Some Final Words on the Power Thread

[Raisinbat]

I feel this conclusively proves that unarmored auxiliaries are a terrible idea outside of ships that absolutely need to save space or extreme glass cannons, and are well worth the weight.

Thanks for the breakdown.

Again, speaking as someone who does not have combat experience with ships this size, i just assume cruiser weapons will turn either of those setups into a fine snortable powder in <1 second. Armored is obviously better for pistols :p

If you dont mind providing an example, what kind of weapons are the cruisers equipped with? Not sure how competitive the ones i've seen were, but i'd assume a good 200-300k mass ship will pull 2-3mil base DPS easily?

 

[Matt_Bradock]

Thanks for the breakdown.

Again, speaking as someone who does not have combat experience with ships this size, i just assume cruiser weapons will turn either of those setups into a fine snortable powder in <1 second. Armored is obviously better for pistols :p

If you dont mind providing an example, what kind of weapons are the cruisers equipped with? Not sure how competitive the ones i've seen were, but i'd assume a good 200-300k mass ship will pull 2-3mil base DPS easily?

Around there, yea, or a 30-50 million alpha missile strike, depending on the preferred weapons config.

 

[Lecic]

Again, speaking as someone who does not have combat experience with ships this size, i just assume cruiser weapons will turn either of those setups into a fine snortable powder in <1 second. Armored is obviously better for pistols :p

Sure, a cruiser level weapon would definitely vaporize either under concentrated fire, but weapons will often only scratch an aux and set off the reaction that way, in which case it will survive longer. Auxes can't have multiple sets of explosions going over it at once, so it doesn't really matter if 1 block or 1000 blocks get destroyed by other weapons.

If you dont mind providing an example, what kind of weapons are the cruisers equipped with? Not sure how competitive the ones i've seen were, but i'd assume a good 200-300k mass ship will pull 2-3mil base DPS easily?

Leaving out the specifics, yes, a ship in the 200-300k range will probably have at least a few million DPS.

 

[Panpiper]

Again, speaking as someone who does not have combat experience with ships this size, i just assume cruiser weapons will turn either of those setups into a fine snortable powder in <1 second. Armored is obviously better for pistols :p

Heh. I think Raisinbat makes a very good point. I know that any ship that looses it's shields facing my battleship will in a matter of seconds have a good couple dozen perforations right through it's entire hull. (A ship with at least ten layers of fully reinforced advanced armor on it's nose might stave this off for a few seconds, assuming the nose is facing, and then maybe still slow a few more down. But that armor won't last long.) Meta weapons are more than capable of doing whatever damage is needed to make deep penetrations, and it won't just be one or two when facing off against ships large enough to carry auxiliaries.

 

[Lecic]

Heh. I think Raisinbat makes a very good point. I know that any ship that looses it's shields facing my battleship will in a matter of seconds have a good couple dozen perforations right through it's entire hull. (A ship with at least ten layers of fully reinforced advanced armor on it's nose might stave this off for a few seconds, assuming the nose is facing, and then maybe still slow a few more down. But that armor won't last long.) Meta weapons are more than capable of doing whatever damage is needed to make deep penetrations, and it won't just be one or two when facing off against ships large enough to carry auxiliaries.

Yes, but auxes at the back/core of the ship are likely not going to take nearly as much damage as whatever is at the front, and will survive longer in the battle after getting scratched by a meta gun.

 

[Tunk]

Just to add to Lecics post, a aux with space filled, or heavier armouring can survive pretty impressive damage.
I'm talking <7% damage per chain reaction cycle (or negligible <3% with heavier designs, or 0.5% in extreme cases), as well as the armour physically protecting the blocks under fire.

 

[DrTarDIS]

Let's test it.

Here's our two sets of reactors. 2000 mass for each of them- one is a pair of unarmored, 10x10x100 (10k) auxiliaries with a sheet of armor between them to prevent them from damaging eachother. The other is a single 10k (+ or - a few hundred from the connecting ribs) with a single armor layer every 2 aux layers.

What happens when I poke the first with a laser pistol?

The pair of unarmored auxes went from 20k blocks to just under 3.7k blocks. Power gen went from 3.8mil to 127k, which is about 3.3% of the original power generation.

