Optimized Steam Engine Setup

[Patric20878]

K, all the kinks worked out on the standard design. 21 boilers and 15 steam engines per row, at 104 blocks long. 5 blocks more compact than Blake's design, at 20.8 tiles per steam engine. Given at least one cold start, will always output 45.9 MW. Pretty satisfied with this design, even if I can't get the battery variant to work, as besides the cold startup requirement, it works pretty cleanly and doesn't introduce any other flaw. Added to original post at the bottom - haven't decided if I want to make this the new setup yet, mainly because I can't get the water battery setup to work with this without dropping output. And actually, I had the standard setup finished like 1.5 days ago, but really wanted to get the water battery variant working too so I could post them both together. But no luck on that.

However, the water battery variant does work if a cold start is given with storage tanks detached, storage tanks are filled, then detached again, then steam engines are connected to power whatever. Then that produces 45.8 MW, connecting the pipes between each 3 offshore pumps actually lowers max output by 0.1 MW. But that's a manual setup, so meh.

I kinda wonder if anyone ever actually builds the water battery variant though, because I sure don't. If anything, coal fields at base are easy to see, and now that I know how to make the coal detector react within 20 seconds or so, coal fields farther from base are easily long enough to give time to setup a new mine. And for very distant mines, train + massive train cargo gives plenty time too. Much rather just build lots of accumulators, which solars eventually need anyways.

[Aru]

I figured out a 3/42/30 which is fully tileable, low tech, and uses strings of 21 boilers.

...

You can compact the engines a bit and use half the poles on them, if you look at my layout on page 14. 4 engines per small pole instead of 2.

Oh... but then, you still have to connect them underneath. But it can't be *more* poles.

[Patric20878]

@Aru: Already done. Doubt it's going to get more compact than this lol. See bottom of original post.

It would've been a 14/15 setup like v1.2 is, but it doesn't compact nearly as neatly as 9/10 does. So no longer compacts as well as it does, but uses 15 engines per row, which is also nice.

[Aru]

Before I started browsing designs on the forums, when I was making my engine layout, I was using the offshore pumps as they are here, adjacent to each other without the gaps. Stretches of straight shoreline are valuable, and it helps take advantage of them. There was a cross pipe connecting some of them (I intuitively suspected there were fluid issues like was being talked about above, where flows impede each other, funny), and they were roughly aligned at the center of the engine array. So, I was using the same layout you see here (my layout I mean), but I also had the pumps compressed better (as in, not always aligned with the engine rows), depending on the shape of the lake shore.

That's why my screenshots stop at the open pipe ends of each row, and don't show the offshore pumps. Because sometimes you should squeeze in as many offshore pumps as you can, or follow an irregular shore line, and then pipe them over to the engines. (You can't pipe all the water from one place to another with a single pipe, of course, it's not big enough. But it gives flexibility for different shore shapes, and for better using smaller shore straights.)

[BlakeMW]

You can compact the engines a bit and use half the poles on them, if you look at my layout on page 14. 4 engines per small pole instead of 2.

The 4 engine per pole layout is pretty similar to the one I use for solar panels (including the row of connecting poles at the bottom), though usually small power pole conservation is not something I care about much because the things almost literally grow on trees and if a build is a little cleaner with more small power poles that seems acceptable.

I do have a great layout for a quad-row of engines using medium power poles:


That is 6 engines per pole. The ends are "ragged" so it's not going to win any beauty contests, but it's very compact and the pole-efficiency is very good, I'm not sure if it even can be beaten in pole efficiency.

[Patric20878]

My old v1.1 design arranged in triple rows using medium poles matches it in pole efficiency, powering 6 engines per pole also. Aru posted the image earlier.

By the way, see if you can figure a working design with water tanks based off the new design I posted at the end of my original post. I tried attaching small pumps right after the offshore pumps, and while it's highly compact, doesn't produce at full power, even with a cold start. Attaching them right after the boilers doesn't work either, and neither does having them both before and after the boilers. Also, that thing you mentioned about water flowing backwards harming power output...was that referring to the water tanks sending water backwards? Like mentioned, I found that disconnecting them recovers output. Tried putting water tanks before the steam engines, along with also trying small pumps right after the tanks...didn't work at all.

I think I need to get a way to disconnect the water tanks after they're full or something.

[BlakeMW]

My old v1.1 design arranged in triple rows using medium poles matches it in pole efficiency, powering 6 engines per pole also. Aru posted the image earlier.

