Optimized Steam Engine Setup

Power Plants, Energy Storage and Reliable Energy Supply. All about efficient energy production. Turning parts of your factory off. Reliable and self-repairing energy.
Qon
Smart Inserter
Smart Inserter
Posts: 2164
Joined: Thu Mar 17, 2016 6:27 am
Contact:

Re: Optimized Steam Engine Setup

Post by Qon »

Patric20878 wrote:I looked at that picture already. Where is there anything about 200% pumps and 1 tile per pump? If you think artificial shore length doesn't count as shore length, you're totally missing the idea. Every pump is still taking the normal space.
It's 6 pumps on 6 tiles vertically. It increases the width, not the shore length. The vertical distance is the same, 6 tiles. The pumps do take the same amount of space, but not vertically since they are stacked horizontally. You can't use that space much more effectivly without also landfilling lots of water to place the steam engines, so it is essentially free extra water that would otherwise go unused.

It's my idea, so I'm not missing it. The idea is to get as much power and water per shore length possible. The shore length is the vertical distance, so the extra landfill strips do not count, because they do not increase the vertical distance.
My mods: Capsule Ammo | HandyHands - Automatic handcrafting | ChunkyChunks - Configurable Gridlines
Some other creations: Combinassembly Language GitHub w instructions and link to run it in your browser | 0~drain Laser
BlakeMW
Filter Inserter
Filter Inserter
Posts: 962
Joined: Thu Jan 21, 2016 9:29 am
Contact:

Re: Optimized Steam Engine Setup

Post by BlakeMW »

I'm impressed by the Qon design. The horizontal space density for the first half of the steam engine setup is basically identical to having a single pipe between each steam engine - that's not the best but it's acceptable, and the second half is much more compact, so on average it works out pretty darn good. The cost of 2x pipe to ground is low (15 iron) and they only need to be used in half the layout, considering a steam engine is 20 iron and a medium pole is 10 iron + 4 copper, the pipe to grounds wouldn't come to more than about a quarter of the cost of the steam engine setup, which is negligible anyway.

I don't think anything is ever really going to beat vertically stacking a bunch of 1x14 boiler strings for me. But if I ever get around to making a GW powerplant I'd definitely consider the 30 long steam engine strings.
Qon
Smart Inserter
Smart Inserter
Posts: 2164
Joined: Thu Mar 17, 2016 6:27 am
Contact:

Re: Optimized Steam Engine Setup

Post by Qon »

BlakeMW wrote:I'm impressed by the Qon design. [...] if I ever get around to making a GW powerplant I'd definitely consider the 30 long steam engine strings.
Thanks BlakeMW. I might be able to compress or upgrade it further though. I'm thinking that it might be possible to merge the pipes slightly earlier without losing efficiency. And I maybe I could use the 2/29/19 setup doubled up to 36-on-a-row interleaved, but then It would be slightly less space and material cost efficient. Not that important when making multi GW powerplants though since it's still much cheaper and smaller than solar if you disregard the fuel cost.
My mods: Capsule Ammo | HandyHands - Automatic handcrafting | ChunkyChunks - Configurable Gridlines
Some other creations: Combinassembly Language GitHub w instructions and link to run it in your browser | 0~drain Laser
BlakeMW
Filter Inserter
Filter Inserter
Posts: 962
Joined: Thu Jan 21, 2016 9:29 am
Contact:

Re: Optimized Steam Engine Setup

Post by BlakeMW »

Hmmm, I think you'd have trouble pushing the steam engine string length much higher, I don't think you could get much more than 50 pipe segments + steam engines in an interleaved row before running up against pretty hard and fast flow limits. If 36 works it'd be very close to the limit.
User avatar
MeduSalem
Smart Inserter
Smart Inserter
Posts: 1686
Joined: Sun Jun 08, 2014 8:13 pm
Contact:

Re: Optimized Steam Engine Setup

Post by MeduSalem »

Qon wrote:
BlakeMW wrote:I've been thinking about rocket fuel in boilers. You lose 10% of the fuel value by converting solid fuel to rocket fuel but otoh you can use productivity modules and with 4x Prod3 module in assembler 3 you get 26% more fuel value - a gross profit of 45MJ of electricity. The net profit is trickier, I think the minimum cost to perform the conversion (fully utilized alternating rows assembler/beacon) is about 15MJ leaving a net profit of 30MJ. Of course, if supplementing with solar power the electricity expense can be further discounted.
Rocket Fuel also saves energy on inserts (for all inserter types, not only burner inserter) but actually increases total factory inserts because the 10 solid fuel had to the inserted into the chemical plant to make the rocket fuel - you're not saving energy on inserts, only changing where those inserts happening and causing more inserts.
I completely forgot about inserts reuired to make the rocket fuel, thanks.