How did the armored aux hold up?

The armored aux went from slightly over 10k aux to just under 6.1k blocks. Power gen went from just under 2mil to slightly over 900k, or 45% of its original power generation.

You would need 7 unarmored auxiliary reactors to match the after-damage power generation of a single armored aux.

I feel this conclusively proves that unarmored auxiliaries are a terrible idea outside of ships that absolutely need to save space or extreme glass cannons, and are well worth the weight.

so 2 10k reactors? Minor point on aux and maintaining groups: armored or not converting the aux into a "tube" design based on the explosion radius helps maintain a "single grouping" after meltdown.
Eg 10k is a little bit "over" the max efficiency per block to begin with.
I noticed a grouping of around 7200 aux should give fairly optimum e/block, churning out about 1.2M/sec. From that I got a 7.2 radius explosion if hit, 8 explosions, at 14400 damage.

SO at 7200 block size and an 11 diameter tube of aux, you use 40 blocks of aux per "slice" of tube, after 180 slices you hit 7200.
no one explosion can "cut off a section" as well as giving each explosion a healthy surface area of non-aux blocks in it's range.
If 180 "long" is a little much for you, you can cut it down to around 88 by having a Thicker outside wall on the tube, or 55-ish with 2 extra layers on the tube.

Whichever way you build it, it minimises the "holes" that get chopped into the aux and helps control the meltdown a bit. a bit of standard armor with some salted advanced at5-6 spaces between covering the surface area really controls the resultant touch-off.

 

[Lecic]

so 2 10k reactors? Minor point on aux and maintaining groups: armored or not converting the aux into a "tube" design based on the explosion radius helps maintain a "single grouping" after meltdown.
Eg 10k is a little bit "over" the max efficiency per block to begin with.
I noticed a grouping of around 7200 aux should give fairly optimum e/block, churning out about 1.2M/sec. From that I got a 7.2 radius explosion if hit, 8 explosions, at 14400 damage.

SO at 7200 block size and an 11 diameter tube of aux, you use 40 blocks of aux per "slice" of tube, after 180 slices you hit 7200.
no one explosion can "cut off a section" as well as giving each explosion a healthy surface area of non-aux blocks in it's range.
If 180 "long" is a little much for you, you can cut it down to around 88 by having a Thicker outside wall on the tube, or 55-ish with 2 extra layers on the tube.

Whichever way you build it, it minimises the "holes" that get chopped into the aux and helps control the meltdown a bit. a bit of standard armor with some salted advanced at5-6 spaces between covering the surface area really controls the resultant touch-off.

9.8k blocks is around the optimum of per-block generation, iirc, with it slowly dropping off after that.

 

[DrTarDIS]

9.8k blocks is around the optimum of per-block generation, iirc, with it slowly dropping off after that.

Regardless of witch discreet number, it's the tube shape that matters to what i said. controlling radius of continuous block grouping to be slightly larger than radius of explosion is the key part I was getting at.

Ed: side note: it's funny how it ends up looking like an empty roll of toilet paper.

 

[Malacodor]

Let's test it.

The pair of unarmored auxes went from 20k blocks to just under 3.7k blocks. Power gen went from 3.8mil to 127k, which is about 3.3% of the original power generation.

The armored aux went from slightly over 10k aux to just under 6.1k blocks. Power gen went from just under 2mil to slightly over 900k, or 45% of its original power generation.

You would need 7 unarmored auxiliary reactors to match the after-damage power generation of a single armored aux.

The armored reactor in your test looks far less damaged than the one we were talking about previously. This is exactly what I said, a single test isn't enough, due to the randomness of the explosions. You have to test it many times and average the results.

And why did you damage both unarmored reactors? You should have damaged only one of them, then they had >50% left instead of 3.3%. Of course two reactors are easier to hit than one, but the internal armor also increases reactor volume and therefore the chance to get hit. Also, the power of the additional reactor could be used to increase damage output, with which an opponents weapons could be disabled faster.

9.8k blocks is around the optimum of per-block generation, iirc, with it slowly dropping off after that.

Smaller reactors groups have less and smaller explosions, so the after-hit optimum is likely lower than 9.8k blocks.


