Power grid design for slug-/solar-heavy?

[myxal]

Cheers. I was wondering if there are any recommended (battle-tested) designs for power grid that is heavily reliant on power sources that cannot be automated, ie plug slugs and/or solar panels. I'm probably missing something, but the only recommendation that comes to mind is a trivial "put more batteries on your spine to take advantage of over-production".

I'm probably freaking out too much over inefficiencies, but I dislike this approach for several reasons:

• The non-controllable sources need a heavi wire (or a mess of small wires + transformers) to hook up. For panels this is just too costly for my liking (surface too far up and away from my mid-game spine), and for slugs it exposes the rancher to hideous decor on a regular basis.
• Every extra battery on the core spine adds capacity (making charging from resource-consuming generators slow) and charge run-off (wasting resources).
• Let's not forget that the capacity mismatch between jumbos and smart batteries makes detecting circuit-wide charge level hideously erroneous when the 2 types are on the same circuit.

I recall seeing "battery flipper" designs coming up in various guides - IIRC the recommendation was to use them as a replacement for spine -> consumer circuit transformers. I never used them, as they make tracking stored charge and prioritizing generators based on it a far tougher challenge than I was ready to take on.

But I wonder, if the principle can be successfully applied to harnessing the absurdly high, but short power output of plug slugs, without spamming the core spine with a dozen jumbos which my dinky power generation capacity will pointlessly struggle to recharge.

I have an idea with expanding the battery flipper to have more branches. All but 1 branch is connected to the slug/solar farm, with that last branch hooked to one (or more) big transformers, feeding onto the core spine. Flip to "the next" branch is triggered by a power meter on the transformer's input, essentially automating any number of jumbo batteries.

[Dracian]

If you plan to rely on slugs, it's not that difficult to avoid the decor penalty of HC Wires. They only need to be far enough from you dupes. IIRC, their penalty only extend to 4 tiles around them, meaning that a 8*12 farm with a HC wire running all along one tile away from the ceiling should not bother your farmer. If you can't afford HC wires, HP wires extend their decor penalty 6 tiles around them, so you'll probably have to build your farm differently (10*10 minus 4 tiles somewhere)

[BezKa]

I was planning on making a slug ethanol battery, where the power surge from them is used to produce fuel for generators that can run during the day. I haven't gotten around to that yet, but I think someone smarter than me might be able to figure it out.

[blakemw]

There are a few different strategies which can be used.

The best is making a power bank out of the rocket Batteries, these have a lot of capacity per tile and no heat or drain making them great for bulk power storage.

Plug Slugs don't need to be and shouldn't be groomed if you just want power from them, their power output since some patch relies only on being tame and fed, but being glum will reduce their food requirements by 80%, making tame-glum-fed the perfect way to harness Plug Slugs for power. Because of this you can isolate your power generating slugs from the breeder slugs, you can have them in a concentration camp and absolutely should just use Heavi-watt wire to simplify the general circuits, dupes don't need to be able to get in there anyway. So basically tame-glum-fed plug slugs on heavi-watt wire is the way to go, and you can run stuff directly on that circuit or export it using transformers.

Solar Power is kind of the opposite of Plug Slug power, often you even struggle to fill a single conductive wire on dimmer planetoids. You absolutely can just run a conductive wire to your Big Battery Bank and feed in the power with a transformer. Something else I often do is feed solar power (or manual generator power) into the SPOM to export more hydrogen, hydrogen happens to be extremely storeable, like a Gas Reservoir full of hydrogen stores as much energy as 30 Jumbo Batteries while taking a fraction of the space. Hydrogen makes a pretty good "multi-cycle" power store, like sometimes I use it to power Glass Forges and stuff which I don't run all the time but when I do run them they suck down a lot of power. Hydrogen is also really easy to prioritize using gas bridges, like say you want a Radbolt setup to have high priority power, you can basically feed the hydrogen into that setup, first going into a Gas Reservoir as a big buffer, then any overflow hydrogen can go on to powering other things. Anyway, coming back, feeding solar power into the SPOM gives you a lot more hydrogen to work with.

Of course these are more super-sustainable strategies in a normal game you kind of just use coal instead of energy storage.

[myxal]

14 hours ago, blakemw said:

rocket Batteries

I should definitely look into that, though - doesn't using these rely on a series of exploits? (Building a rocket module without an engine, and perhaps melting the platform to get that space back..) I haven't abused rocket modules yet, but there's a first time for everything... ¯\_(ツ)_/¯ 

14 hours ago, blakemw said:

Anyway, coming back, feeding solar power into the SPOM gives you a lot more hydrogen to work with.