But the last part, did you account for the inserters used to make rocket fuel will not be burner inserters but stack inserters or fast ones? The number of inserts isn't the only thing that determines how much inserter energy is used since different inserters use different amounts of power. The inserters that are working often are more energy efficient/movement and the ones moving the rocket fuel into boilers have no drain.

Also if inserter energy used goes up from making rocket fuel it doesn't matter that much with how much extra is gained by productivity modules. Burners are still the energy efficient alternative at your boilers with rocket fuel now, so maybe people will stop using electric ones for the tiny energy usage advantage electric ones have for coal. Using electric ones is just a big risk and hassle.

But getting rid of overdraw killing your power plant completely requires burner mining drills, burner inserters at coal train stations for un/loading cargo wagons, no bot transport of fuel. For oil it's not possible to get rid of your electric dependence. The most important part is inserters to boilers though since it requires a pretty short overdraw to activate the death spiral. You should also also use burner inserters from your buffer storage at the power plant to belts. If you place stack inserters higher up on the belt then the burners will idle until the stack inserter lose power, so they will only activate in an emergency anyways.

But ultimately inserters are not a big power drain for your factory anyways.
I tried the Rocket Fuel solution today for my Steam Engine setup just for fun. Just let me say that it is a huge waste of resources.

The overhead introduced by converting Solid Fuel into Rocket Fuel is outweighing the gains by far. Even with Productivity Module 3s in Assembler 3s using alternating rows.

I remember that BlakeMW and I had a discussion in another thread somewhere if Speed Beacons and Productivity Modules are reasonable in Solid Fuel production... to which I never got to respond due to lack of time until now ( sorry :/ )... but now that I have a chance I came to the conclusion that it doesn't pay off.

The additional Fuel gained by the Productivity Modules doesn't outweigh the electricity used during the production by the Productivity Modules and Beacons. Not even close.

In my map the power consumption overhead of creating power increased 6-7 times compared to the same setup using Efficiency Modules. Basically it increased from 5% (15MW of 300MW total power, Efficiency Modules) to about 33% (100MW of 300MW total power, PM3s everywhere, Beacons with SM3s everywhere, and converting Solud Fuel to Rocket Fuel).

So basically Productivity Modules and Beacons in the Oil Industry for Fuel production are the worst thing you can do if you are using Steam Only. Rocket Fuel doesn't pay off either. The fuel value of Rocket Fuel would have to be somewhere around 300-400MJ to make it worthwhile, otherwise you lose too much in the conversion process.
Qon
Smart Inserter
Smart Inserter
Posts: 2164
Joined: Thu Mar 17, 2016 6:27 am
Contact:

Re: Optimized Steam Engine Setup

Post by Qon »

MeduSalem wrote: I tried the Rocket Fuel solution today for my Steam Engine setup just for fun. Just let me say that it is a huge waste of resources.

The overhead introduced by converting Solid Fuel into Rocket Fuel is outweighing the gains by far. Even with Productivity Module 3s in Assembler 3s using alternating rows.

I remember that BlakeMW and I had a discussion in another thread somewhere if Speed Beacons and Productivity Modules are reasonable in Solid Fuel production... to which I never got to respond due to lack of time until now ( sorry :/ )... but now that I have a chance I came to the conclusion that it doesn't pay off.

The additional Fuel gained by the Productivity Modules doesn't outweigh the electricity used during the production by the Productivity Modules and Beacons. Not even close.

In my map the power consumption overhead of creating power increased 6-7 times compared to the same setup using Efficiency Modules. Basically it increased from 5% (15MW of 300MW total power, Efficiency Modules) to about 33% (100MW of 300MW total power, PM3s everywhere, Beacons with SM3s everywhere, and converting Solud Fuel to Rocket Fuel).

