Niggles and nitpicking - player preferences

[Honktown]

I like to take my starting boiler banks and swap out the engines for turbines, and then use it as an emergency brownout supplemental power source. By using strictly burner inserters in the setup it can run on zero power and has zero drain.

Steam-generator output is dependent on steam temperature and fluid consumption alone. Turbines unfortunately can't output more energy than steam engines when boiler steam is input. Vice-versa, steam engines waste the extra energy from nuclear steam, as they're only permitted to use 165-15 degrees of steam energy, compared to the 500-15 of turbines.

I put eff1 modules in almost everything until it's ready for prod3 + beacons. Miners to reduce pollution in addition to electrical burden. Anything over 90kw gets 2, and over 120kw or so gets 3, since they're cheaper than a solar panel.

I sphaget until it's time for prod3 + beacons.

I space my miners so that patches last longer and then colonize more patches.

Interesting on the efficiency modules. Never considered it. I know solar panels are expensive and have poor power density, and can't even be used at night, but never thought about the cost compared to efficiency modules. 27.5 copper, 15 iron, 5 steel (25 more plates) vs 32.5 copper, 15 iron, and 10 plastic bars (who cares about coal). That's 67.5 plates for 42 kW over the day, vs 47.5 for 30% (26.66% average for 3 modules). Took a few tries for the math so hopefully it was done right, but I think if a machine uses more than 98 kW, it's more efficient to use 1-2 modules, not even accounting for accumulators, and even with three modules, if the machine uses over 110 kW it saves more power per ore. Great observation! Fewer finished products are required for science now (gun turrets and drills *cough*) but this is still very useful, especially in biter-heavy games (significantly fewer resources, less space, and even less pollution/item). I want to take this the nearly full way, comparing production chain costs (miners, furnaces, refineries, chem plants, assemblers) and maybe pollution but can't muster the focus at the moment.

[foamy]

I like to take my starting boiler banks and swap out the engines for turbines, and then use it as an emergency brownout supplemental power source. By using strictly burner inserters in the setup it can run on zero power and has zero drain.

Steam-generator output is dependent on steam temperature and fluid consumption alone. Turbines unfortunately can't output more energy than steam engines when boiler steam is input.

Yes, I know. Swapping in turbines doubles the power output without increasing the footprint beyond the steam tanks needed to supply them. So, when the plant is running primary power, it's plain engines, but when I swap in turbines, it can produce twice what the boilers can supply. If it's now running as a backup this is no problem, since the boilers can run continuously and charge steam tanks, which can then discharge and allow for the full output.

Turbines and engines also have slightly different behaviours in that their loading is based on their assumed maximum power generation, i.e. 5.8 MW for turbines and 900 kW for engines, regardless of the steam actually supplied. This means a backup plant running turbines has a higher priority than a nuclear plant, which is suboptimal but acceptable for the footprint convenience since, in a backup role, the nuclear plant'll be running at full anyway.

(The key advantage of panels over modules is that panels can reduce your pollution and resource drain from power production to zero, and also have no UPS penalty for bases where that's relevant. The best combo for pure ecofriendliness in a relatively compact space is probably using both at once, which would get you down to a panel and a half per AM2. Plus an accumulator farm somewhere, but accumulators are smaller than panels.)

[Honktown]

Yes, I know. Swapping in turbines doubles the power output without increasing the footprint beyond the steam tanks needed to supply them. So, when the plant is running primary power, it's plain engines, but when I swap in turbines, it can produce twice what the boilers can supply. If it's now running as a backup this is no problem, since the boilers can run continuously and charge steam tanks, which can then discharge and allow for the full output.

Turbines and engines also have slightly different behaviours in that their loading is based on their assumed maximum power generation, i.e. 5.8 MW for turbines and 900 kW for engines, regardless of the steam actually supplied. This means a backup plant running turbines has a higher priority than a nuclear plant, which is suboptimal but acceptable for the footprint convenience since, in a backup role, the nuclear plant'll be running at full anyway.

(The key advantage of panels over modules is that panels can reduce your pollution and resource drain from power production to zero, and also have no UPS penalty for bases where that's relevant. The best combo for pure ecofriendliness in a relatively compact space is probably using both at once, which would get you down to a panel and a half per AM2. Plus an accumulator farm somewhere, but accumulators are smaller than panels.)

