So I'm taking a crack at a low Pollution run, and I'm running into major problems figuring out the trade off (and ratios) of Spd/Prod modules when it comes to Pollution statistics and possible AM3 compacting.
Right now I'm running eff1 in pretty much everything to bring the total pollution generation down. However, trying to do the math for the Spd/Prod (with or without beacons), I can't wrap my mind around how to approach the question of increasing throughput/mat efficiency without increasing pollution by huge amounts. This is assuming power isn't a problem (nuclear), setup ROI isn't a major concern, and the overall goal is to keep cloud spread as low as feasible.
I get the impression this isn't possible to end up in a net gain for compacting a setup, without huge increases in pollution output, and that realestate expansion might be the only viable (if not optimal) option. But since I'm using 0 biter attacks as a metric for success, expanding outward is something I want to avoid. I understand that mixing Eff modules with the others are is a waste of effort and space, since the energy per unit cost heavy favors Prod/Spd combos... I just can't visualize that curve with how it affects pollution production.
Am I out of my mind? Mostly.... but thats unrelated. But am I looking for an impossible solution?
Module trade offs for Low Pollution Run.
Re: Module trade offs for Low Pollution Run.
It's not impossible... just VERY EXPENSIVE.
Put productivity modules in AM3s. Surround them with efficiency beacons. If you have any spare beacon slots, add speed modules (and more efficency modules!) to reduce the AM3 count.
Put productivity modules in AM3s. Surround them with efficiency beacons. If you have any spare beacon slots, add speed modules (and more efficency modules!) to reduce the AM3 count.
Re: Module trade offs for Low Pollution Run.
Sychalis made a calculator, check out the post: viewtopic.php?f=5&t=41475
My own personal Factorio super-power - running out of power.
Re: Module trade offs for Low Pollution Run.
But it doesn't show any figures or even mention pollution output.Amarula wrote: ↑Sat Dec 14, 2019 7:23 pmSychalis made a calculator, check out the post: viewtopic.php?f=5&t=41475
Re: Module trade offs for Low Pollution Run.
Ok now this is starting to make sense to me. Unfortunately a Biter nest moved right next to my base and I shut everything down for a bit "to clear the air" (bad pun, and will go throw myself in a centrifuge later). I'll deal with them after I retool the factory a bit.
So basically what I came up with is 3 Prod3 + 1 Spd3 in each AM3, on an 8 beacon setup - which negates the speed penalty (with +5 left over) and pushes it down to only 30% total power usage. But with 30% less machinery, and that sweet stacking efficiency for Intermediate mats.
On the assumption that science labs have no pollution, 2 Prods and speed beacons, I can realistically cut the number labs I need in half.
So the next question..... should I rebuild Top down, or bottom up; since I have no research running atm?
(if anyone cares, I'm 30 hours in (mostly fumbling around), and only at 0.31 evolution.)
So basically what I came up with is 3 Prod3 + 1 Spd3 in each AM3, on an 8 beacon setup - which negates the speed penalty (with +5 left over) and pushes it down to only 30% total power usage. But with 30% less machinery, and that sweet stacking efficiency for Intermediate mats.
On the assumption that science labs have no pollution, 2 Prods and speed beacons, I can realistically cut the number labs I need in half.
So the next question..... should I rebuild Top down, or bottom up; since I have no research running atm?
(if anyone cares, I'm 30 hours in (mostly fumbling around), and only at 0.31 evolution.)
Re: Module trade offs for Low Pollution Run.
A key is that the bonuses/subtractions are additions to the base value:
[ 1 + (modifiers) ] = value, for everything except labs (whose speed is a multiplicative bonus for every stack, I think, it's weird, but they don't make pollution).
For pollution, the calculation is power ratio * pollution of building = pollution emitted. For a base miner, the pollution is 1 * 10. The reason efficiency modules lower that is they lower the power's contribution to the overall pollution equation.
For a fully beaconed assembler 3, the speed calculation is:
1.25 * (1 + 4 * -.15 + 12 * .5 * (2 * .5) ) = 8 * 1 craft rate. With 1.4x productivity, that's a craft rate of 11.2.