So basically Productivity Modules and Beacons in the Oil Industry for Fuel production are the worst thing you can do if you are using Steam Only. Rocket Fuel doesn't pay off either. The fuel value of Rocket Fuel would have to be somewhere around 300-400MJ to make it worthwhile, otherwise you lose too much in the conversion process.
Just some quick calculations:
((rocket_fuel_value*productivity_bonus - solid_fuel_value*10)*boiler_efficiency) - energy_used_in_conversion
energy_used_in_conversion = (assemby_machine_power*(module_energy_increase)+beacon_power) * recipe_energy * crafting_speed * productivity_bonus
(225*1.4 - 25*10)*0.5 - (0.210*(1+8.8)+0.48)*30/5.5/1.4 = 22.61 MJ net gain (with boiler efficiency considered) solid fuel -> rocket fuel
Assuming infinite rows and no energy for all the insertions and bots used. But that's a big increase in energy.
And then there's the savings in oil, and the fact that you can reduce power used for beaconised pumpjacks, refineries, heavy cracking all by 30%.

Did I forget something?

So obviously your setup is not infinite. But it still shouldn't be that bad if you have a big enough factory. Pics?

Edit: math error.
Last edited by Qon on Thu Jul 07, 2016 5:12 pm, edited 4 times in total.
My mods: Capsule Ammo | HandyHands - Automatic handcrafting | ChunkyChunks - Configurable Gridlines
Some other creations: Combinassembly Language GitHub w instructions and link to run it in your browser | 0~drain Laser
BlakeMW
Filter Inserter
Filter Inserter
Posts: 962
Joined: Thu Jan 21, 2016 9:29 am
Contact:

Re: Optimized Steam Engine Setup

Post by BlakeMW »

MeduSalem wrote: In my map the power consumption overhead of creating power increased 6-7 times compared to the same setup using Efficiency Modules. Basically it increased from 5% (15MW of 300MW total power, Efficiency Modules) to about 33% (100MW of 300MW total power, PM3s everywhere, Beacons with SM3s everywhere, and converting Solud Fuel to Rocket Fuel).
Okay, with beacons you have to be careful because they constantly draw power whether their benefit is needed or not. This means that if the assemblers/chemical plants are not running absolutely 100% of the time without exception you are wasting large amounts of electricity - of course this isn't a problem with solar/accu because the energy is wasted anyway if you don't use it, but if you're relying on solid fuel for electricity you have to be careful.

But what you can do to guarantee utilization is cut power using a power switch.

First have a storage tank on the light oil input, if the storage tank becomes depleted cut power. Next have a buffer chest for the rocket fuel, if the buffer chest becomes full, cut power. This was you can guarantee that when the system does run, it does so at 100%.

The theoretical profit when you have an infinity of alternating rows of speed3 beacons and prod3 assembler3's, is as follows:
Energy Consumption: 2058kW + 480kW (1/8th x 8 beacons) = 2538kW
Production Time: 30s
Crafting Speed: 5.5
Electricity consumption per rocket fuel: 2538kW * 30s / 5.5 = 13843kJ (14MJ)

Energy increase over solid fuel: 225 * 1.4 - 10 * 25 = 65MJ
Net electricity profit: 65MJ / 2 - 14MJ = 18.5MJ

The final profit is approximately +15% compared with burning solid fuel (and btw as of version 13.5 these calculations are dead on, the assemblers use exactly as much energy as they say they do, in previous versions the game could deviate significantly although mainly for fast recipes)

But of course actually getting that +15% profit requires a lot of things be done exactly right.

I would generally consider this most plausible when you already have alternating rows beacons which you are committed to using, consider for example the plausible case that you're already producing rocket parts including rocket fuel in alternating rows setup for the purpose of launching rockets. Extending this setup to create excess fuel for burning will be much more economically favorable than creating an entire setup from scratch.

The other condition which makes it much more energetically favorable is if you gain most your energy from solar - for example when you're creating fuel for a backup (100% discount) or night steam plant (70% discount). If you can discount some or all of the electricity cost then the energy profit is much higher.

And as Qon points out, you're also saving oil and associated processing costs - essentially this involves taking electricity and making fuel out of thin air - although this argument only holds weight if you plan to do things other than burn the fuel for electricity.