If I'm understanding the turbines correctly, you feed in nuclear steam when you can produce enough (i.e. had extra in tanks), but then add boiler steam if it's insufficient? In my experience steam mixes and temperature is averaged, so it doesn't interact like two different fluids. No pumps would be needed if the math evens out.

The pollution/cost-per-energy drain of solar panels is obvious. UPS is for megabasers. Comparing the cost with efficiency modules is not obvious, and if you're a productivity/overbuilding/just-spam-the-biters-with-more-damage kind of person, I don't think it'd even come up.

[foamy]

If I'm understanding the turbines correctly, you feed in nuclear steam when you can produce enough (i.e. had extra in tanks), but then add boiler steam if it's insufficient? In my experience steam mixes and temperature is averaged, so it doesn't interact like two different fluids. No pumps would be needed if the math evens out.

No: they're two totally separate plants. The nuclear one just goes and operates at whatever load it needs to, with accumulators to cover limited surges beyond that, but if there's a brownout situation, a switch closes and a boiler plant -- usually, my old primary plant, but with turbines swapped for engines and buffer tanks added to allow for the doubled draw -- will start feeding into the network as well. 20x6x2 turbines = 432 MW, which is a nice piece of additional generation for middling factories.

[Honktown]

If I'm understanding the turbines correctly, you feed in nuclear steam when you can produce enough (i.e. had extra in tanks), but then add boiler steam if it's insufficient? In my experience steam mixes and temperature is averaged, so it doesn't interact like two different fluids. No pumps would be needed if the math evens out.

No: they're two totally separate plants. The nuclear one just goes and operates at whatever load it needs to, with accumulators to cover limited surges beyond that, but if there's a brownout situation, a switch closes and a boiler plant -- usually, my old primary plant, but with turbines swapped for engines and buffer tanks added to allow for the doubled draw -- will start feeding into the network as well. 20x6x2 turbines = 432 MW, which is a nice piece of additional generation for middling factories.

Either there's a description issue or you're making a mistake. A turbine has a maximum generation from 60 units of steam per second, period. That is what the output is based on. If you're using boiler steam, you can only generate 1800 kW/sec of power if you're consuming 60 fluid per second.

[foamy]

Either there's a description issue or you're making a mistake. A turbine has a maximum generation from 60 units of steam per second, period. That is what the output is based on. If you're using boiler steam, you can only generate 1800 kW/sec of power if you're consuming 60 fluid per second.

That's correct (aside from the fact that kW already includes the /sec you added, by definition :p). However, steam engines can consume a maximum of 30 units/sec.

Therefore, if you replace all your engines with turbines, they can provide twice the power for the same number of entities, but draw twice as much steam. Therefore you need twice as many boilers if you want continuous output at that level, but for a backup system you don't need continuous output; you can stuff tanks on the end of the turbine rows, and so what happens is the tanks supply the extra steam per second needed to drive the turbines.

Each offshore pump can sustain 36 MW of continuous production. I have six of them. So the maximum sustained production is 216 MW (which I've tested), and that doesn't change if you use turbines or not. But on engines, that's also your peak power. If you replace them with turbines (equivalent for this purpose to just building twice as many engines, although it messes with how load gets allocated as I noted earlier), you can double your peak power output without needing to expand the generator array, to the 432 MW figure I quoted in my prior post. Each row of six turbines draws 360 steam/sec at maximum, so you get an additional 140 seconds of peak production per tank you add to the row (180 steam/sec is supplied by the boilers, 180 by the tank). So even two tanks gets you almost five minutes of peak production before it collapses down to the sustained number. That's quite a while.

[Honktown]

Either there's a description issue or you're making a mistake. A turbine has a maximum generation from 60 units of steam per second, period. That is what the output is based on. If you're using boiler steam, you can only generate 1800 kW/sec of power if you're consuming 60 fluid per second.

That's correct (aside from the fact that kW already includes the /sec you added, by definition :p). However, steam engines can consume a maximum of 30 units/sec.