Energy consumption (ignoring tiny drain):
.375MW * (1 + 4 * .8 + 12 * .5 * (2 * .7) ) = 4.725 MW on the machine. Up to 4 beacons can be shared with a machine, and one may or may not use roboports to insert ingredients/remove products. Each beacon uses .480 megawatts, so it's not an insignificant component. For some recipes and longer beaconed set-ups, you'd need higher-tier belts and stack inserters, which aren't a low start-up cost.
Pollution emitted (per minute?):
(4.725/.375) * 2 * (1 + 4 * .1)) = 35.28. That's 17.64x more than the base assembler 3, but with a craft speed of 11.2, it's making 8.96x more products per time, for a final pollution ratio of 1.96875. I shouldn't be surprised it's higher than base, but it's only 2x the pollution per craft, for 8.96x faster crafting, with ~29% of that product being free (minus needing the power infrastructure, which for nuclear is pretty cheap compared to other generators).
If we're disregarding the importance of time here, an assembler 3 with 3 efficiency modules will be the easiest low energy/craft. Someone mentioned the speed 3's with efficiency beacons which makes sense because beacons and nuclear make little/no pollution (we do need to mine uranium and process it, but I'd expect it to be low overall contribution with efficiency modules and nuclear potency). Technically solar is fuel-free, so in theory at the limit of time, solar+accumulators is less pollution than nuclear, but in practice, you'd need a long time to recoup any difference. I'd question the value of beacon'd assembler 3's, because the sheer amount of oil and resources needed to make module 3's to module 1's is massive. For the amount of module 1's you need alone, it's 80 for every single module 3. Pollution reduction has a limit, so no matter the set-up, in vanilla, the least pollution you can emit is 80% of the original per craft. Ever. (unfortunately assembler 2's actually emit more pollution than 3's, so we can't try using a slower machine).
Comment: the old property in data used emissions_per_second_per_watt, but it was never based on active power, only the ratio of power.
[ 1 + (modifiers) ] = value, for everything except labs (whose speed is a multiplicative bonus for every stack, I think, it's weird, but they don't make pollution).
For pollution, the calculation is power ratio * pollution of building = pollution emitted. For a base miner, the pollution is 1 * 10. The reason efficiency modules lower that is they lower the power's contribution to the overall pollution equation.
For a fully beaconed assembler 3, the speed calculation is:
1.25 * (1 + 4 * -.15 + 12 * .5 * (2 * .5) ) = 8 * 1 craft rate. With 1.4x productivity, that's a craft rate of 11.2.
Energy consumption (ignoring tiny drain):
.375MW * (1 + 4 * .8 + 12 * .5 * (2 * .7) ) = 4.725 MW on the machine. Up to 4 beacons can be shared with a machine, and one may or may not use roboports to insert ingredients/remove products. Each beacon uses .480 megawatts, so it's not an insignificant component. For some recipes and longer beaconed set-ups, you'd need higher-tier belts and stack inserters, which aren't a low start-up cost.
Pollution emitted (per minute?):
(4.725/.375) * 2 * (1 + 4 * .1)) = 35.28. That's 17.64x more than the base assembler 3, but with a craft speed of 11.2, it's making 8.96x more products per time, for a final pollution ratio of 1.96875. I shouldn't be surprised it's higher than base, but it's only 2x the pollution per craft, for 8.96x faster crafting, with ~29% of that product being free (minus needing the power infrastructure, which for nuclear is pretty cheap compared to other generators).
If we're disregarding the importance of time here, an assembler 3 with 3 efficiency modules will be the easiest low energy/craft. Someone mentioned the speed 3's with efficiency beacons which makes sense because beacons and nuclear make little/no pollution (we do need to mine uranium and process it, but I'd expect it to be low overall contribution with efficiency modules and nuclear potency). Technically solar is fuel-free, so in theory at the limit of time, solar+accumulators is less pollution than nuclear, but in practice, you'd need a long time to recoup any difference. I'd question the value of beacon'd assembler 3's, because the sheer amount of oil and resources needed to make module 3's to module 1's is massive. For the amount of module 1's you need alone, it's 80 for every single module 3. Pollution reduction has a limit, so no matter the set-up, in vanilla, the least pollution you can emit is 80% of the original per craft. Ever. (unfortunately assembler 2's actually emit more pollution than 3's, so we can't try using a slower machine).
