froze the CSS into Ice/Brine first before distilling the Brine?  This would reduce the Saltuner's hot side liquid throughput by 77% and Brine has lower SHC as well.

Could that turn your failure into success?

[Fradow]

31 minutes ago, tuxii said:

@Fradow

Do you think you could make your CSSG-taming Saltuner work if you froze the CSS into Ice/Brine first before distilling the Brine?  This would reduce the Saltuner's hot side liquid throughput by 77% and Brine has lower SHC as well.

Could that turn your failure into success?

A CSSG outputs Brine, not Salt Water. Brine and Ice Brine only state change to each other and both have the same SHC, so unfortunately there is no trick to be used there.

That SHC is the problem: it's lower than the resulting elements (Steam/Salt), which means even with a perfect heat exchanger, the heat requirement is still too high, which is why I deemed a CSSG Saltuner a failure. Of course it's possible to make it work by just throwing heat/power at the problem, but that's an inelegant solution that's way worse than the alternative (the Desalinator).

[tuxii]

My mistake.  It was a cool idea, for a moment.

I understand your methodology.