Summary:
Exploding reactor usefulness depends on

• power generation when undamaged
  
• power generation after an explosion event, which depends on
  • initial block count
  • explosion damage, which in turn depends on
    
    • explosion count
    • explosion size
      
    • reactor shape
    • internal armor
    • internal hollow space
    • luck
  • possible additional weapon damage
• mid-explosion event power generation, depending on
  • explosion event duration
  • power generation before and after the event
  • luck
• chance to get hit, depending on:
  • size, due to
    • block count
    • internal armor
    • internal hollow space
  • reactor shape
  • position inside the ship
  • luck
    
  • situation, including, but not limited to
  • ship orientation
  • already suffered damage
    
  • enemy position(s)
  • enemy weapon types
  • enemy fire power, depending on
    • initial fire power
    • suffered damage to
      • weapons
      • power generation, depending on
        • goto start (aka infinite loop)

A first approximation to what might be a relatively good design shouldn't take more than 10^10 to 10^20 tests. Happy testing.

 

[Tunk]

Do the tests yourself Malacodor and stop theory crafting.

 

[Lecic]

The armored reactor in your test looks far less damaged than the one we were talking about previously. This is exactly what I said, a single test isn't enough, due to the randomness of the explosions.

The armored reactor we were originally talking about only has a main body 4 blocks thick. The one I built is 10 blocks thick. That is why more of it is still connected. I would suggest testing things yourself if you want to disagree with my data.

Of course two reactors are easier to hit than one, but the internal armor also increases reactor volume and therefore the chance to get hit.

Internal armor also reduces the chance that a hit to the reactor will actually overheat, because it could just hit armor. Furthermore, while this heavily armored design is 1.5x the volume of an unarmored one, this is also one of the most heavily armored versions. Less armored versions that still provide a decent amount of power generation after damage with significantly less volume than this one still exist.

 

[Malacodor]

Do the tests yourself Malacodor and stop theory crafting.

A first approximation to what might be a relatively good design shouldn't take more than 10^10 to 10^20 tests. Happy testing.

Did you even read my post?

The armored reactor we were originally talking about only has a main body 4 blocks thick. The one I built is 10 blocks thick. That is why more of it is still connected. I would suggest testing things yourself if you want to disagree with my data.

So you "disproved" what I said about one design, by making a test with a different design?
Look, I don't really care which one of two or three designs is better. My point is that I could easily come up with a hundred different shapes, each with hundreds of variants, and each one has its own more or less situational advantages and disadvantages. Even thinking of how I could test all those makes me want to hurt somebody. And even if I did, the next patch could make it all obsolete and I had to start over.

 

[refirendum]

so i can see that people don't like heat boxes as a thing replacing power gen

why not keep power generation and capacity and add heat? is that also something people don't want?

i'm totally not responding much to more recent discussion in this thread; i'm just giving my take on the concept here.

Energy and its associated factors/interactions

energy generation:
- energy generation rate (linear)
- heat generation rate multiplier per block (non linear exponential* [base+e/s])
- heat dissipation rate (linear)
- losses in power of damaged reactors

energy consumption:
- output of all blocks of system type (linear)
- power consumption efficiency multiplier (nonlinear exp * [base+output/s])
- insufficient power efficiency penalty

control
- user-set limits on outputs of systems
- user-set limits on power generation

Basically what i'd like to see first is that energy generation scales linearly with block count while heat generation has some significant base heat generation that is added to the e/s before multiplying the sum by an increasing exponential. This ensures that small single reactors would generally be governed by relatively easy to manage base heat factors while larger reactors are more governed by more difficult to manage block count heat factors. This also ensures that when comparing polyreactor power generation to monoreactor power generation, polyreactors output significantly more heat. This is, I feel, a critical detail of power generation to both prevent massive arrays of micro reactors, and lend an intuitive quality to power generation considerations when designing a ship. Heat dissipation can be linear or non-linear per block, though linear will feel much easier on players in general, and likely is easier to program. Losses of power generation in damaged reactors must be present in a palpable way, maybe not 100 percent power generation loss, but a severe drop nonetheless is necessary to the dynamic between polyreactor vs monoreactor design philosophies.