By the way, see if you can figure a working design with water tanks based off the new design I posted at the end of my original post.

It's not possible without small pumps. The only "water battery" designs which work use 1/14. The reason is that storage tanks can draw "limitless" water through a boiler string. The 1.5/21/15 design works because the 15 steam engines regulate (limit) the flow through the 21 boilers based on how much a steam engine consumes per second. But a storage tank has no real limit (it's at least as high as 300/s), this will result in more water being sucked through one boiler string than the other.

You can get something which almost kind of works by placing 25 steam engines (or steam engines + pipes) and then the storage tanks, in this case the sheer length of the pipe system limits the flow to a degree. However the problem is if you allow the system to become empty you get serious problems with the flow being too fast because of the voracious appetite of the steam engines and tanks, that results in serious degradation if the powerplant is a west-coaster or mild degradation if it is an east-coaster. So you have a system which kind of works okay if isn't allowed to become empty, but isn't exactly following any proper ratios. Personally I'd stick with the 1/14 designs as found in the link in my sig, they follow ideal ratios and work great.

Though speaking of those designs, they're pretty old and I've built better designs in my factory (just not better enough to bother updating the guide). I designed a new one using everything I've learned which is optimized for perimeter:


That's for solar/steam, it has 68 steam engines whereas the ideal number is 66.67 - having extra steam engines does no harm. When connected to a load which consumes the entire output, and to the same mega-wattage of solar panels the boilers will run at 100% all day long and the water should be steady at 100 degrees. The main feature difference between my older designs is this one has a direction connection from the boilers to both the steam engines and the storage tanks, using a shared water pipe, this direct connectivity is really helpful for ensuring proper filling of tanks and proper direct access to the boilers for the steam engine, it's effectively an increase in parallelism which is always good in these designs, parallelism is great for overcoming pipe capacity problems - much how in electric circuits parallel resistors have reduced resistance. It also uses the "burn the candle from both ends" principle, both ends can provide approximately equal amounts of water flow when the system is under full load.

Btw notice the sneaky long handed inserter? Yes, there is a stealth 29th boiler! Making it a 2/29/68. The reason for the 29th boiler is because when a 1/14 system is under a lot of suction (like from storage tanks) you need at least 15 boilers to heat the water (it means some of the boilers are underutilized due to the water dwell time being too short). The 29th shared boiler seems to work fine to keep the water at 100 degrees and the long-handed inserter and extra boiler just fit in without making the design any larger.

[Patric20878]

Is there a way to use small pumps to make the 15 steam engine setup work with water tanks? I need to know if there's a way to shut off small pumps once tanks are full, and while combinations can provide for this inverted logic, I can't think of any that can react based on how full a storage tank is, or based on water flow in a pipe.

And if there is no way, I still question if making water tanks is really worth it at all. I originally added it as a feature, but since last year, my consideration for aspects of a factory to optimize also includes ones like time, resources, and others, and I wonder if making all those water tanks is a better idea than making lots of accumulators. What use do massive amount of water tanks still have once the player moves entirely to solar power? If there's no use, it's not an efficient use of time for teching up.

[Qon]

Is there a way to use small pumps to make the 15 steam engine setup work with water tanks? I need to know if there's a way to shut off small pumps once tanks are full, and while combinations can provide for this inverted logic, I can't think of any that can react based on how full a storage tank is, or based on water flow in a pipe.

And if there is no way, I still question if making water tanks is really worth it at all. I originally added it as a feature, but since last year, my consideration for aspects of a factory to optimize also includes ones like time, resources, and others, and I wonder if making all those water tanks is a better idea than making lots of accumulators. What use do massive amount of water tanks still have once the player moves entirely to solar power? If there's no use, it's not an efficient use of time for teching up.

Small pump condition: water < 2500. Just wire them to the tanks. That's it.

But the question is, do you want storage tanks? The reason for accumulators with solar panels is because solar stops outputting energy during the night. But with steam you have energy stored as coal as a perfectly valid energy buffer that can be converted to energy as fast as you need it if you have enough steam engines and boilers. Hot water is just another way to store energy. But with hot water you still need a steam engine setup massive enough to handle converting your hot water to electricity. The only advantage really is that you don't need as many boilers when converting from hot water to electricity. So if you onnly use steam engines to handle spikes in energy requirements then you can use less boilers combined with water tanks to slowly fill up the tanks with hot water ready for instant conversion. But if your boiler setup is big enough to heat water as quickly as steam engines use them then hot water storage is pretty useless. Having storage tanks when you have enough boilers is a bit like having solar panels that work at night too and enough to handle any spike, but still insist on crafting expensive accumulators.