If none of those mitigating factors can be applied, even though it's theoretically possible to turn a good energy profit by building a large beacons/assembler alternating rows setup and only powering it on when required, it will almost certainly never justify the enormous sunk cost in expensive modules - and I should stress, the margins are so low this doesn't work at all with cheaper modules. If you don't have the prod3/speed3 modules lying around you're much better off using eff1 modules in solid fuel creation and skipping the rocket fuel.
User avatar
MeduSalem
Smart Inserter
Smart Inserter
Posts: 1686
Joined: Sun Jun 08, 2014 8:13 pm
Contact:

Re: Optimized Steam Engine Setup

Post by MeduSalem »

Yeah, I remember from the past discussions that mathematically it should work somehow, which is why I tried the setup, but in reality it doesn't for some reason.

My guess is that the setup isn't infinite IS the problem why it isn't efficient at all.

And yeah, the Solid Fuel production line is basically running 24/7 non-stop. The Rocket Fuel is a little bit piled up, but not that much and I disabled the Beacons that are superfluous to reduce the unnecessary overhead and still it doesn't break even.

Here is the setup:

I basically adapted my old 0.12 map with the ideas of beacons and rocket fuel. It might not be the most efficient layout, but maybe you guys catch something I have overlooked... at this point I guess I must have done something wrong which I am just too blind to see:
0.13 - Oil Industry 1.jpg
0.13 - Oil Industry 1.jpg (502.05 KiB) Viewed 9738 times
Don't wonder about the weird cracking setup... it doesn't get used at all except for Heavy Oil -> Light Oil cracking, so basically no cracking is done. The Light Oil gets used up completely before anything could be cracked.

Here is the power graph for the Oil Industry shown in above picture (all the necessary stuff like pump jacks, refineries, cracking, solid fuel, rocket fuel and the beacons are seperated from the rest of the base to prevent a power death spiral):
0.13 - Oil Industry Power.jpg
0.13 - Oil Industry Power.jpg (184.63 KiB) Viewed 9738 times
Here is the power graph for the base (the factory itself is located behind the accumulators to gurantee that turrets get priority, which is still the 0.12 approach, haven't tried to fiddle with power switches yet):
0.13 - Factory Power.jpg
0.13 - Factory Power.jpg (212.04 KiB) Viewed 9738 times
Last edited by MeduSalem on Thu Jul 07, 2016 12:43 pm, edited 1 time in total.
BlakeMW
Filter Inserter
Filter Inserter
Posts: 962
Joined: Thu Jan 21, 2016 9:29 am
Contact:

Re: Optimized Steam Engine Setup

Post by BlakeMW »

You get high power efficiency out of having more beacons and fewer assemblers, strange but true.

That is you should start by making sure that *every* assembler is covered by 8 beacons, this is more important than making sure that every beacon covers 8 assemblers! So the beacon rows should extend as far as necessary beyond the assembler rows.

The assemblers are the ultra expensive energy hogs which need to be maximized as much as possible, beacons are comparatively cheap both in terms of cost and energy.
Last edited by BlakeMW on Thu Jul 07, 2016 12:50 pm, edited 2 times in total.
User avatar
MeduSalem
Smart Inserter
Smart Inserter
Posts: 1686
Joined: Sun Jun 08, 2014 8:13 pm
Contact:

Re: Optimized Steam Engine Setup

Post by MeduSalem »

BlakeMW wrote:You get high power efficiency out of having more beacons and fewer assemblers, strange but true.

That is you should start by making sure that *every* assembler is covered by 8 beacons, this is more important than making sure that every beacon covers 8 assemblers! So the beacon rows should extend as far as necessary beyond the assembler rows.

The assemblers are the ultra expensive energy hogs which need to be maximized as much as possible, beacons are comparatively cheap both in terms of cost and energy.
I thought of that already and strangely I implemented my Smart Furnaces that way... but I somehow forgot about that behaviour in the Oil setup. :D

I'll make the necessary adjustments... thanks so far for the advice. I'll come back later and post the new results.
User avatar
MeduSalem
Smart Inserter
Smart Inserter
Posts: 1686
Joined: Sun Jun 08, 2014 8:13 pm
Contact:

Re: Optimized Steam Engine Setup

Post by MeduSalem »

@BlakeMW: I now have more Beacons than Assemblers. Also more Beacons than Solid Fuel Plants. Well... it still doesn't work. It only optimized the consumption "a bit".