Hmm, what exactly do you mean by this - your SPOM is powered exclusively by solar? An interesting choice, for sure - it's not like I'm struggling to find water to burn into hydrogen.

Anyway, regarding the battery flipper - this is what I had in mind:

I played around in sandbox mode where I used a signal distributor and signal counter to switch between battery rows. The "signal" being counted was a 0 watts sensor on the spine feed's input:

Spoiler

(The transformers on the left are used with pliers tool to vary the generator power available on the slug/solar rail.)

I'm looking for a lower-tech alternative to the logic so I can build this a lot earlier in a playthrough, and also to be able to scale to a different, arbitrary number of battery rows. Not to mention the flip-signal logic has some kinks to work out:

• getting stuck if it switches to a depleted row of batteries
• spazzing out when there is some power generated on the slug/solar rail, but not enough to meet the needs of the spine.

EDIT: Fixed up the logic to address the issues above, and also added OR gates to sync signals to shutoff and the distributor's address bits (Thanks, TonyAdvanced https://www.youtube.com/watch?v=1-lvvOripp0

EDIT 2: Stackable memory toggle-based design. I have not thought of a way to safely and automagically start it if all toggles end up in reset position, or kill it if more than one end up active (though that's probably less of an issue).

[psusi]

The transformer flipper is by far the best way to do a power grid.  No need for heavy wire, easy to prioritize different generators, and if you have a lot of solar/plug slugs/steam turbines, you can build a big battery bank to absorb power when all other generators are off and turn around to supply power before any other generators kick on.

 

[ToiDiaeRaRIsuOy]

Solar panels are really different from your, for instance, fossil power sources. And thus, applying the same logic concerning efficiency and optimization is not going to work.

Regarding the batteries: I like to put them high up, right into the space biome, below the solar panels. They are nicely tucked away there. I have my own system regarding this. I will show some pictures of this later when I get home. What I will always do though, is turning the gravitas facility into a base of its own from where my dupes can operate and bring in material easy for the the build.

You have to embrace the inefficiency here, bc while jumbo batteries have a lot more heat generation and power run off, the double capacity makes up for it.

You can however set up a system where a separate smart battery bank only becomes usable hen total energy in the banks drops below a certain point. See it as emergency power. Also have some fossil power generation as backup. Like 3-4 coal generators running if total energy stored falls below a certain point.

 

EDIT:

[myxal]

18 hours ago, ToiDiaeRaRIsuOy said:

You can however set up a system where a separate smart battery bank only becomes usable hen total energy in the banks drops below a certain point. See it as emergency power. Also have some fossil power generation as backup. Like 3-4 coal generators running if total energy stored falls below a certain point.

This - how is this implemented? Do you have automated transformers  with the backup batteries on the high side and main spine on the low side? Or a simple shutoff operated by a smart battery on the "main spine" side?

Either way, how do you charge this backup battery bank?

[ghkbrew]

On 8/17/2023 at 7:46 AM, myxal said:

This - how is this implemented? Do you have automated transformers  with the backup batteries on the high side and main spine on the low side? Or a simple shutoff operated by a smart battery on the "main spine" side?

Either way, how do you charge this backup battery bank?

What I've done in the past is connect a bank of dumb batteries to my power spine with separate input and output transformers.  Then control those transformers based on the charge level of smart batteries on the power spine.  E.g. when the smarter batteries are above 95% charge the battery bank.  When they're below 90% discharge the battery bank.  Connect solar (or slugs) directly to the power spine then to charge the battery bank when there's excess power.  Have your backup power sources (coal, hydrogen, etc) controlled by their own smart batteries at lower levels depending on preference (e.g. hydrogen at 85% and coal at 75%)

[Farsight]

For the slugs themselves, you can use conductive wire in place of heavy watt. You just run it backwards into a transformer in the next room. Or just put the heavy watt a few tiles away from your grooming station.

For automation of slugs/solar, I wouldn't worry about anything too complicated. The main thing I do is make sure that all my other power sources defer to the "must-use" sources like solar, like only starting them at 50% battery level. Then I just watch to see if my batteries are overfilling when all the other sources are disabled, and if that happens, I add more batteries.