Therefore, if you replace all your engines with turbines, they can provide twice the power for the same number of entities, but draw twice as much steam. Therefore you need twice as many boilers if you want continuous output at that level, but for a backup system you don't need continuous output; you can stuff tanks on the end of the turbine rows, and so what happens is the tanks supply the extra steam per second needed to drive the turbines.

Each offshore pump can sustain 36 MW of continuous production. I have six of them. So the maximum sustained production is 216 MW (which I've tested), and that doesn't change if you use turbines or not. But on engines, that's also your peak power. If you replace them with turbines (equivalent for this purpose to just building twice as many engines, although it messes with how load gets allocated as I noted earlier), you can double your peak power output without needing to expand the generator array, to the 432 MW figure I quoted in my prior post. Each row of six turbines draws 360 steam/sec at maximum, so you get an additional 140 seconds of peak production per tank you add to the row (180 steam/sec is supplied by the boilers, 180 by the tank). So even two tanks gets you almost five minutes of peak production before it collapses down to the sustained number. That's quite a while.

Turbines cost way more than steam-engines, and you can just as easily put tanks on the nuclear reactor, and only let them discharge via a pump. A single pump can provide 116 MW of nuclear steam back into the system during brown-out. No pumps are needed at all if you add extra turbines/tanks to the reactor.

If you're using boiler steam, energy per fluid is the same regardless of consumer, but for nuclear steam, you're storing/supplying roughly 3.23x the amount of energy per unit of steam.

Only time I could see separate boiler/turbines being useful is if you're exactly on the edge of the nuclear reactor, so close that you can't recharge backup steam in time. At that point you could still attach boilers to the opposite end of the nuclear turbines or tanks (as long as there are extra turbines to overpower). The steam would mix into the turbines, which can't all possibly be full because you've ran low on steam.

[foamy]

Turbines cost way more than steam-engines, and you can just as easily put tanks on the nuclear reactor, and only let them discharge via a pump. A single pump can provide 116 MW of nuclear steam back into the system during brown-out. No pumps are needed at all if you add extra turbines/tanks to the reactor.

By the time you're needing any of these kinds of assemblies the cost of turbines over engines is pretty minor all told. Turbine costs 170 plates vs. 2x steam engine's 62. By comparison, one L3 module is 3080 plates, plus a not-insignificant amount of plastic.

As for putting tanks on a reactor array, it's doable, yes. I've considered just adding some tanks and turbines to the end of a turbine row, and then toggling the power connection out to the main grid on and off (no need for additional pumps) based on accumulator behaviour. The drawback is that nuclear plants, especially of the tileable variety, are already very large across -- nearly 300 tiles for a 2xN tileable! -- and that they share a single common point of failure in fuel production. Which generally shouldn't be a problem, but a completely separate plant avoids that and can be used to ensure you have power to the uranium chain.



Also, I already have the dang thing anyway from when it's the primary plant. :p

[Honktown]

By the time you're needing any of these kinds of assemblies the cost of turbines over engines is pretty minor all told. Turbine costs 170 plates vs. 2x steam engine's 62. By comparison, one L3 module is 3080 plates, plus a not-insignificant amount of plastic.

As for putting tanks on a reactor array, it's doable, yes. I've considered just adding some tanks and turbines to the end of a turbine row, and then toggling the power connection out to the main grid on and off (no need for additional pumps) based on accumulator behaviour. The drawback is that nuclear plants, especially of the tileable variety, are already very large across -- nearly 300 tiles for a 2xN tileable! -- and that they share a single common point of failure in fuel production. Which generally shouldn't be a problem, but a completely separate plant avoids that and can be used to ensure you have power to the uranium chain.



Also, I already have the dang thing anyway from when it's the primary plant. :p

My biggest game has two separate nuclear plants because I don't copy blueprints and didn't want to mess things up while I added more reactors and stuff. After 2x3 the average return per additional reactors is poor and gets poorer, so I only imagine by culture or by "must be infinitely tileable", people don't usually make square-ish reactors. Two 2x3s vs a 2x6 is a difference of 160 megawatts. A 10% increase (1600 vs 1760 MW) from 2x3 for the significant increase in complexity and design limitations. An unchainable 2x4 or 2x5 would have even less of a difference.