The Energy consumption still far outweights anything that could be done with Efficiency Modules. It still requires almost the same amount of power (average being 80MW now compared to the 90MW from before). Though it dropped for about 20 minutes to 70MW, probably an effect where everything tried to balance itself out, but then the consumption increased again. So the relocation of the Beacons squeezed about 10MW out.

I must be doing something so horribly wrong... I don't know. That or the math foundation is completely wrong all together.



But really... How is 20% or even 40% productivity surplus supposed to outweigh Energy efficiency Modules?

The efficiency modules decrease the consumption to 20% of the original value. Meaning I can build 5 times the chem plants/refineries and thereby gain 5 times the throughput without additional energy requirement.

On the other hand with Productivity Modules + Speed Beacons the Chem Plant to produce Solid Fuel for example gets a speed increase of 5.875 (+370%), but also an Energy Consumption of 1.7MW (+720%)... So basically the consumption goes up by a magnitude of 7.2 from the Chem Plants alone. Additionally the Beacons also require Energy (which is harder to determine due to its effects being distributed unevenly among the chem plants). But in sum it feels like the gained Productivity of 20% additional fuel is wasted in the process of gaining said productivity.

The same thing with Rocket Fuel, just that Rocket Fuel is additionally handicaped due to only gaining 225 MJ of Fuel value compared to 10 x 25MJ

All the math done until now doesn't really add up to what I am measuring.


[edit]

I have further tweaked it... with the suggestion of using Power switches. With Power Switches kicking in depending on buffer storages I can cut down the consumption for short periods as low as 70MW when Solid Fuel hits the cap and sometimes to 35MW when Rocket Fuel hits the cap or something. And with short periods I mean split seconds.

But still... most of the time it is at 95MW... Maybe the overall average is now at 80-85MW. Can't tell exactly due to how the switches mess with the stats. But seems like I managed to minimize the leakage quite "well" already without the Power Switches.

[edit 2]

I'm getting closer... but one really has to pull all kinds of tricks.

I now hooked up the Refineries to a power switch as well, which works on Petroleum Gas cap (my refineries usually stalled anyways due to hitting Petroleum gas cap so might as well use that behaviour to my advantage), I hooked the Pump Jacks up to power switches with Crude Oil cap.

That gets the average power consumption of the Oil Industry really, really low. About 30-40MW or something (Total Power Output is still 300MW from last page... so the overhead is now down to 10%).... which should be alright for what we are trying to accomplish here. I think it couldn't be much better with Efficiency Modules at this point. The 30MW can easily be handled by 60 steam engines or so.

Now the big problem is... the damn power spikes when EVERY single power switch turns on at the same time. Then the spike gets up to 100-120 MW or so.

To handle that I started to use arrays of accumulators... which fill up when the Power Usage of the Oil Industry is very low, and get drained during the spikes when every Power Switch is on.

The approach has its advantages... I really start to notice now how less and less Crude Oil is consumed by eliminating all the drains and using the advantages of additional productivity.

... the work that has to go into that build though... urgh. A real project. xD
Qon
Smart Inserter
Smart Inserter
Posts: 2164
Joined: Thu Mar 17, 2016 6:27 am
Contact:

Re: Optimized Steam Engine Setup

Post by Qon »

BlakeMW wrote: Electricity consumption per rocket fuel: 2538kW * 30s / 5.5 = 13843kJ (14MJ)

Energy increase over solid fuel: 225 * 1.4 - 10 * 25 = 65MJ
Net electricity profit: 65MJ / 2 - 14MJ = 18.5MJ
Your electric consumption per rocket fuel doesn't consider that you get 40% more rocket fuel for each successful craft cycle, decreasing the time to get a rocket fuel and energy to craft it by almost 30%.
But I had some errors in my calculation that I had to fix too. Hope I'm more correct now.
BlakeMW wrote: And as Qon points out, you're also saving oil and associated processing costs - essentially this involves taking electricity and making fuel out of thin air - although this argument only holds weight if you plan to do things other than burn the fuel for electricity.
It always holds weight though. Why wouldn't it apply for fuel? If you can extract more net energy out of the same amount of crude oil then you can scale down your crude oil pumping until you don't produce more rocket fuel than you consume.
You just can't count the same gain twice. You either count as decrease of pumping and refining or as increased net energy profit (or a mixture of both, as long as you don't count any profit twice).
My mods: Capsule Ammo | HandyHands - Automatic handcrafting | ChunkyChunks - Configurable Gridlines
Some other creations: Combinassembly Language GitHub w instructions and link to run it in your browser | 0~drain Laser
Qon
Smart Inserter
Smart Inserter
Posts: 2164
Joined: Thu Mar 17, 2016 6:27 am
Contact:

Re: Optimized Steam Engine Setup

Post by Qon »

MeduSalem wrote: The Energy consumption still far outweights anything that could be done with Efficiency Modules. It still requires almost the same amount of power (average being 80MW now compared to the 90MW from before). Though it dropped for about 20 minutes to 70MW, probably an effect where everything tried to balance itself out, but then the consumption increased again. So the relocation of the Beacons squeezed about 10MW out.

But really... How is 20% or even 40% productivity surplus supposed to outweigh Energy efficiency Modules?
The point of productivity isn't to minimize electricity used in that process. The point is that making 1 solid fuel ideally takes 10MJ but it's worth both the cost of losing 10% fuel value (without productivity) and the cost of the conversion. And the 40% extra of 225 MJ is 90 MJ. That gain exceeds both the conversion process energy cost and the loss of the solid fuel used as input even after halving it in the boiler.

It's not a competition with efficiency at all. Even if your efficiency modules could remove 100% of the energy used to convert solid fuel to rocket fuel you would still prefer not converting solid fuel to rocket fuel at all since 10 solid fuel is worth more than 1 rocket fuel. Making rocket fuel with the intention of burning it is never a good idea without productivity. Of course your energy used goes up a lot when making the rocket fuel with productivity. But the amount of fuel you extract more than weighs up the losses if you do it right.

I'm confused what you are comparing the energy drain on the system with. If you are comparing the energy cost of making rocket fuel with efficiency modules then efficiency modules are going to win. But then you are disregarding the fact that all energy used for the conversion with efficiency is energy you pay to lose fuel.
My mods: Capsule Ammo | HandyHands - Automatic handcrafting | ChunkyChunks - Configurable Gridlines
Some other creations: Combinassembly Language GitHub w instructions and link to run it in your browser | 0~drain Laser
BlakeMW
Filter Inserter
Filter Inserter
Posts: 962
Joined: Thu Jan 21, 2016 9:29 am
Contact:

Re: Optimized Steam Engine Setup

Post by BlakeMW »

MeduSalem wrote: But really... How is 20% or even 40% productivity surplus supposed to outweigh Energy efficiency Modules?
Chances are you overestimate eff modules. The basic recipe to create a solid fuel in an unmoduled chemical plant requires 432kJ (180kW * 3 / 1.25) - that is all the efficiency module has to work with, even 2x eff2 module only saves you 2.7% of the 12.5MW of electricity fuel value. The productivity modules are working with the much larger number of 12500kJ, 20% of 12500 is much bigger than 80% of 432. This basically is why productivity modules kick the ass of efficiency modules.
The only real reason to use efficiency modules is that a watt saved by an eff1 module is usually cheaper than a watt generated by solar/accu, they are really great if you want to minimize the time for return on investment. eff2 modules though are an extremely dubious investment because the cost per watt saved is so high, although it's quite possible that a watt saved by an eff2 module is cheaper than a watt generated thanks to prod3/speed3 beacons ;).
Additionally the Beacons also require Energy (which is harder to determine due to its effects being distributed unevenly among the chem plants).
Actually beacons are pretty easy, just count the number of beacons, multiply by 480kW, then divide that over the chemical plants, so if for instance you have 8 beacons and 16 chemical plants, you can just add 480 * 8 / 16 = 240kW to each chemical plant - I know it's not absolutely perfect when plants operate at different speeds, but it should be close enough.
Now the big problem is... the damn power spikes when EVERY single power switch turns on at the same time. Then the spike gets up to 100-120 MW or so.

To handle that I started to use arrays of accumulators... which fill up when the Power Usage of the Oil Industry is very low, and get drained during the spikes when every Power Switch is on.
Since you're using steam energy you might consider hot water storage - it is 100% oil free and very tunable to deal with longer or shorter spikes by adding more tanks or more steam engines. It's not as compact by kW, but can be much more compact by kJ stored because a hot water tank stores 42.5x as much energy as an accumulator. Hot water storage only really makes sense when you're devoted to steam power, but then I find it really makes sense.
The approach has its advantages... I really start to notice now how less and less Crude Oil is consumed by eliminating all the drains and using the advantages of additional productivity.