This dynamic means that ships using polyreactor designs have done so with the knowledge that in exchange for their design's combat resilience, they had to compromise on power generation, heat generation, volume commitment to other systems due to heat dissipation requirements, or some combination of these. On the flip side, ships that were designed with larger mono-reactor systems have done so with the knowledge that in exchange for power efficiency, they have sacrificed valuable combat resilience. The balance between the design philosophies here is controlled by the factor of power loss upon damage.

The other part of an energy ecosystem is consumption. For each type of system, it's block count should determine its theoretical total output, whether that is thrust or damage or shielding or anything else. Power consumption would then use a significant base value added to the output/s of the group all multiplied by an increasing exponential much like power generation.. This would ensure that as weapon size increases, its total inefficiency also increases, and also that single group weapons have a significantly better efficiency than multi-group weapons of the same total damage output (also applies to basically every other power-consuming system)

Next, the over-drawing of power that leads to insufficient power in systems should mean that the systems exhibit a loss in output that is equal to the fraction (ship total power generation)/(ship total power consumption including efficiency multipliers). This will lead to the behavior that overdrawn ships will output exponentially less power as the size of the systems overdrawing the reactor increase. It places pressure on ship designs to have power generation or capacity leeway/headroom rather than letting ships fall into power shortage.

Lastly, control of systems is important. Being able to quickly and simply set/adjust your own limits to prevent power generation, weapon damage, thrust, or other systems from operating at 100% is important to the ecosystem. A reactor array that produces too much heat at 100% output can be adjusted down to a level in which it doesn't at some loss of energy generation, or a weapon that draws too much energy when thrusters are not limited can be supplied with enough energy by limiting thruster output to lower power consumption. This is important to design, because of flexibility. A user should not be forced to design a ship that has a reactor large enough to supply all systems of their ship at all times when all systems are operating at 100%, but instead, they should be given the option of limits to let them build ships that are capable of more than one role. Another scenario: limiting power generation/ship output during combat upon damage that reduces heat dissipation capability to maintain function. The point is that these systems should have control. similar to thrust allocation, but more easily accessible by hotkey or modifier. Using a power use indicating bar as seen earlier in this thread as a tool, one could, based on sight, easily adjust limit sliders so that the total power consumption lines up with power generation.

this is probably going to have some issues in the explanation, or edge cases that I haven't thought about, but in concept, i feel that it is my ideal solution to this whole power thing.

 

[Tunk]

Personally I am down for a heat on consumption mechanic to force ships to cool down after extended power consumption, in addition to a conduit mechanic.
Powered systems require at least a powered conduit connection (or bridged via powered docker/rail) in order to receive power from a generation system.
A actual power grid, basically.

 

[Lecic]

So you "disproved" what I said about one design, by making a test with a different design?

It's literally the exact same design, just proper reactor thickness instead of a thin slice of it made for demonstration purposes. Stop acting dense.
[doublepost=1487808561,1487808153][/doublepost]So I've had a thought- what if we added heat on top of power generation, and faster firing weapons generated more heat than slower firing weapons? This would make large weapons more heat effective. Small ships could be more effective at dissipating heat, though, so they could still use rapid fire weapons fine.

 

[refirendum]

Personally I am down for a heat on consumption mechanic to force ships to cool down after extended power consumption, in addition to a conduit mechanic.
Powered systems require at least a powered conduit connection (or bridged via powered docker/rail) in order to receive power from a generation system.
A actual power grid, basically.

okay, think about it this way. starmade is like some similar-to-our-own alternate universe, set far into the future where space travel is super mundane and ppl jump like 10 ft high. (lol) in our own present day, we already have solutions to wirelessly transmit electricity. and consumer electronics have been rapidly adopting wireless charging and such. In a future advanced universe like starmade, they will probably have far surpassed this low-output wireless power transmission and gotten to high output power transmission. Maybe wired power transmission is still technically better, and as such you can have ships that don't use power wires get a slight power consumption bump to simulate the losses through wireless power. Requiring power lines is not a positive thing for the freedom of gameplay/ship design that i feel starmade would benefit from. This ensures that floating ships still have a method to functioning, and broken power lines doesn't mean that you lose all functions connected to those lines even if you do get a bit of a penalty on power consumption.

heat, however, should probably be manageable by use of heat sink blocks. maybe instead of a blocks linear relation like i presented, maybe it could be linear based on surface area of heatsink blocks -open faces of the blocks that have no directly adjacent blocks covering the sides- and all of the heatsink blocks might have to at least touch the reactor/heatbox or a group that makes contact with the reactor or heatbox.