Since boilers is actually lower tech than storage tanks I don't really see any reason someone would have lots of storage tanks but not being able to get enough boilers for your power plant. They are cheap as dirt (literally, they are crafted from stone which is used for almost nothing else). And storage tanks instead of boilers don't give you the ability to have high output all the time.

Storage tanks for hot water would be really useful if we had solar panels that heated water instead of being photovoltaics. Or if we had an electric boiler (a mod does this, but it prevents accumulators from working).

[Patric20878]

I'm inclined to agree. The more I think about it, the less worth it storage tanks seem to be. Just isn't worth using the time making storage tanks instead of accumulators to tech up, and there's no reason you can't just have chests with fast inserters storing coal into chests, then taking them back out when the coal belt is depleted. Since you need a lot of accumulators anyways eventually, it's not really wise to go for compactness with the storage tanks, since then you're trading tech speed for it.

I think I'll probably remove the water battery variant. The standard variant with 15 engines works really well, and it's only flaw with needing to cold startup first is pretty natural to do anyways.

[Aru]

Is there a way to use small pumps to make the 15 steam engine setup work with water tanks? I need to know if there's a way to shut off small pumps once tanks are full, and while combinations can provide for this inverted logic, I can't think of any that can react based on how full a storage tank is, or based on water flow in a pipe.

And if there is no way, I still question if making water tanks is really worth it at all. I originally added it as a feature, but since last year, my consideration for aspects of a factory to optimize also includes ones like time, resources, and others, and I wonder if making all those water tanks is a better idea than making lots of accumulators. What use do massive amount of water tanks still have once the player moves entirely to solar power? If there's no use, it's not an efficient use of time for teching up.

I use small pumps as valves more often than I do as actual pumps for long distance transmission. That's why I suggested they add some kind of valve mechanism in the update thread, one cheaper than pumps, but it didn't look like the devs cared.

[BlakeMW]

And if there is no way, I still question if making water tanks is really worth it at all. I originally added it as a feature, but since last year, my consideration for aspects of a factory to optimize also includes ones like time, resources, and others, and I wonder if making all those water tanks is a better idea than making lots of accumulators. What use do massive amount of water tanks still have once the player moves entirely to solar power? If there's no use, it's not an efficient use of time for teching up.

It's fantastic for saving oil. Both solar and steam are pure iron/copper, while accumulators require a surprising amount of oil. Whether saving oil matters or not depends almost entirely on the map. Sometimes you get 14 oil patches in the starting area, you probably don't care about saving oil. Other times, you get 2 oil patches on a death world. Also there is precious little use for the coal on a map, you need a bit for plastic and can smelt with it if you like, but there will still be a great deal left over. Going full steam munches through coal pretty quickly, quickly enough to be annoying. It lasts much longer (3.33x longer) when using solar/steam. You wouldn't want to use it for a mega factory or a game running 50+ hours unless as a challenge, but for launching a few rockets or having fun it's great.

Is there a way to use small pumps to make the 15 steam engine setup work with water tanks? I need to know if there's a way to shut off small pumps once tanks are full, and while combinations can provide for this inverted logic, I can't think of any that can react based on how full a storage tank is, or based on water flow in a pipe.

Yes, you just use them as flow-control valves. Put 3 in parallel on one of the boiler strings. You don't really need to put any on the other string because pumps are aggressive movers, they tend to aggressively move 1/2 the water down one string, leaving the other half to go down the other string.
There is no need to turn the pumps off because they use no power if they are not pumping anything, the draw is proportional to the liquid actually moved. Or at least, there's no reason I can see to turn them off.

I strongly dislike small pumps in steam power, besides being not handcraftable (I mean the electric engine component) and adding a parasitic drain, you need to give them dedicated power because otherwise under brownout conditions their throughput drops off. I consider a steam engine with small pumps to be a design fail. For hot water storage just use 1/14, for backup steam, just use a circuit network condition on the offshore pumps. Technically you can make steam backup designs which are a little more responsive with small pumps (that is pumping already hot water into steam engines gives faster response by a good fraction of a second), but I'd prefer to save the bother of a dedicated power supply and just have the offshore pump condition trigger a few seconds before the power runs out so the steam engines are primed and ready to respond. The strength of steam is it's cheap and simple (as in quickly handcraftable) and my philosophy is it should be as cheap and simple as possible.