Comment: the old property in data used emissions_per_second_per_watt, but it was never based on active power, only the ratio of power.
I have mods! I guess!
Link
Link
Re: Module trade offs for Low Pollution Run.
Thanks Honktown this is very helpful...
I double checked the wiki https://wiki.factorio.com/Pollution, and it says
So modules https://wiki.factorio.com/Module, specifically productivity modules, are going to make the pollution worse, from 5% for level one modules, to 10% for level three.Modules that list "+x% pollution" increase pollution multiplier, not a flat pollution rate. Final pollution value is (pollution multiplier * energy usage multiplier * base pollution),
My own personal Factorio super-power - running out of power.
Re: Module trade offs for Low Pollution Run.
Ah yeah, I didn't phrase it the way the wiki does, but the algebra's the same. I grouped "pollution multiplier" with my "pollution of building". Unfortunately, although prod 3's increase productivity 10% for 10% more product, they also increase pollution multiplier the exact same 10%, so even if you reduce energy use by 80%, you still have +10% pollution for 10% more product, and 110/110 is 1x - no change in pollution per craft [disregarding energy change]Amarula wrote: ↑Mon Dec 16, 2019 3:54 pmThanks Honktown this is very helpful...
I double checked the wiki https://wiki.factorio.com/Pollution, and it saysSo modules https://wiki.factorio.com/Module, specifically productivity modules, are going to make the pollution worse, from 5% for level one modules, to 10% for level three.Modules that list "+x% pollution" increase pollution multiplier, not a flat pollution rate. Final pollution value is (pollution multiplier * energy usage multiplier * base pollution),
I have mods! I guess!
Link
Link
Re: Module trade offs for Low Pollution Run.
Thanks both you guys.
The intended goal was to to see if I could at least keep the pollution down enough that the surrounding tiles could clean them up. I didn't realize it at the time, but the starting location I got was well suited for this experiment..... Water gap so less walls to watch, and enough trees to soak up nearly all of the pollution from the factory.
I wasn't expecting absolute lowest possible, but seeing if containing it was feasible enough at least get up to rocket launch.... which I finally managed.
For the record..... yeah, the cost for module and beacon production ate up too much material, and forced train expansion (which creates higher risk of biter run ins, and another source of pollution). I meet most of my goals, but got messed up on the Pacifist aspect since the biter nest got too close to my train.... Shockingly they were just diagonal to a chunk with pollution, and dissipated before getting to them. Unfortunately that nest was too close to my train tracks...... if one of the twerps were at the edge of their nest, they'd aggro the train as it passed by. That happened 3 times, and accounts for all the attacks I had during the run.
Lessons learned here:
- Beacons arrays need to be scaled up way too much to make them practical for anything less then a med- large base. Too much realestate and really complicated belt pathing. Which in turn leads to a lot of material cost.
- While power and material weren't too major an issue, the level of production compacting for a full beacon setup wasn't enough to offset the effort. 2-3 beacons and speed modules is the best way to go.
- I found mixed modules in beacons do allow for some precise effect tuning, so spd3/eff3 got me +25% Spd, but net total power of +10%. (and stack eff1s in the machines themselves) This is incredibly nice for smelters, since 2x eff1 in slot and the mixed beacon leads to -50% power and +25% speed.
- Single or double beacons seems to be the ideal way to boost production at minimal risk and space cost. With stuff that benefits with production mods (end chain) can justify more beacon use.
-Total pollution control is actually realistic if you use the terrain to your advantage. But it also makes scaling up much more difficult.
- Cutting a clearing in the middle of a large forest chunk is basically a super power, if you can get the space you need. All my adv products was done this way.
- Ore fields in a forest are actually good for once. Soaks up all your pollution for you. With eff keeping it down to 2/m each, it rarely got more then 2 chunks in any direction.