Before long, I'll be using way more power than sources like solar can provide, so building a very complex system would be a waste of time. Unless you've got a massive amount of refined metals and are planning to go heavily into slugs, you'll likely outpace their power production as well.

[blakemw]

5 hours ago, Farsight said:

For the slugs themselves, you can use conductive wire in place of heavy watt. You just run it backwards into a transformer in the next room. Or just put the heavy watt a few tiles away from your grooming station.

A fed tame slug generates 1600 W (even if glum), a conductive wire can thus handle only slightly more than 1 tame plug.

It's fine to have slugs on conductive wire if there is no transformer, like a slug produces 150 W on average over a cycle, so if you had say 8 slugs every night they would put in the batteries enough power to sustain 1200 W for the whole cycle, and charging the batteries doesn't cause circuit overloads, in fact I often use this pattern.

But if you want to distribute the slug-power over multiple circuits you do need transformers (or battery flippers) and the ratio of slugs to transformers is just not favorable.

[Farsight]

3 hours ago, blakemw said:

A fed tame slug generates 1600 W (even if glum), a conductive wire can thus handle only slightly more than 1 tame plug.

It's fine to have slugs on conductive wire if there is no transformer, like a slug produces 150 W on average over a cycle, so if you had say 8 slugs every night they would put in the batteries enough power to sustain 1200 W for the whole cycle, and charging the batteries doesn't cause circuit overloads, in fact I often use this pattern.

But if you want to distribute the slug-power over multiple circuits you do need transformers (or battery flippers) and the ratio of slugs to transformers is just not favorable.

Yes, it would be one transformer per tame slug, which... seems fine to me? You spend a bit on transformers, and save some on wires and batteries.

But it's probably easiest to just run heavy-watt outside the range of the grooming station. Doing that and locking any other entries will cause pretty close to 0 decor issues.

[blakemw]

10 hours ago, Farsight said:

Yes, it would be one transformer per tame slug, which... seems fine to me? You spend a bit on transformers, and save some on wires and batteries.

But it's probably easiest to just run heavy-watt outside the range of the grooming station. Doing that and locking any other entries will cause pretty close to 0 decor issues.

No you don't groom. Tame-glum-fed all the way, full power for 20% of the food requirements. You can generate a lot of power in a small space with tame-glum-fed Plug Slugs. Like say you have a metal volcano you don't mind devoting to feeding Plug Slugs, an average metal volcano produces enough food for 15 plug slugs. Once you've bred the pop up, you can shove them in a ~40 tile room with 15 heavi-watt wires and just sweep the eggs into a separate hatching chamber (separated by a locked open door). No dupe ever has to enter it.

Given the high density it's pretty much impossible to save metal by using less heavi-watt wire and more transformers.

[myxal]

I've implemented the design above (4-row battery flipper) in a proper playthrough (Excuse the misplaced smog slug and the hideous-decor puft ranch).

Spoiler

The slug rail continues down to a room with wild slugs. I should probably move them elsewhere, they don't need duplicant attention.
The shutoffs are both on one side, so the design can expand/contract by simply cutting off/connecting batteries from the left side.

Thanks for the fed-tame-glum slug idea @blakemw, so far my breeder ranch (1 slug) is always on, and excess is left to starve. Do you meter the food to keep the slugs barely alive, or just keep it topped up with no constraint? My thinking is that 5 glum-fed slugs would consume as much as 1 breeder slug, but IIRC the starvation ranch usually has more slugs in it - the starving slugs live up to 37 cycles (27 cycles from calories + 10 starving). I'm not sure if I can trust even the calorie level thresholds from the wiki.

[blakemw]

23 hours ago, myxal said:

My thinking is that 5 glum-fed slugs would consume as much as 1 breeder slug, but IIRC the starvation ranch usually has more slugs in it - the starving slugs live up to 37 cycles (27 cycles from calories + 10 starving). I'm not sure if I can trust even the calorie level thresholds from the wiki.

Interesting question.
So yes the wiki is outdated, now it seems the rule is:

• At least 5000 kcal: 1600 W
• Between 3000 and 5000 kcal: 1200 W
• Between 0 and 3000 kcal: 800 W
• Starving: 40 W

The Slug loses 2000 kcal per cycle, which is only 400 when tame-glum.
A Slug should start adulthood with 8000 kcal then lose 400 per cycle, for the first 7.5 cycles they produce 1600 W power, for the next 5 cycles they produce 1200 W, for the next 7.5 cycles they produce 800 W, then for the remaining 10 cycles of their life they produce 40 W power.

