Factorio Forums

Edit based on feedback from cbhj1 where he pointed out a huge flaw in my math.


Assuming you have a 2x2 nuclear (neighbour bonus) which you want to run with a 50% duty cycle and steam storage. You need 80MW worth of fuel to power it. Each fuel cell is 8GJ so you need a fuel cell every 8GJ/80MW = 100 seconds. I'm ignoring the fact that the minimum runtime is 200s as it impacts your steam storage not ore consumed.

<edit=cbhj1>One U235 makes 10 fuel cells so you only need 1 per 1000 seconds</edit>

Ignore U238 since it is plentiful. You need one U235 every 1000 seconds from your centrifuges. To get one you need to process, on average, 10/0.007 = 1428.6 ore or let's say 1500. It takes 10 seconds to process 10 ore so 1500 seconds to process the ore to get one U235. If you want that every 1000 seconds you need to be running 1.5 centrifuges. The ore also has to be mined at a rate of 0.525 per drill so to get 1428.6 / 0.525 = 2721 so at the required rate of 100/s you need 2.8 drills.


Results for the power only

With optimal steam storage and fuel loading a 2x2 reactor on a 50% duty cycle produces 40*4*3/2 = 240MW
To feed it you need 1.5 centrifuges and 2.8 miners
Kovarex

Alongside this you'll need to get Kovarex running which takes 40 U235. Getting there takes 40*10/0.007 = 57K Uranium ore.

To feed into the Kovarex process you'll need centrifuges and miners. If you have the same number 1.5 centrifuges and 2.8 miners then you'll need 57K for Kovarex and 57K for Nuclear power.


Results

If you want 240MW and you are willing to build 3 centrifuges and 5.6 miners you can get Kovarex up and running while also generating power if your available ore exceeds 114K.

Summary

Consider starting up your powerplant before you've acquired your 40 U235 Kovarex stash. Doesn't take many drills or centrifuges to run a decent powerplant.

there are a couple points for the math.

1. you get 10 cells per U235 - 80GJ total
2. each reactor can't burn for less than 200 seconds each cycle, but they do store an immense amount of heat internally 5GJ per reactor, .5GJ per heat pipe segment or heat exchanger, both of these values are usable heat between 500 and 1000

edit:
also a 2 reactor setup will use 16GJ of fuel(2 cells) for each 200 second burn, but produce 32GJ of heat

In my first 0.15 game I used nuclear power with a few reactors before I accessed kovarex. I used 8 centrifuges to process the ore. It took a long time to get the excess U-235 to start up the enrichment, but I had plenty of other things to do to kill the time and I supplemented my power with traditional coal+steam setup. I did not bother with trying to control the reactors, I just fed them a steady supply of fuel cells and they just ran -- so no steam tanks or anything like that. Once I started the enrichment I was swimming in U-235 and made lots of bombs and there was much biter destruction.

After about 60 hours played on that map I felt I had learned enough about the new features that I decided to abandon that map and start up a new one. That first map was pure vanilla, but my second map added a couple mods including RSO. This time I just saved up 40 units of U-235 before I started up nuclear power at all. Those sat in a box for a little while until I unlocked the tech for kovarex, but I was able to jump right in as soon as that happened. Up to that point I ran on coal+steam and that was plenty sufficient.

I suspect that is the method I will use on all future maps as it feels like it is the most hassle-free for me -- just save up the 40 units before even bothering with the nuclear power. I think the trick for me is to just make sure I unlock the nuclear power early and make sure I have the acid available so I can be mining uranium ASAP. If you do that right, then you will have plenty of U-235 for both kovarex jump start as well as fuel cells for reactors. All assuming of course that you have enough uranium ore within reach.

cbhj1 wrote:there are a couple points for the math.

1. you get 10 cells per U235 - 80GJ total

Off by a factor of 10x well that won't change much damn it!

Attacking it from a different angle. Let's say you want to get Kovarex up and running in one hour. To do that you need 31 miners and 16 centrifuges. One hour later you'll have processed 57600 ore (centrifuge bottleneck) and have your 40 U235.

Simple enough. But what if you ran Nuclear at the same time? With the same number of miners and centrifuges, how much would you delay your Kovarex start and how much extra ore would it cost?

R is number of reactors, assume optimal (rectangle) layout. MW is heat. Delay is measured in seconds and % is the percentage of the delay.

