Factorio Forums

TL;DR: nuclear reactors should produce 10x the MW heat energy that they currently do, and steam turbines should be 2:1 turbine/reactor ratio



I haven't played .15 yet, in fact, I haven't played since .13 or so. I have been waiting for some of these big updates.

I watched a handful of videos on Youtube documenting the new release and its features - and I was a little disappointed with the new nuclear power.

From what I understand, one fission reactor will produce roughly 40MW of energy. That's awfully low. The smallest reactor in the United states, a single reactor, produces more than 500MW of power. The largest reactor in the United States produces ..... 4,000MW of power.

A steam engine in Factorio produces 510kW of power - which is fair - but nuclear fission has roughly 1-million times the energy density of coal and oil. ie, in line with the real-world examples I have given.

I think that these nuclear reactors are significantly underwhelming.

Not only that, but they're unreasonably expensive. For sake of argument, let's compare the raw resources cost (iron and copper plates) per MW generated, roughly.

Steam engines like to run 2:1 with boilers, and an offshore pump will provide water for a number of them. I won't include the cost of offshore pump, nor connecting pipes:

2 steam engines: 62 iron plates
1 boiler: 4 iron plates, 5 stone

total: 71 resources per 1.530MW generated.

total: 46.4 resources per 1 MW

Nuclear reactors like to run .... in a number of different ways. Since I provided reference to the smallest and largest reactors in the States, I'll do the same here, for sake of simplicity, based on the table from this thread:

1 reactor:
500 concrete
3,000 copper plates
1,000 iron plates
1,000 plastic
500 steel/2,500 iron plates

4 exchangers:
400 copper plates
10 iron plates
10 steel/50 iron plates

7 steam turbines (40MW/5.8MW):
350 copper plates
840 iron plates

total raw:

500 concrete
3,750 copper plates
1,850 iron plates
1,000 plastic
510 steel/2,550 iron plates

total that: 7610 resources per 40MW

190.25 resources per 1 MW (lol)

Now I understand that this is the worst practical efficiency, but that's where we're all going to start every new world with nuclear power. I think that's worth noting. This doesn't include all the set up on the infrastructure to get to that point, either - the mining and processing of uranium ore, which could potentially even double the cost per MW. I don't feel like doing the math for that, but that's my speculation.

so, the biggest reactor on the linked table:

25 reactors (sigh):

12,500 concrete
75,000 copper
25,000 iron
25,000 plastic
12,500 steel/62,500 iron plates

420 heat exchangers:
42,000 copper
4,200 iron
4,200 steel/21,000 iron

2,940 turbines (again 7 per heat exchanger):
147,000 copper
352,800 iron

total # of raw resources: 700,200 items

per 4,181.8MW

700,200 / 4,181.8 = 167.4 resources per 1 MW.

or, in other words, more than 4x the resource cost (realistically, more like 5x or 6x if you count everything I skipped!) per MW. and I don't feel like figuring it out 100%, but I would also bet that it takes up a lot more space using nuclear than it does solar or steam.

so where is there any incentive to use nuclear power? If these were real world numbers, we would never have seen more than a single nuclear power plant ever in history. It just wouldn't be worth it.

Now I am not too concerned about the resource cost itself - I think that's a pretty fair and balanced cost ...... if nuclear power were 10x what it is. because, then, we have:

steam: 46.4 resources / 1MW
small nuclear: 19 resources / 1MW
large nuclear: 16.7 resources / 1MW

and that's pretty reasonable, all else considered, I think. Nuclear power has been nerfed one order of magnitude of what it is relative to coal and oil, while being still 4 or 5 or 6 times the cost in resources, never mind the crafting time!

and, as for the turbines themselves .... most of the world tends to have 2 turbines per reactor (per plant!) at 1/2 load each, though the United States tends to have 1 steam turbine per reactor. Here, having 7 per reactor, turns into a really ridiculous an enormous and TEDIOUS layout that, I don't think, should be part of the reward of end game. It's tedium for the sake of tedium.

Building a factory, building a logistical nightmare with conveyor belts and inserters - that's fun, that makes you think, that's just the right level of tediosity.

Placing 2,940 of the same object for two hours? ...... why? because, even with logistics robots, one would still have to place a large number of them even to make a blueprint. I'm sorry, but it just seems like a really lazy way of trying to balance the game.