- Would I do it again? I might. Knowing what I know now (and scumming map rolls for a good starting area), this is actually challenging in a fun way. Having to read the terrain for advantages, and making placement just as important as overall layout to keep things efficient/fed.
heres screen shots of the factory, and the pollution map right at the end before Rocket launch. The kills were from train aggro, and not pollution aggro, so I consider it a decent success.
https://imgur.com/a/8IF9OuV
The intended goal was to to see if I could at least keep the pollution down enough that the surrounding tiles could clean them up. I didn't realize it at the time, but the starting location I got was well suited for this experiment..... Water gap so less walls to watch, and enough trees to soak up nearly all of the pollution from the factory.
I wasn't expecting absolute lowest possible, but seeing if containing it was feasible enough at least get up to rocket launch.... which I finally managed.
For the record..... yeah, the cost for module and beacon production ate up too much material, and forced train expansion (which creates higher risk of biter run ins, and another source of pollution). I meet most of my goals, but got messed up on the Pacifist aspect since the biter nest got too close to my train.... Shockingly they were just diagonal to a chunk with pollution, and dissipated before getting to them. Unfortunately that nest was too close to my train tracks...... if one of the twerps were at the edge of their nest, they'd aggro the train as it passed by. That happened 3 times, and accounts for all the attacks I had during the run.
Lessons learned here:
- Beacons arrays need to be scaled up way too much to make them practical for anything less then a med- large base. Too much realestate and really complicated belt pathing. Which in turn leads to a lot of material cost.
- While power and material weren't too major an issue, the level of production compacting for a full beacon setup wasn't enough to offset the effort. 2-3 beacons and speed modules is the best way to go.
- I found mixed modules in beacons do allow for some precise effect tuning, so spd3/eff3 got me +25% Spd, but net total power of +10%. (and stack eff1s in the machines themselves) This is incredibly nice for smelters, since 2x eff1 in slot and the mixed beacon leads to -50% power and +25% speed.
- Single or double beacons seems to be the ideal way to boost production at minimal risk and space cost. With stuff that benefits with production mods (end chain) can justify more beacon use.
-Total pollution control is actually realistic if you use the terrain to your advantage. But it also makes scaling up much more difficult.
- Cutting a clearing in the middle of a large forest chunk is basically a super power, if you can get the space you need. All my adv products was done this way.
- Ore fields in a forest are actually good for once. Soaks up all your pollution for you. With eff keeping it down to 2/m each, it rarely got more then 2 chunks in any direction.
- Would I do it again? I might. Knowing what I know now (and scumming map rolls for a good starting area), this is actually challenging in a fun way. Having to read the terrain for advantages, and making placement just as important as overall layout to keep things efficient/fed.
heres screen shots of the factory, and the pollution map right at the end before Rocket launch. The kills were from train aggro, and not pollution aggro, so I consider it a decent success.
https://imgur.com/a/8IF9OuV
Re: Module trade offs for Low Pollution Run.
directly posing screenshots
Re: Module trade offs for Low Pollution Run.
Ugh I'm an idiot. I didn't consider that pollution of a building is dependent on per minute power and pollution, and not per craft. THANK YOU. I need to make a spreadsheet for possible speed/efficiency combos... unless you know of one already.starlinvf wrote: ↑Thu Dec 19, 2019 9:05 pmThanks both you guys.
The intended goal was to to see if I could at least keep the pollution down enough that the surrounding tiles could clean them up. I didn't realize it at the time, but the starting location I got was well suited for this experiment..... Water gap so less walls to watch, and enough trees to soak up nearly all of the pollution from the factory.
I wasn't expecting absolute lowest possible, but seeing if containing it was feasible enough at least get up to rocket launch.... which I finally managed.
For the record..... yeah, the cost for module and beacon production ate up too much material, and forced train expansion (which creates higher risk of biter run ins, and another source of pollution). I meet most of my goals, but got messed up on the Pacifist aspect since the biter nest got too close to my train.... Shockingly they were just diagonal to a chunk with pollution, and dissipated before getting to them. Unfortunately that nest was too close to my train tracks...... if one of the twerps were at the edge of their nest, they'd aggro the train as it passed by. That happened 3 times, and accounts for all the attacks I had during the run.