[Aru]

Oh, yes, you can do whatever you want with storage tanks and combinators, especially for fine control of cracking, like I just described in the other post viewtopic.php?f=8&t=8854&start=130#p164184

I use the "storage tank" signal to set the tank count, and the red and green signals for upper and lower cracking thresholds, and all three on the same constant combinator. So, I could configure it remotely if I wanted, and could also view fluid levels remotely, via the network display on the right. (My whole oil processing area is, incidentally, remote in my current game, not right next to the main base. I have all four pipe lines running adjacent. It's easier to remote oil processing than anything else, because pipes have good throughput, and very good velocity, and are inexpensive.)

Reminds me of those hand-held heavy duty machine controls, with the thick cable, that you see someone with a hard hat in a factory grabbing to move something around.

Now that I think about it, not only should there be a controllable fluid valve (that doesn't require an electric engine), but also a controllable gate on the conveyor belts, or a special belt tile that you can turn on and off. Such things are realistically simpler and cheaper than switchable huge fluid pumps (ironically named "small pump") and arrays of robotic arms. It doesn't make sense, that the only real way to control item movement is with an array of smart inserters continuously transferring from one belt to another, when simply turning a small piece of conveyor belt off and back on would accomplish the same thing. (But more cheaply, compactly and easily.)

And why are engines materials for electric engines... that makes no sense at all. They are completely different. Electric engines are smaller, denser, usually cheaper, higher torque, have less (or no) need of gears.

[Patric20878]

Put 3 in parallel on one of the boiler strings.

What do you mean by that?

I strongly dislike small pumps in steam power, besides being not handcraftable (I mean the electric engine component) and adding a parasitic drain, you need to give them dedicated power because otherwise under brownout conditions their throughput drops off. I consider a steam engine with small pumps to be a design fail. For hot water storage just use 1/14, for backup steam, just use a circuit network condition on the offshore pumps. Technically you can make steam backup designs which are a little more responsive with small pumps (that is pumping already hot water into steam engines gives faster response by a good fraction of a second), but I'd prefer to save the bother of a dedicated power supply and just have the offshore pump condition trigger a few seconds before the power runs out so the steam engines are primed and ready to respond. The strength of steam is it's cheap and simple (as in quickly handcraftable) and my philosophy is it should be as cheap and simple as possible.

Wait, are you saying you can shut offshore pumps off with circuitry? Is this new in 0.12? How does that work? Back when my water battery variant was originally made, I don't think this was possible.

Also, I played around with attaching two way valves between the steam engines and storage tanks. Unfortunately though, just adding pipes themselves cause power drops and there's no way to attach small pumps directly after the steam engines since that'd be one way only. And then the only way to have it go back into the steam engines would be routing long ugly pipes back to the beginning of the engines, not to mention one small pump per engine row would be way too slow.

[BlakeMW]

Wait, are you saying you can shut offshore pumps off with circuitry? Is this new in 0.12? How does that work? Back when my water battery variant was originally made, I don't think this was possible.

Apparently it was introduced in 0.12.0 at the same time as storage tank and small pump connectivity. The thing is that it's quite unintuitive, the offshore pump doesn't use electricity so you tend to think of it something which is always passively on, right? That's what I thought until I learned somewhere you can give them circuit network conditions. In a way it's not that surprising because other parts of the circuit network also do not require electricity, chests and storage tanks and I think constant combinators do not require electricity to provide a signal so the idea that there are unpowered components which can use a signal is not that surprising.

Also, I played around with attaching two way valves between the steam engines and storage tanks. Unfortunately though, just adding pipes themselves cause power drops and there's no way to attach small pumps directly after the steam engines since that'd be one way only. And then the only way to have it go back into the steam engines would be routing long ugly pipes back to the beginning of the engines, not to mention one small pump per engine row would be way too slow.