Overall, I am suggesting that nuclear reactors produce 10x the MW heat energy that they currently do, and that steam turbines should produce a proportional amount of MW with whatever losses .... to the degree where there are 2 turbines per 1 reactor. given the current cost of all this, I really don't think it's OP at all, being that this is directly proportional to the power output of real life power plants.

Thanks for reading. I hope you all will agree - and please, feel free to consider the math involved in everything I did not include; I am certain it will only drive further home the point I am making.

You're focussing on just a part of the scene : reactors are just designed to be used with neighbor bonus. With 10 reactors, you won't get 400 MW, but over 1.4 GW.
It's like saying trains are a very inefficient way of transporting stuff because there's no room in the locomotive for cargo.

Yes, that aspect was taken into consideration to the extreme - 25 reactors, 420 heat exchangers, 2,940 steam turbines. By my metric, it was still 4x the resource cost of steam power per MW.

leeknivek wrote:Yes, that aspect was taken into consideration to the extreme - 25 reactors, 420 heat exchangers, 2,940 steam turbines. By my metric, it was still 4x the resource cost of steam power per MW.

Yes, and yet somehow, people spend Lots more solar panels and use Lots more space, even though steam engines are cheaper and take up way less space.

Solar panels to trade space for lack of upkeep.
Nuclear power to trade resources for reduced space usage.

I'm not doing the math.

From what I understand, one fission reactor will produce roughly 40MW of energy. That's awfully low. The smallest reactor in the United states, a single reactor, produces more than 500MW of power. The largest reactor in the United States produces ..... 4,000MW of power.

Aside from you always wanting to use a minimum of 4 reactors together, that produces 480MW. That is close enough to 500MW to not care.
This is a makeshift reactor, made from scratch. It shouldn't be as efficient.
The setup isn't very large. Not doing the math, but again, I doubt it is much bigger overall compared to that smallest reactor.

(Do we have numbers for size of 1 tile vs real life? I assume one is about one meter, but not really sure).

leeknivek wrote:I haven't played .15 yet

If you'd actually played the game, or payed more close attention to the forums, you would see that the nuclear reactors are not in need of this massive buff. Literally two forum-posts downwards as I'm typing this, is a post called, "I have always too much uranium". For the same wattage, nuclear power is already smaller than normal steam engines, and much smaller than solar.

Yes, and yet somehow, people spend Lots more solar panels and use Lots more space, even though steam engines are cheaper and take up way less space.

Solar panels to trade space for lack of upkeep.
Nuclear power to trade resources for reduced space usage.

As I mentioned in my initial post, I would argue that nuclear uses as many tiles as steam per MW generated, and those require very, very little upkeep. Automation is the name of the game, here, after all.

I'm not doing the math

You don't have to. I did all the math relevant to my point in my initial post. Did you read all of it?

Aside from you always wanting to use a minimum of 4 reactors together, that produces 480MW. That is close enough to 500MW to not care.
This is a makeshift reactor, made from scratch. It shouldn't be as efficient.

If that is the case and applies to how a majority of people play, then I still think my suggestion about changing the ratio of turbines remains valid.



-------------------------------

I do not see the incentive for an overwhelmingly expensive power source that does roughly the same thing and requires more work. The appeal of nuclear power is its energy density, no? the fact that nuclear fission is 1,000,000 times as powerful as fossil fuels? That is the key to its very existence, why diminish that?








AileTheAlien - I have been playing this game since the 0.6 release, I think that's a fair qualification to pass judgement.
>I have always too much uranium yes, I even replied to that thread before your reply was posted here. He feels the game could do with some rebalancing on a new feature - so do I.

There is also another factor that you may not be taking into account here: Gameplay balance. All other things being equal, the reactors should offer a challenge proportionate to their benefit while also being proportionally beneficial compared to the other options, based on their ease of use and similar factors (in Factorio's case, ability to automate and uniqueness in design choices are paramount).