The breeder slug produces 1600 W, it takes it 6 cycles to make a new slug, so on average it'll have 1.25x 1600 W children, 0.833x 1200 W children, 1.25x 800 W children and 1.66x 40 W children.

This means the total power produced is: 1600 + 1.25 * 1600 + 0.833 *  1200 + 1.25 * 800 + 1.66 * 40 = 5666 W.

That is less than the 8000 W of just having 5x glum-fed Slugs, while also requiring 1 more Slug to be computed. In exchange for less power you get more egg shells and meat, the five glum-fed slugs only drop an egg once every 20 cycles, compared with once every 6 cycles for the groomed slug.

23 hours ago, myxal said:

Do you meter the food to keep the slugs barely alive, or just keep it topped up with no constraint?

The power curve makes food metering tricky. Obviously there's no point in keeping slugs fully fed at below 5000 kcal, because food produces twice as much power above 5000 kcal.

There is the obvious exploit of feeding slugs only before night time and letting them starve throughout the day. This would basically require keeping slugs in individual "Slug Power Units" otherwise you can't control how much food each slug gets. Say you have a starving slug, if you feed it enough metal for 100 seconds right before night time, you get 800 W of power at night instead of 40 W. Basically compared with fully-fed, you get 1/2 as much power for 1/6th as much food, for 3x the power-to-food ratio. Slugs start adulthood with 8000 kcal so it'd be slightly more optimal to not feed the Slug for the first 20 cycles of the Slug's life, reducing food requirements by an additional 20%, the automation to detect when a new Slug is delivered to the SPU would be woolly but entirely doable, though for a mere 20% reduction in food requirement on top of an already very efficient system would make it a "stupid dupe trick" IMO.

Personally I think that ye basic "bunch of tame-glum slugs in a room" is a perfectly good compromise between simplicity, power density and efficiency, already making a metal volcano rival top-tier power producing volcanoes (like at least as much as a Volcano powering Steam Turbines and feeding Stone Hatches) and refined metal is basically trash on planetoids with metal volcanoes. But an SPU system would not be terribly difficult to engineer and lets you get 3x as much power for 6x as many slugs and hugely more space and complexity.

When I have used SPU's, I've just made 12 tile cells and I drop an egg or young slug in each cell, and let the replacement egg just hatch in the cell, so each cell contains 1 or 2 slugs and dupes never have to touch the eggs or slugs. It's really simple and would easily work with the scheme of drip-feeding food because if there are two slugs in the cell it doesn't really matter which one eats the food you'll still definitely be meeting power requirements. I've never used SPUs on a large scale, only to generate a few hundred watts for random circuits with low feeding requirements.

[myxal]

19 hours ago, blakemw said:

• At least 5000 kcal: 1600 W
• Between 0 and 5000 kcal: 800 W
• Starving: 40 W

There's at least 1 more threshold where they drop to 1200W:

Spoiler

The grooming station and the feeder in the top room were actually disabled, ignore them.

I 

I figure I'll have to dive into debug mode (metering out metal to a starving tame slug before night) to get to the bottom of this.

[blakemw]

I just found a fun fact: It seems Slugs decide what their power production will be when they get "drowsy", this is also when you have to record the kcals. I tested this by giving a starving slug a tiny amount of food right before night, but the slug was already starving again by the time it plugged in. Nevertheless, it produced 800 W.

From sandbox testing I am thinking that the 1200 W range is 3000-5000 kcal, at least to within a few hundred kcal, but Klei tends to like nice round numbers.

I edited my previous post with this new assumption, and the new expected power production for the single groomed slug and its starving descendants is increased from 5333 W to 5666 W.

Getting back to the exploit potential of the power output being calculated at drowsy time and not updated if the Slug crosses a threshold: this means the actual food requirements to have 800 W at night trends towards infinitesimal, with the main limitation probably being how quickly the Slug AI reacts to the food, as it has to react to the food before it becomes drowsy. While it takes 2 kg of food per cycle to feed a slug all night, just tiding it over until it gets drowsy could probably be easily done with 0.5 kg or perhaps even 0.1 kg, in any case easily multiplying the power-per-food several times.
 

[myxal]

10 hours ago, blakemw said:

From sandbox testing I am thinking that the 1200 W range is 3000-5000 kcal, at least to within a few hundred kcal, but Klei tends to like nice round numbers.