Is the 0.7% 235/238 ratio showing up in everyone else's game? My ratio seems to be running closer to 0.4-0.5%. I wasn't getting enough 235 to power one reactor and two turbines (2 miners on first 10k patch. Got 6 pieces from first patch.) so I shut it down to wait to get the 40 pieces from the 135k patch I connected. I'm at 9 pieces out of 20k ore. Making me feel a little underwhelmed with nuclear so far.

AndrewIRL wrote:Attacking it from a different angle. Let's say you want to get Kovarex up and running in one hour. To do that you need 31 miners and 16 centrifuges. One hour later you'll have processed 57600 ore (centrifuge bottleneck) and have your 40 U235.

Simple enough. But what if you ran Nuclear at the same time? With the same number of miners and centrifuges, how much would you delay your Kovarex start and how much extra ore would it cost?

Code: Select all

3600 seconds, 31 miners, 16 centrifuges

R  Ore    MW   Secs  Delay %
-  -----  --   ----  ----- -
0  57152  0    3572  -28   0
1  59824  40   3739  139   4
2  62752  160  3922  322   9
3  65984  280  4124  524   15
4  69568  480  4348  748   21


7200 seconds, 16 miners, 8 centrifuges

R  Ore    MW   Secs  Delay %
-  -----  --   ----  ----- -
0  57144  0    7143  -57   0
1  62752  40   7844  644   9
2  69568  160  8696  1496  21
3  78056  280  9757  2557  36
4  88896  480  11112 3912  55

R is number of reactors, assume optimal (rectangle) layout. MW is heat. Delay is measured in seconds and % is the percentage of the delay.

math

Kovarex requires a fixed amount of 57143 ore. Each nuclear reactor (no fancy duty cycles and steam storage, just load them all) needs:
1 fuel cell per 200s
1/200 cells per second

1 U235 = 10 fuel cells
1/10 U235 = 1 fuel cell

(1/200) * (1/10) = 1/2000 U235/sec

1 U235 = 10/0.007 ore = 1428.57

So ore/sec = 1428.57/2000 = 0.714


Now solve for time = t

16 centrifuges process 1/ore per second each

(57143 + 0.714*t) / 16 = t

57143 + 0.714*t = 16*t
16*t - 0.714*t = 57143
15.286*t = 57143
t = 57143 / 15.286 = 3739 seconds

Running one reactor would delay your Kovarex start by 138 seconds if your setup was going to get you there in 3600. You've delayed Kovarex by 4% but gotten the benefits of an early switch to nuclear power.

You are assuming a 100% duty cycle. What if you use 4 reactors but only use 40MW? Or even 8 reactors? Fuel consumption goes way down.

mrvn wrote:You are assuming a 100% duty cycle. What if you use 4 reactors but only use 40MW? Or even 8 reactors? Fuel consumption goes way down.

Absolutely. Good point.

You need fancy logic and steam storage to do efficient fuel use. I assumed an early game throw-it-together style dumb reactor fuel feed. However if you use one of the many excellent blueprints to make efficient use of the fuel then you are definitely correct, my numbers are worst case. Further, as you point out, bigger plants will perform much better.

Steam storage is pretty over rated. Nuclear components will store an impressive amount of energy on their own. Everything between 500C-1000C is potential energy storage and it's a very easy bank to use.

Steam tanks are still necessary as a point for the circuit network to plug in. But it only takes 1 tank for that.

Using accumulators along with the reactor itself for energy storage instead of steam has a couple benefits:
1. zero combinator fuel regulation, extract burnt fuel at low charge, insert new fuel on burnt fuel being in the extraction hand, set insertion stack size to 1
2. reduced UPS impact from the fluid reduction

biggest downsides are energy density/size, but doesn't need to sustain demand for more than a few seconds, installation cost of accus

cbhj1 wrote:Using accumulators along with the reactor itself for energy storage instead of steam has a couple benefits:
1. zero combinator fuel regulation, extract burnt fuel at low charge, insert new fuel on burnt fuel being in the extraction hand, set insertion stack size to 1
2. reduced UPS impact from the fluid reduction

biggest downsides are energy density/size, but doesn't need to sustain demand for more than a few seconds, installation cost of accus

With accumulators I found the biggest problem is throughput. A steam tank with 16 steam turbines can deliver a lot of power at a moments notice for a short time. An accumulator is limited to 300kW. So I rather use a steam tank as trigger. Logic remains the same. Except I would always trigger at not fully charged. If you wait till low charge you don't have enough buffered to heat up the reactor again.