Whether or not these are achieved is not something you can decide with math, but with gameplay. I'll admit I've built a total of 2 reactor setups, one in sandbox mode to figure out what to do and once in a world to replace my boiler/engine setup at the time I'd normally switch to solar. However, in my opinion the nuclear setup was cleaner, smaller, easier to manage, far less ugly, and far more rewarding than boilers and engines. I had my entire factory (that was big enough to produce all the science to get nuclear power in a reasonable time, ie while I was busy doing the dozens of other things you have to do in the game) with a single reactor, and when that reactor got overtaxed adding a second one not only solved the problem, but made it so I have yet to have power problems. My world is ready to start churning out rockets and I've not half-taxed the 2-reactor setup.

So no, I don't think they're underpowered. For the work you put in they seem just fine.

I dont think size actually matter much in Factorio too a pretty large extent, the fact people build massive solar fields is proof of that. At least not unless the devs make the biters a lot more aggressive so that having to actually defend a large low density area is a problem.

I think the important thing for nuclear to be viable is for it to be easier to get enough uranium in the post 1GW factory than it is to get that much coal or oil (solid fuel) on an on-going basis with the standard or other "balanced" map settings, while still having quicker setup (or being more fun ) than clearing an area and repeating the solar blueprint enough times.

Not played much with fuel blocks yet in 0.15, but I am definitely finding nuclear easier than coal in 0.14. The challenge of getting enough coal/oil in 0.14 without constantly building mines was a reason id nearly always go to solar (along with 1 pump only doing 14 steam engines for 7MW being a pain on a random map).

If anything, nuclear is "too good". A 50K uranium ore patch, with enrichment, can power a 580 MW reactor setup for nearly 48 hours. In terms of coal, 580 MW is the equivalent of 2 blue belts full of coal! That's a challenge to keep up and store, but you can process that 50K uranium patch into a few storage chests of Uranium and not have to worry for literally days.

Okay - so I guess everyone is content with the power output of reactors, which, after thinking on the mechanic more, they do seem ok. They scale well enough to be underpowered or overpowered, as far as one wishes to take them.

My bigger issue, I suppose, is with the reactor-heat exchanger-turbine ratio, which is 1-4-7, respectively. The size and cost with the larger setups becomes an almost exponential growth. One setup I listed called for 25-420-2940 for 4,000MW. I think a 1-2-2 ratio is much more reasonable, especially if the recipes are altered to reflect the same or more resource cost. I would much rather spend more on turbines to have to place fewer of them.

This is, I feel, no different than the recent rebalance of steam engines, where 1 pump-14 boilers-10 engines became .05 pumps-1 boiler-2 engines (or 1-20-40). And it isn't so much the space, per se, but the time and tedium to fill that space with the same thing. Sure, one is almost decidedly going to have robots at the nuclear stage, but the blueprints still, for the most part, need to be hand-placed to some degree.

Modern steam driven plants - coal, oil and nuclear - all contain around the same amount of space, though nuclear produces 1,000,000x more power per volume of fuel. Solar fields are actual fields. Why should nuclear become a field of turbines?

I love these people, not trying something but KNOWING FOR SURE how overpowered / useless it is.

But OP is still not getting it, he doesn't have to try, he just knows. Because theory and math are perfect, and need no contact with the real world. On the other hand this thread is nice display of failures of such thinking, and might be good cautionary tale for others.

Tbh OP should refrain from posting unless he tries in game to compare coal based setups in GW range, solar farms in GW range and making two 2x2 nuclear reactors. Maybe he will realize simple tradeoffs all actually playing people recognized right away . . .
It's simple, really:

Initial resource cost:
Nuclear > Solar >>> Coal

Implying nuclear = bad, right? Hell no, once again, PLAYER TIME IS NOT FREE, therefore this is way more important - Logistics setup and maintenance costs:
Coal >>> Nuclear >>> Solar

However space taken directly translates into tediousness, so "fun" cost is
Solar >>> Coal ~ Nuclear

And sometimes it is important to consider pollution costs:
Coal >>> Nuclear ~ Solar
(nuclear makes some pollution from minings and refining but solars will eat your forests, reducing pollution absorption rate)

So in actual gameplay terms, solar is pretty cheap for what it does, and coal is very expensive. And since nuclear is way more fun than solar, it wins easily, because it is just a bit worse than solar. Once you figure out efficient setups and considering you need u235 and related refining for nukes anyway, nuclear just crushes everything.

But hey, nuclear is more expensive + omg realism, therefore it needs a buff. Right?

EDIT: and yeah let's just ignore how actually expensive coal power is when coal is valuable resource now because of coal liqef.