I am not sure what you're attempting there? Why would you want a two-way valve? Flow is generally one-way in steam engine setups. Small pumps are best placed directly after the boilers. To maintain throughput there are 3 approaches, the first is parallelism, the second is "burn the candle from both ends" where you have storage tanks at both, the third is partitioning. Say you're using 21 boilers, that means 3 small pumps worth of water. So you place 3 small pumps after the boilers for flow-limiting. Now take the steam engines - lets say there are 21 for argument's sake. So each 7 steam engines will consume 1/3rd the water. So after the 7th steam engines you can put 2 small pumps, and after the 14th steam engine you place 1 small pump. Now each partition of 7 steam engines gets exactly 1/3rd the water. You can add storage tanks as you please (probably best at the end of the steam engine strings). I have built serial steam setups using this principle, where you just use fewer small pumps between each partition to ensure even filling/consumption of water. It actually works really well, but is strictly inferior to parallelism owing to the use of small pumps, unless you really want long designs. Also resistance becomes a problem, the small pumps help especially when they suck directly from a tank outlet, but storage tanks just have innately high flow resistance, so for designs above a certain length you need to short-circuit the storage tanks so that flow that is filling the system can completely bypass the storage tanks. Arguably short-circuiting the storage tanks is even more effective than using small pumps - an example I recently built was a 200MW hot water storage on the reddit community game map:

It uses partitioning with tank short-circuiting and no small pumps. The design is not optimized for anything in particular, but fulfills its function of providing oodles of electricity. It is an extensible design because it starts as a conventional 16/224/160 and then additional blocks of 20 tanks and 80 steam engines can be added on the end basically indefinitely, flow isn't a major limit, because of the short-circuit you only really need to care about the total length from the offshore pumps to the final steam engine. According to DaveMcW's flow post the limit for 60 fluid/s is over 300 pipe segments while this design has only about 50 segments, Still, small pumps can assist a lot in *completely* filling tanks because of the tendency for the system to equalize itself rather than systematically filling all the tanks, but using more tanks and only having them 80% full is a fair substitute for using small pumps to ensure 100% filling.

[Patric20878]

Refer to the water battery variant on my original post. The storage tanks are filled from the left side. Since circuitry can control offshore pumps now, it no longer needs the small pumps for cold water flow prevention, but the storage tanks still do, for pressure control. The condition I need is that when there's a steady supply of coal, only the pumps that push water towards the right into the storage tanks are active, since in a 15 engine setup, without them the tanks only fill about halfway due to lack of pressure. When the tanks are full, those pumps should be shut off, to remove the storage tanks as a valid destination (since adding any length of tube seems to drop output). Then, if coal runs out, a second set of small pumps that pump water left into the steam engines should be activated, and the first set that pumps water right is deactivated. In this way, water flow between the steam engines and water tanks only ever flows one way at any given time. But like said, just adding pipes drop output.

It's a bit hard to digest what you're describing in text form, but how would you add 4 tanks to the ENDS of each 15-engine row on MY design (towards end of original post)? I need to preserve compactness here so changes must have a minimal impact on compactness. And I specifically want to be able to add tanks to the end (or beginning if possible), as I want to be able to upgrade the standard setup to the water battery variant with minimal manual reconstruction.

[BlakeMW]

Personally I'd do something like this:



Use 3 tanks per 10-15 extra steam engines. Without the short-circuit pipe performance is horrible, with it, you should get tanks to about 80% full before the system starts to slack off, with small pumps slave-driving the boilers it'll get close to 100% before starting to slack off. I just have to say, that these short circuit pipes are amazing for performance. Tanks with a direct or near-direct connection to the pipe network work so much better than those stuck behind other tanks. It's really quite un-intuitive but for whatever reason flow between tanks is just really really slow. What happens is that in a normal pipe system there is a performance optimization which could be described as "flow momentum", future water movement is accelerated along the path previous water has flowed (I finally figured out a setup that conclusively proves this - I'll make a thread later once I figure a good way to screenshot it), but this optimization isn't applied to flow through tanks, so whenever a tank appears in series it breaks the flow momentum dramatically lowering the throughput of liquid. Several tanks in series reduce the flow to a trickle. The short-circuit pipes let the flow momentum optimization do its magic and liquid can speed through to the end of the system, nearly evenly filling all tanks along the path.

The condition I need is that when there's a steady supply of coal,

Okay I don't really get the point of the coal detector, steam engines consume cold water pretty quickly just due to how the performance thing works (okay I do get the point, but not the usefulness unless you plan to be running out of coal all the time). But when I want to do empty belt detection I use a smart chest:

The output inserter sideloads so that when the lane is full the inserters become inactive. It can also double as a buffer. So even though I don't really get the point, it must be said it's pretty darn cheap and easy to add one of those and link it to the offshore pumps.

[Patric20878]

The coal detector is probably another of those things that are outdated. If your chest system can detect the lack of coal and shut off offshore pumps, welp, it probably didn't exist back in 0.10, heh. Good to see there's no longer any need for this workaround solution. I update this design faithfully but I quit playing Factorio a year ago, so I have no clue about all the new features heh.