Only one game could I have kept my boilers around after I switched to nukes (340 steam engines+tank accumulators). My limitation wasn't resources, just the sheer amount of peak power and quickness I needed for killing dudes (was switching from flame turrets to laser in a .16 game with completely maxed-out enemies). Every game I've stopped building my previous power system, and quite rapidly it's usefulness as a back-up is dwarfed by the capacity of my newer system and secondary savings (pollution, space, ease of construction, fuel). The biggest solar game I had at least 10k solar panels, which sounds like a lot... but it's an average of 420 MW, when I could just slap down a 2x3 reactor and have 800 MW 24/7, and even a 2x2 is 480 MW continuous. Don't have to build/put down accumulators either.

As for modules, you don't need modules anywhere in the uranium chain. The up-front cost of kovarex is through the roof (1500 science, phew, but it costs much less than it used to).

Contrasting, a single reactor provides the energy of 22 boilers (and no pollution...). Two reactors is worth 88. A fuel cell is worth at least 2000 coal. Two fuel cells in a box is 67% more energy than a steel chest filled with coal, without neighbor bonuses (6 cells for fuel blocks). Four tanks can hold more than a single unbonused fuel cell worth of energy and is "instantly" deliverable. That's not accounting for the energy cost of producing thousands of coal/fuel blocks. In the flamethrower->laser game I fed nuclear steam to steam engines for a little while because yeah, I could afford the energy loss in exchange for additional power and nuke testing at that moment (you don't need to mine/logistic fuel to a heat exchanger - saves a lot of time and is impossible if you don't have enough coal/fuel block production/space).

Space/laziness is the only reason I can imagine replacing steam engines with turbines, as it's super inefficient across the board. I WILL NITPICK UNTIL THE END.

Edit: Also learned a burner arm inserting wood is not fast enough for a boiler. Neither is a basic inserter. Two burner arms work, but chug power.

[Cribbit]

My biggest game has two separate nuclear plants because I don't copy blueprints and didn't want to mess things up while I added more reactors and stuff. After 2x3 the average return per additional reactors is poor and gets poorer, so I only imagine by culture or by "must be infinitely tileable", people don't usually make square-ish reactors. Two 2x3s vs a 2x6 is a difference of 160 megawatts. A 10% increase (1600 vs 1760 MW) from 2x3 for the significant increase in complexity and design limitations. An unchainable 2x4 or 2x5 would have even less of a difference.

I've never found tileable nuclear to be overly complex. The repeatedness is almost simpler since you end up just replicating the same few rows of tiles, vs a "square" reactor where making the initial blueprint means figuring out all the surroundings. Of course you could use someone else's blueprint, but still.

Then again, bots and blueprints remove any complexity of repeating a construction.

[Honktown]

My biggest game has two separate nuclear plants because I don't copy blueprints and didn't want to mess things up while I added more reactors and stuff. After 2x3 the average return per additional reactors is poor and gets poorer, so I only imagine by culture or by "must be infinitely tileable", people don't usually make square-ish reactors. Two 2x3s vs a 2x6 is a difference of 160 megawatts. A 10% increase (1600 vs 1760 MW) from 2x3 for the significant increase in complexity and design limitations. An unchainable 2x4 or 2x5 would have even less of a difference.

I've never found tileable nuclear to be overly complex. The repeatedness is almost simpler since you end up just replicating the same few rows of tiles, vs a "square" reactor where making the initial blueprint means figuring out all the surroundings. Of course you could use someone else's blueprint, but still.

Then again, bots and blueprints remove any complexity of repeating a construction.

A tileable/blueprintable reactor means finding the surroundings too, related to the 300 wide tile comment. I've seen some reactors normally built on lakes so the water pumps go to the heat exchangers with no pipes, but I don't know if you can auto-landfill under everything else.

[Cribbit]

A tileable/blueprintable reactor means finding the surroundings too, related to the 300 wide tile comment. I've seen some reactors normally built on lakes so the water pumps go to the heat exchangers with no pipes, but I don't know if you can auto-landfill under everything else.

Non-lake reactors are a huge pain the butt. Default map gen has plenty of massive lakes though, I can make a 2x40 on the nearest lake in 90% of maps.

You're right, it does take a little bit of effort to set the pumps before landfilling, but I would still put it on a lake even if it was a 2x6 square unit, and clearing cliffs/trees + finding nearby water to pump is IMO worse for a land-based reactor.