 

[Haalon]

so i can see that people don't like heat boxes as a thing replacing power gen

why not keep power generation and capacity and add heat? is that also something people don't want?

i'm totally not responding much to more recent discussion in this thread; i'm just giving my take on the concept here.

Energy and its associated factors/interactions

energy generation:
- energy generation rate (linear)
- heat generation rate multiplier per block (non linear exponential* [base+e/s])
- heat dissipation rate (linear)
- losses in power of damaged reactors

energy consumption:
- output of all blocks of system type (linear)
- power consumption efficiency multiplier (nonlinear exp * [base+output/s])
- insufficient power efficiency penalty

control
- user-set limits on outputs of systems
- user-set limits on power generation

Basically what i'd like to see first is that energy generation scales linearly with block count while heat generation has some significant base heat generation that is added to the e/s before multiplying the sum by an increasing exponential. This ensures that small single reactors would generally be governed by relatively easy to manage base heat factors while larger reactors are more governed by more difficult to manage block count heat factors. This also ensures that when comparing polyreactor power generation to monoreactor power generation, polyreactors output significantly more heat. This is, I feel, a critical detail of power generation to both prevent massive arrays of micro reactors, and lend an intuitive quality to power generation considerations when designing a ship. Heat dissipation can be linear or non-linear per block, though linear will feel much easier on players in general, and likely is easier to program. Losses of power generation in damaged reactors must be present in a palpable way, maybe not 100 percent power generation loss, but a severe drop nonetheless is necessary to the dynamic between polyreactor vs monoreactor design philosophies.

This dynamic means that ships using polyreactor designs have done so with the knowledge that in exchange for their design's combat resilience, they had to compromise on power generation, heat generation, volume commitment to other systems due to heat dissipation requirements, or some combination of these. On the flip side, ships that were designed with larger mono-reactor systems have done so with the knowledge that in exchange for power efficiency, they have sacrificed valuable combat resilience. The balance between the design philosophies here is controlled by the factor of power loss upon damage.

The other part of an energy ecosystem is consumption. For each type of system, it's block count should determine its theoretical total output, whether that is thrust or damage or shielding or anything else. Power consumption would then use a significant base value added to the output/s of the group all multiplied by an increasing exponential much like power generation.. This would ensure that as weapon size increases, its total inefficiency also increases, and also that single group weapons have a significantly better efficiency than multi-group weapons of the same total damage output (also applies to basically every other power-consuming system)

Next, the over-drawing of power that leads to insufficient power in systems should mean that the systems exhibit a loss in output that is equal to the fraction (ship total power generation)/(ship total power consumption including efficiency multipliers). This will lead to the behavior that overdrawn ships will output exponentially less power as the size of the systems overdrawing the reactor increase. It places pressure on ship designs to have power generation or capacity leeway/headroom rather than letting ships fall into power shortage.

Lastly, control of systems is important. Being able to quickly and simply set/adjust your own limits to prevent power generation, weapon damage, thrust, or other systems from operating at 100% is important to the ecosystem. A reactor array that produces too much heat at 100% output can be adjusted down to a level in which it doesn't at some loss of energy generation, or a weapon that draws too much energy when thrusters are not limited can be supplied with enough energy by limiting thruster output to lower power consumption. This is important to design, because of flexibility. A user should not be forced to design a ship that has a reactor large enough to supply all systems of their ship at all times when all systems are operating at 100%, but instead, they should be given the option of limits to let them build ships that are capable of more than one role. Another scenario: limiting power generation/ship output during combat upon damage that reduces heat dissipation capability to maintain function. The point is that these systems should have control. similar to thrust allocation, but more easily accessible by hotkey or modifier. Using a power use indicating bar as seen earlier in this thread as a tool, one could, based on sight, easily adjust limit sliders so that the total power consumption lines up with power generation.

this is probably going to have some issues in the explanation, or edge cases that I haven't thought about, but in concept, i feel that it is my ideal solution to this whole power thing.

Well, your idea is very close to my proposed one. However, in some cases I took opposite ways of dealing with issues.