So I still need small pumps after the boilers on my design if I don't want any power output loss? I want the full 45.8-45.9 MW or so output with 90 steam engines on my extended design, either right away or eventually. Or can I just attach the tanks in the way you showed? I don't care for how fast the tanks fill up as long as power output doesn't drop permanently. At full consumption, it'd fill far slower than the liquid transfer bottleneck between the tanks anyways.

And as for the usefulness of the coal detector, automation, of course. Why manually check your coal field when you can just build the simple detector you showed? Having one offers a degree of security for such a critical part of your base, at least until you move onto solar power as the primary source.

[OdinYggd]

Here's how I set mine up. Each row is a complete 1:14:10 arrangement, containing 14 boilers, 1 accumulator, and 2 pumps each driving 5 engines.

I use a dedicated 'pilot engine' to power these pumps, which is itself controlled by a variation of the liquid op-amp posted in another thread.
For fuel, a bit of clever wiring allows the boiler to use either fuel blocks or coal depending on availability. Boxes of fuel with inserters to load and unload them are used as a buffer between the factory belting and the power station, that way if the mine dries up or I have a surge load elsewhere my power station continues operating.



It is possible to make this same layout using wooden poles instead of the substations, but I found it to be rather cluttered and reworked it to use the substations as soon as I had access to them. What you see here are 2 assemblies out of my total of 10, they are identical and blueprinted with few exceptions.


This is the pilot assembly. The pump that my mouse has highlighted is powered by an isolated grid using a solar panel and accumulator to maintain an always-on primary supply for the feed pump. Said pump draws steam from the tankr in the top right corner, allowing my pilot to share boilers with the #1 engine bank in the array.

The 3 pumps to the right of the pilot engine and left of my character then remove steam from the line feeding the pilot engine, forcing the pilot engine into cutoff from lack of steam. These discharge into the tank right below them, and are controlled by the monitoring grid. The tank is emptied back into the line for bank #1 by another trio of pumps. I found I had to add the tank and additional pump stages to avoid issues where the source tank was too full to return steam to causing the pilot engine to run constantly. Like this, the pilot tank empties in the morning when the main tank hasn't yet refilled all the way but there is enough power in the system to run the pumps.

For control, the monitoring grid has 4 solar panels and that one accumulator on it, while powering a total of 6 pumps- 3 to hold the pilot in cutoff, the other 3 to empty the pilot tank back into bank #1 when there is room. Like so at sunset the steam engines remain off leaving the system on accumulator power until the pilot accumulator is empty and the cutoff pumps stop. This causes the pilot to come on right away, sending the entire ~50MW main array to full throttle instantly instead of having to wait for boilers to heat up as done by the traditional method of turning off the stokers. Once the sun rises, the monitoring grid solar panels start to power the cutoff pumps again, slowing the pilot. Usually it surges a little as the pumps start and stop by available power in the monitoring grid, I am still working on a solution to that. But eventually the solar panels on that grid make enough power to force the pilot to stop- in turn stopping the main grid. This then leaves the accumulators only partly full so that they are charged up by solar power during the day, and the steam power just fills the deficit during the night as well as cutting in during surge loads such as turret creeps.

[Aru]

Is there a way to use small pumps to make the 15 steam engine setup work with water tanks? I need to know if there's a way to shut off small pumps once tanks are full, and while combinations can provide for this inverted logic, I can't think of any that can react based on how full a storage tank is, or based on water flow in a pipe.

And if there is no way, I still question if making water tanks is really worth it at all. I originally added it as a feature, but since last year, my consideration for aspects of a factory to optimize also includes ones like time, resources, and others, and I wonder if making all those water tanks is a better idea than making lots of accumulators. What use do massive amount of water tanks still have once the player moves entirely to solar power? If there's no use, it's not an efficient use of time for teching up.

I never considered water tanks to be useful. Boilers are cheap, there's no major logistical challenge in connecting them to engines, they're small, reliable, so it seems like one of the more inappropriate places in factorio for some kind of buffer (that is, between boilers and steam engines). A chest with coal would be better.

Seems like, if you were to put a buffer chest for copper wire, between your wire assembler and your green circuit assembler. You wouldn't do that. It's easier to consider them as a single unit, with nothing between but a single inserter, and it's more worthwhile to store the copper plates further 'upstream', because they are denser and have more uses than the wire.