[foamy]

Fun fact: I was doing some math on what it would take to guarantee a refinery can always output, no matter what the balance in the draw between lubricant, petroleum, and light oil is. I came up with some interesting numbers, centered around a 5:2:6 prod/12 beacon layout.

Pure petroleum draws are easily accomodated through cracking with only a very minor excess accumulation of light oil, with the excess burned. Pure light oil draws require making roughly 1826.5 solid fuel a minute made from the excess petroleum gas. This can be fed into the rocket fuel assembly line, which is what draws the light oil anyway.

That leaves lubricant. Since lubricant can only be made from heavy oil, in order to guarantee heavy oil can be produced, you need to get rid of all the light oil and petroleum gas that ordinarily comes with it. That means you need to generate 1826.5 solid fuel a minute from the excess gas and another 2988.7 a minute from the light oil. Since the rocket fuel line in this scenario isn't active, there's only one thing that can be done with all that solid fuel: Burn it.

Which requires ~963 MW worth of boiler power. Oof.

[jodokus31]

That leaves lubricant. Since lubricant can only be made from heavy oil

You could also check coal liquefaction. This might be better for lubricant only

[foamy]

That leaves lubricant. Since lubricant can only be made from heavy oil

You could also check coal liquefaction. This might be better for lubricant only

Yeah, it's better balance-wise since it a. incorporates its own flare and b. starts primarily with heavy oil. Trouble is the enormous pile of coal you need to shovel in to get the same production as from an oil plant, and still leaves you with 'how do I balance my oil plant' as a problem :v

[jodokus31]

Yeah, it's better balance-wise since it a. incorporates its own flare and b. starts primarily with heavy oil. Trouble is the enormous pile of coal you need to shovel in to get the same production as from an oil plant, and still leaves you with 'how do I balance my oil plant' as a problem :v

The use case to produce only blue belts and nothing else is the most common use case for heavy oil only, I would assume.

I checked:
https://kirkmcdonald.github.io/calc.html

To produce 1 blue belt / s (with coal liquefaction, prod 3 where possible except miners, no speed) you need:
- 1.5 yellow lines of iron ore
- 0.2 yellow lines of coal into 1.7 refineries
- the needed steam for liquefaction comes from solid fuel which reduces the surplus below a small bit.
You have a surplus of 4.6 light oil / s and 2.4 gas / s

To produce 1 blue belt / s (with advanced oil, prod 3 where possible except miners, no speed) you need:
- 1.5 yellow lines of iron ore
- 47.3 crude oil / s into 4.4 refineries
And you have a surplus of 27.7 light oil / s and 33.8 gas / s

[mrvn]

pass stuffs, particularly aforementioned ammos, by simply having inserters leaving things on ground, without using any chests or belts.

That costs you UPS because the inserters won't go to sleep when blocked. For all the turrets on your boundary that adds up.

[Inglonias]

My standard factory is typically a belt bus with boutiques along one side or the other. I've never used a giant mall until my current game. Up until recently, I also seriously handicapped myself by not using a lot of what the game has to offer. Nuclear power and combat robots being the two big ones.

[Koub]

[...] I also seriously handicapped myself by not using a lot of what the game has to offer. Nuclear power and combat robots being the two big ones.

You don't miss much with vanilla combat robotics They are so underwhelming compared to alternatives.
More on this here : viewtopic.php?f=16&t=68384

[Inglonias]

[...] I also seriously handicapped myself by not using a lot of what the game has to offer. Nuclear power and combat robots being the two big ones.

You don't miss much with vanilla combat robotics They are so underwhelming compared to alternatives.
More on this here : viewtopic.php?f=16&t=68384

I'm confused. The recipes were changed midway thru .17 to not need the flying frames anymore.

I find defenders to be quite handy because it's like having a dozen extra guns follow me around wherever I go (at least for the lifetime of the robot). That being said, I do wish they weren't tied to the player. It looks like Rampant Arsenal lets you deploy Capsule turrets which can launch the robots, but I've never tried it. If I could wave a wand and change how they work, you'd put combat robots in some flavor of roboport and they'd deploy out to a fixed range from that port when needed, kill whatever needs to die, and return when done.