... the work that has to go into that build though... urgh. A real project. xD
Indeed. And in the end you create the most expensive electricity possible ;). With all the high end modules it probably ends up like 10x as expensive per kW as solar/accu. But we do these things not because they are easy, but because they are hard, right?
Last edited by BlakeMW on Thu Jul 07, 2016 5:46 pm, edited 2 times in total.
BlakeMW
Filter Inserter
Filter Inserter
Posts: 962
Joined: Thu Jan 21, 2016 9:29 am
Contact:

Re: Optimized Steam Engine Setup

Post by BlakeMW »

Qon wrote:
BlakeMW wrote: Electricity consumption per rocket fuel: 2538kW * 30s / 5.5 = 13843kJ (14MJ)

Energy increase over solid fuel: 225 * 1.4 - 10 * 25 = 65MJ
Net electricity profit: 65MJ / 2 - 14MJ = 18.5MJ
Your electric consumption per rocket fuel doesn't consider that you get 40% more rocket fuel for each successful craft cycle, decreasing the time to get a rocket fuel and energy to craft it by almost 30%.
I meant per production cycle, not per rocket fuel created. I do the calculation per production cycle, and multiply the output by 1.4
User avatar
MeduSalem
Smart Inserter
Smart Inserter
Posts: 1686
Joined: Sun Jun 08, 2014 8:13 pm
Contact:

Re: Optimized Steam Engine Setup

Post by MeduSalem »

Qon wrote:I'm confused what you are comparing the energy drain on the system with. If you are comparing the energy cost of making rocket fuel with efficiency modules then efficiency modules are going to win. But then you are disregarding the fact that all energy used for the conversion with efficiency is energy you pay to lose fuel.
I was comparing

Solid Fuel production using Efficiency Modules (No Rocket Fuel)

with

Solid Fuel + Rocket Fuel using Productivity Modules + Speed Beacons.

for the use in Boilers.


The original discussion I had with BlakeMW in the other thread was if using PM+SMs is actually more efficient than Efficiency Modules, to which BlakeMW responded with a calculation that showed that it PM+SMs are actually a small percentage better even with all the losses involved. The discussion was about 0.12 though, so back then Rocket Fuel couldn't be burned in boilers so we only talked about using Solid Fuel + PMs/SMs. That said BlakeMW's calculation was done assuming infinite rows of Assemblers/Beacons, which in reality never happens due to the processing area being finite.

So back then we didn't know if it is worth going for with all the additional overhead involved. Probably not. But now that Rocket Fuel can be burned in boilers we get an additional layer to use Productivity Modules... when turning Solid Fuel to Rocket Fuel using 40% extra productivity... which I thought now might be enough to make it worthwhile.


BlakeMW wrote:Chances are you overestimate eff modules. The basic recipe to create a solid fuel in an unmoduled chemical plant requires 432kJ (180kW * 3 / 1.25) - that is all the efficiency module has to work with, even 2x eff2 module only saves you 2.7% of the 12.5MW of electricity fuel value. The productivity modules are working with the much larger number of 12500kJ, 20% of 12500 is much bigger than 80% of 432. This basically is why productivity modules kick the ass of efficiency modules.
The only real reason to use efficiency modules is that a watt saved by an eff1 module is usually cheaper than a watt generated by solar/accu. eff2 modules are an extremely dubious investment because the cost per watt saved is so high, although it's quite possible that a watt saved by an eff2 module is cheaper than a watt generated thanks to prod3/speed3 beacons ;)
Probably it is like that.