I ll start with energy interactions too.

energy generation:
- (potential) energy generation rate per block (exponential)
- (potential) energy capacity rate per block (exponential)
- (potential) heat generation rate per block (exponential, MUCH more rapid than energy generation rate)
- heat dissipation rate (~ linear)
- losses in power of damaged reactors

energy consumption:
- output of a system per energy used (linear)
- heat generation of a system per energy used (exponential)
- efficiency of power usage per system block count (logarithmic)
- heat generation of a system per block count (asymptotic and decreasing)

control
- user-set amounts of power supplied to a system, which can be changed at any time.


(potential) is there, because in my solution layout of a reactor system GREATLY affects its stats, even more than block count does.

So, I solved monoreactor vs polyreactor issue differently. One big reactor potentially generates MUCH more energy than several small ones.
In terms of power generation, it's MUCH better to have one or two big reactors, than 10 or 20 small ones with the same total volume.
However, big and power-efficient reactors will produce much more heat than several small ones. This not only solves monoreactor vs polyreactor issue, it also prevents monoreactor system from being, literally, too overpowered. Because heat generation of efficient reactor rises greatly with its size. And heat distortion doesn't rise that fast. Moreover, cooling systems can be volatile in some cases.

As for power consumption. Player can choose at any time how much energy he want for a system to use. More energy for a system means better output, while much higher heat generation. Logarithmic dependency of system output on block count means that size of a system doesn't affect output much. What does it affect much - is a heat generation. I ll better quote myself to explain

Ok. For example we have weapon group of 100 blocks. All values below are for demonstration.

• If we supply to it 1 000 000+ energy, it wont even shoot, but will generate INSANE amounts of heat and shut down. This will limit capacity-based suicide ships.
• If we supply from 100 000 to 1 000 000 energy - the weapon will shoot once with very high damage, generate extreme amounts of heat, and shut down for some time or until reboot. Shutdown time goes from 0 sec at 100 000 power, to the max value at 1 000 000 power.
• If we supply from 10 000 to 100 000 energy - there will be no shutdown, damage output is great, but heat generation is still high. So big ships with well-built power and cooling systems could use this overcharged weapons for large amounts of time or even non-stop. And small ships can use such overcharge for single-strike weapons.
• If we supply from 1 to 10 000 energy - weapon will generate amounts of heat manageble for inherent cooling, so small ships can shoot for a long amount of time or non-stop.

So, this makes block count of a power-consuming system less valuable, while still very important - its block count determines how much energy can this system use without turning your ship into a melting pot.

As for control energy ecosystem. Or power distribution, in my case. I want to allow player a control of a ship's power consumption. He will be able to choose how much energy to supply to any of his system - thus he will be able enhance any of his system(or systems) while sacrificing the others. So, for example in a battle, you will be able to sacrifice thrust and some of your weapons power, in order to make your shields and other weapons stronger.

 

[Captain_Boroski]

So I've had a thought- what if we added heat on top of power generation, and faster firing weapons generated more heat than slower firing weapons? This would make large weapons more heat effective. Small ships could be more effective at dissipating heat, though, so they could still use rapid fire weapons fine.

Been thinking of the same, throw in thrusters generating heat and you can have the basics down for having a thermal-sensor system that'll detect thermal sources closer depending on how many thermal-sensor blocks you have (like scanners) and how strong the thermal signature is.

Make the thermal sensors capable of detecting the heat of cloaked ships and vola, more depth for the game.

 

[refirendum]

Well, your idea is very close to my proposed one. However, in some cases I took opposite ways of dealing with issues.

I ll start with energy interactions too.

energy generation:
- (potential) energy generation rate per block (exponential)
- (potential) energy capacity rate per block (exponential)
- (potential) heat generation rate per block (exponential, MUCH more rapid than energy generation rate)
- heat dissipation rate (~ linear)
- losses in power of damaged reactors

energy consumption:
- output of a system per energy used (linear)
- heat generation of a system per energy used (exponential)
- efficiency of power usage per system block count (logarithmic)
- heat generation of a system per block count (asymptotic and decreasing)

control
- user-set amounts of power supplied to a system, which can be changed at any time.