Lessons learned here:
- Beacons arrays need to be scaled up way too much to make them practical for anything less then a med- large base. Too much realestate and really complicated belt pathing. Which in turn leads to a lot of material cost.
- While power and material weren't too major an issue, the level of production compacting for a full beacon setup wasn't enough to offset the effort. 2-3 beacons and speed modules is the best way to go.
- I found mixed modules in beacons do allow for some precise effect tuning, so spd3/eff3 got me +25% Spd, but net total power of +10%. (and stack eff1s in the machines themselves) This is incredibly nice for smelters, since 2x eff1 in slot and the mixed beacon leads to -50% power and +25% speed.
- Single or double beacons seems to be the ideal way to boost production at minimal risk and space cost. With stuff that benefits with production mods (end chain) can justify more beacon use.
-Total pollution control is actually realistic if you use the terrain to your advantage. But it also makes scaling up much more difficult.
- Cutting a clearing in the middle of a large forest chunk is basically a super power, if you can get the space you need. All my adv products was done this way.
- Ore fields in a forest are actually good for once. Soaks up all your pollution for you. With eff keeping it down to 2/m each, it rarely got more then 2 chunks in any direction.
- Would I do it again? I might. Knowing what I know now (and scumming map rolls for a good starting area), this is actually challenging in a fun way. Having to read the terrain for advantages, and making placement just as important as overall layout to keep things efficient/fed.
heres screen shots of the factory, and the pollution map right at the end before Rocket launch. The kills were from train aggro, and not pollution aggro, so I consider it a decent success.
https://imgur.com/a/8IF9OuV
I have mods! I guess!
Link
Link
Re: Module trade offs for Low Pollution Run.
To quote Robot Chicken doing a Shawshank parody with DC characters.......
"Again........not so good with the math"
"Again........not so good with the math"
-
Itemfinder
- Inserter
- Posts: 29
- Joined: Thu Feb 21, 2019 4:46 pm
- Contact:
Re: Module trade offs for Low Pollution Run.
As a fan of minimalism, compact builds, and efficiency modules, I appreciate your factory. Thanks for sharing.
Re: Module trade offs for Low Pollution Run.
I've been playing with this a lot lately so I want to add a little bit to it:
My best designs are with assemblers running at 1.4375x production speed (+5%) with +40% prod and -80% power. Furnaces run a lot faster since they only take 2 prod modules. In all cases, both the production speed and power cost reduction contribute to reducing pollution, primarily the latter. The capital cost of building factories with this is very high though, as it makes use of maximum beacons and a lot of eff3 modules which are normally never worth using, but using just productivity and efficiency modules is really even worse because production speed falls to 0.5x. If there's interest, I could post BPs. Picture shows design (reds aren't tileable as is). And no, I wasn't able to add even a speed1 there anywhere because then you won't get power down to -80% which is crucial for pollution.
The secret (?) upside is that you will never ruin forests again and they will forever soak up all pollution for you.
I suppose you don't have to necessarily use productivity modules everywhere - I just want to out of principle.
Edit: Actually, I happen to have a version of the BP book with some other ones as well. (I hope it's up to date.) Take it away:
My best designs are with assemblers running at 1.4375x production speed (+5%) with +40% prod and -80% power. Furnaces run a lot faster since they only take 2 prod modules. In all cases, both the production speed and power cost reduction contribute to reducing pollution, primarily the latter. The capital cost of building factories with this is very high though, as it makes use of maximum beacons and a lot of eff3 modules which are normally never worth using, but using just productivity and efficiency modules is really even worse because production speed falls to 0.5x. If there's interest, I could post BPs. Picture shows design (reds aren't tileable as is). And no, I wasn't able to add even a speed1 there anywhere because then you won't get power down to -80% which is crucial for pollution.
The secret (?) upside is that you will never ruin forests again and they will forever soak up all pollution for you.
I suppose you don't have to necessarily use productivity modules everywhere - I just want to out of principle.
Edit: Actually, I happen to have a version of the BP book with some other ones as well. (I hope it's up to date.) Take it away:
- Attachments
-
- maxpe_a10.jpg (399.56 KiB) Viewed 3935 times