I remember that I somewhere made an excel sheet for Smart Furnaces where I compared various setups using EM2s, PM3s and SM3s... and from there I only remembered that PM3/SM3s always have to go in tandem, but are never as energy efficient as EM2s. But that calculation was only for consuming resources... and I never applied the calculation for generating additional energy because I found the calculation necessary for that way too nasty.
BlakeMW wrote:Since you're using steam energy you might consider hot water storage - it is 100% oil free and very tunable to deal with longer or shorter spikes by adding more tanks or more steam engines. It's not as compact by kW, but can be much more compact by kJ stored because a hot water tank stores 42.5x as much energy as an accumulator. Hot water storage only really makes sense when you're devoted to steam power, but then I find it really makes sense.
Yeah well it might be worth considering... but if you store the energy as Hot Water instead one would have to be able to convert the Hot Water on the fly to Power... meaning in my case I would have to use 240 Steam Engines just to cover a spike of 120MW. Of course the energy density in Hot Water tanks is much higher, but the space required by the Steam Engines is also nasty.

On the other hand using accumulators I can use 30 Steam Engines running 24/7 on 100% and instead build an array of accumulators storing the surplus power directly.

Good question which approach would take less space in the end. The energy storage variants aren't exactly my expertise.
BlakeMW wrote:Indeed. And in the end you create the most expensive electricity possible ;). With all the high end modules it probably ends up like 10x as expensive per kW as solar/accu. But we do these things not because they are easy, but because they are hard, right?
Exactly.

The entire contraption doesn't make any sense anyways... Not in a regular game with a predetermined end. If one puts a few thousand hours into the map then it might be worthwhile, don't know. xD

I already researched everything on that map etc so basically I am just fooling around now to see what's possible. I don't really care anymore how many resources go to waste for Modules... As long as the contraption looks nice and has some highly theoretical benefits. :D
Aru
Fast Inserter
Fast Inserter
Posts: 212
Joined: Wed Apr 13, 2016 11:31 pm
Contact:

Re: Optimized Steam Engine Setup

Post by Aru »

This is my fault, I focus very much on optimizing early and mid game designs, in order to progress faster. I'm really not even making optimized designs with beacons yet, hah. That's just how I am, my design progress is just like the game progress, early game first. I'll get to that stuff once I feel like I have the early and mid game perfected, which could very well be never. (edit: I just read changelog, looks like none of the numbers changed, so designs are still good, yay.)

But, it's about low tech, low resource cost (minimal everything), being easily tileable, highly optimized, integrated, and balanced, and at least relatively space efficient. I might start with 1 engine and 2 boilers, and then from there go straight into a row of 10 engines with 14 boilers (or whatever the new numbers are, haven't really played with 0.13 yet), and then eventually get to solar, get combinator-controlled oil processing / storage / cracking going, train stations, electric smelting, Efficiency 1 modules, and THEN beacon designs, arrays of flexible bot-fed assembler mk3s, and productivity & speed modules. I'm still at the first part, I'm going to get back into it now, probably, in the next few days, because I was looking forward to v0.13 for so long, and it finally hit, and I don't even read the actual change log until 2 months after :P (I stopped playing for a while)
mrvn
Smart Inserter
Smart Inserter
Posts: 5969
Joined: Mon Sep 05, 2016 9:10 am
Contact:

Re: Optimized Steam Engine Setup

Post by mrvn »

Skellitor301 wrote:One thing I can see about that layout is the furnaces at the end will be a lower priority when it comes to coal. The first line will get half while the third one down will get like 1/8th. Other than that it's a good layout, I'll be sure to add it to the TDT
I actually think this is a good thing. If you do not have enough coal the first rows get all the coal (1/32th for the 5th row is basically nothing). That means you still get 100°C water in the first row. Then if there is enough coal you get 100°C in the second row and so on. Driving steam engines with luke warm water makes no sense so you want as many rows with 100°C water as possible and then quickly waste no coal on the rest.
Gnark
Inserter
Inserter
Posts: 43
Joined: Tue Oct 18, 2016 10:15 am
Contact:

Re: Optimized Steam Engine Setup

Post by Gnark »

Building my nest power plan :

Image
Image
Image

I have 3 14 boiler line for 2 engine line.
I have 21 engine per line and 3 tank per line.

My tank are managed by pumps that make them deliver the full power for 88% of the load of the tank, they also made my boiler work full load until the tank are 90% full.

I use a 5*5 design for my engine 5 engine 1 tube 1 electric pole.
nevniv
Inserter
Inserter
Posts: 49
Joined: Wed Sep 21, 2016 1:37 am
Contact:

Re: Optimized Steam Engine Setup

Post by nevniv »

This sticky needs an update for .15 soon.. unless there are some poor souls out there still playing .14
Post Reply

Return to “Energy Production”