(potential) is there, because in my solution layout of a reactor system GREATLY affects its stats, even more than block count does.

So, I solved monoreactor vs polyreactor issue differently. One big reactor potentially generates MUCH more energy than several small ones.
In terms of power generation, it's MUCH better to have one or two big reactors, than 10 or 20 small ones with the same total volume.
However, big and power-efficient reactors will produce much more heat than several small ones. This not only solves monoreactor vs polyreactor issue, it also prevents monoreactor system from being, literally, too overpowered. Because heat generation of efficient reactor rises greatly with its size. And heat distortion doesn't rise that fast. Moreover, cooling systems can be volatile in some cases.

As for power consumption. Player can choose at any time how much energy he want for a system to use. More energy for a system means better output, while much higher heat generation. Logarithmic dependency of system output on block count means that size of a system doesn't affect output much. What does it affect much - is a heat generation. I ll better quote myself to explain

So, this makes block count of a power-consuming system less valuable, while still very important - its block count determines how much energy can this system use without turning your ship into a melting pot.

As for control energy ecosystem. Or power distribution, in my case. I want to allow player a control of a ship's power consumption. He will be able to choose how much energy to supply to any of his system - thus he will be able enhance any of his system(or systems) while sacrificing the others. So, for example in a battle, you will be able to sacrifice thrust and some of your weapons power, in order to make your shields and other weapons stronger.

basically the heat generation vs power generation dynamic between polyreactor and monoreactor is solved similarly, but instead of making heat the factor that makes polyreactor systems less efficient, instead you use energy generation scaling at high volumes. my version is based on my current understanding of how energy efficiency works irl, where the many small engines (engines in automobiles) being replaced by electric motors diverts the power usage to the power grid, whose energy is generated by large singular power plants. this is what gives electric cars operational efficiency vs fossil fuel vehicles, because it pulls its energy from a more efficient producer of energy that could still use fossil fuel rather than a less efficient mobile fossil fuel motor.

power generation, i feel, is a fairly linear thing for the most part irl. joules of energy being extracted from reactions scale linearly with the mass of reactants for chemically produced energy or rate of fission events in the case of nuclear energy. heat and useable electricity when converted to the same form of energy will add up linearly. so i suppose we could change the heat-e/s generation to a changing fraction, where at low total e/s the portion of that potential energy that is turned into heat is relatively low (plus some baseline which dominates the dynamic until you reach maybe 1.5 million or so e/s) and at high potential e/s, that proportion grows large enough to start controlling the heat rate.

with this re-thinking of the mechanic, maybe later, blocks can be added to attempt to reclaim heat to extract a little more power (with fairly severe diminishing returns).

maybe we can set some theoretical maximum efficiency % for energy generation vs heat generation in addition to the baseline, and then the reactor's intrinsic efficiency takes that theoretical % as its maximum percentage. It would then begin to drop after the "softcap" so to speak, and then the only way at those sizes to properly extract max efficiency is to use the heat reclamation mechanic to reclaim maybe up to 20% of the lost heat per full reclamation stage up to maybe 2 or 3 groups before the bonuses drop to 0.

so let's say a large reactor has an intrinsic efficiency of 80% and the theoretical maximum was set to 80%. then the actual efficiency of the reactor would be 64% with some linearly scaling additional heat amount. then let's say the player added 2 fully filled out reclamation stages, which gives back 20% of the missing 16%, or 3.2%, and then an additional 20% of the remaining 12.8%, or ~2.5%. then with the two reclamation stages, this reactor would have an efficiency of 69.7% instead of 64% which would mean 69.7% of the block's energy is generated into electricity, while 30.3% of its energy is generated into heat.

basically an optional addition of complexity that gives a small bonus, but one can still build without. you could link it like the heat sink systems and maybe they would have to be outside the heat affect box to function best.

then the framework would mean that each block has a linear energy value per second with some linear heat value (to balance poly vs monoreactor) which then is used in the efficiency calculation to determine how much of that is going to be heat and how much is going to be energy.

the controls for limiting power then just adjust this block energy value down from its maximum per block and the rest follows.

though if they don't want that many calculations, then meh. but it would all be calculated together and not adjusted downwards upon damage until reeboot, similarly to how HP and AHP and sysHP is calculated now.