Those thresholds are spot on, just finished testing in sandbox myself - the only issue is that even if fed just the hour before night time, it takes a while between the moment the slug eats and when it enters Drowsy state, so a bit of extra feed is a good idea to make sure the slug is above the x-thousand threshold. 500 grams buys a duplicant-hour, which should be plenty if the conveyor meter is close to the slug room.

Re: feeding just before night-time: 4% of feed for 50% of power generations are huge savings, though the build below requires a lot of extra machinery per-slug to work. I can also imagine limiting the critter feeder, and turning on the sweeper for a limited time, for an earlier-game, no-plastic alternative. Not sure how I would retrieve meat and eggshells from the room though.

 

[blakemw]

There's an extremely simple way of metering food.

Basically Conveyor Meter Valves suck for dispensing particular amounts, but they are great at packet-splitting. So you get the conveyor meter valve to split the packets into meal sized amounts (I've been using 500 g llke you) and then just send a 1 second green signal to the Conveyor Chutes to drop a meal for each slug.

So like that setup works by having a Cycle Sensor set to 81% send a green signal. Because the Cycle Sensor is dumb and can't send a 1 second green signal, I use a "rising edge detector" (a Signal Counter set to 1, and with its output fed into the reset port) to send a green signal pulse, which is turned into a 1 second green signal using a buffer gate, thus dropping a meal in each Slug Power Unit. Each individual SPU requires only a Conveyor Chute and all the other machinery is shared, and not very much.

I kind of favor my SPU design where a young Slug or egg is dropped in each SPU, and each tame-glum slug lays one egg per lifetime, so there will be 1 or 2 slugs in each SPU, and there's zero need for egg management, the adult only gets the cramped debuff after laying the single egg it can lay in a lifetime.

FWIW I don't see any benefit at all to coming up with pre-conveyor solutions. That period of the game is short and you can easily afford to fully feed the slugs, and will probably want to anyway while breeding up a population to be tame-glum power units. I normally want to breed like 20 slugs and chances are I'd have a mechatrons by then.

[LadenSwallow]

Can slugs eat from a critter feeder that is on top of an open automatic airlock? If they can't, then having that door open during most of the day, closing just before night, with a sweeper automated to the same circuit might offer a pre-plastic solution?

[blakemw]

41 minutes ago, LadenSwallow said:

Can slugs eat from a critter feeder that is on top of an open automatic airlock? If they can't, then having that door open during most of the day, closing just before night, with a sweeper automated to the same circuit might offer a pre-plastic solution?

Yes but no because Slugs will eat more than "need" given the opportunity.

[LadenSwallow]

But you can set feeders to hold only 1kg.

[Bluefoxfire]

On 8/16/2023 at 4:43 AM, myxal said:

I should definitely look into that, though - doesn't using these rely on a series of exploits? (Building a rocket module without an engine, and perhaps melting the platform to get that space back..) I haven't abused rocket modules yet, but there's a first time for everything... ¯\_(ツ)_/¯ 

What I find works is to first build an engine on the platform (The steam and petrol engines have the largest height limit, but you can do other engines if the available space isn't sufficient), then you build the battery stack as normal. Next, you deconstruct the engine and hook up the power after.

While it is possible to also deconstruct the platform and still have the batteries, I don't know if that will end up causing issues down the road. Note that this can also be done with the cargo modules and long as the platform is intact and has the proper loaders/unloaders attached.

[myxal]

On 9/12/2023 at 7:28 PM, Bluefoxfire said:

What I find works is to first build an engine on the platform (The steam and petrol engines have the largest height limit, but you can do other engines if the available space isn't sufficient), then you build the battery stack as normal. Next, you deconstruct the engine and hook up the power after.

While it is possible to also deconstruct the platform and still have the batteries, I don't know if that will end up causing issues down the road. Note that this can also be done with the cargo modules and long as the platform is intact and has the proper loaders/unloaders attached.

I've played around with rocket modules abit since, and AFAICT once the platform is built, the game lets me build any module, no rocket engine needed (am not in sandbox or debug mode either). ¯\_(ツ)_/¯

Of note, all batteries in a stack like this are implicitly on the same circuit, without having to build wires hooking them up.

[Bluefoxfire]

9 hours ago, myxal said:

I've played around with rocket modules abit since, and AFAICT once the platform is built, the game lets me build any module, no rocket engine needed (am not in sandbox or debug mode either). ¯\_(ツ)_/¯

It'll let you, yes, but you can't put more than 2 without having an engine first, unless something changed since.