160MW Nuclear Setup, no waste w/ minimal storage tanks
Posted: Thu May 25, 2017 4:17 pm
by BlakeMW
This is a fairly compact 160MW nuclear setup that doesn't use an excessive amount of storage tanks. Under normal usage it can easily buffer a complete fuel cycle - perhaps this post is more intended as a demonstration that heaps of steam tanks are not required as it is basically a dumb reactor which has been upgraded with circuit network to make it smart, and this works out well: see the maths below.
The setup has requester chests, but you could just as easily use a belt as nothing is wired to the chests.
Each heat pipe and heat exchanger can buffer 1MJ of energy per degree and reactors can buffer 10MJ per degree, from 500 to 1000 degrees.
This setup has 2 Reactors, 16 exchangers and 24 heat pipes, allowing it to theoretically buffer 30GJ of heat.
In addition, the two steam tanks can store 4.8GJ of heat in the form of steam.
The two reactors create 160MW of heat for 200 seconds, a total of 32GJ of heat per fuel cycle.
As such this basic setup can buffer 34.8MJ of heat, while a fuel cycle generates 32GJ of heat. Now - a smart person would point out that not all that theoretical buffer is available, because in principle the heat pipes furthest from the reactors could be at, say 900 degrees, while the reactors are at 1000 degrees and "venting" heat and as such the full range from 500 to 1000 degrees is not available. Note though, that if for example your factory is using 32MW, then only 80% of the fuel cycle actually needs to be buffered because the other 20% is consumed immediately. In testing, I've found this setup does not waste heat at loadings as low as 10% (16MW) and even if it were to be used at even lower loading it wouldn't waste more than a few percent of the fuel and would consume much less fuel than a setup that inserts fuel whether required or not (for example say you run it at 1% load, it might consume 2% more fuel than if there was enough heat capacity to definitely buffer a full fuel cycle: but it'll still use 1/100th the fuel as a dumb reactor would at 1% load).
Circuit Controller
When using storage tanks only for steam measuring and not storage it is important to make sure the reactors are on when the steam starts dropping (actually there's about 10s grace, but you can't just check if steam is low every 200s). So this setup uses a condition which can be summarized as: "Insert fuel IF steam is depleting AND at least 200s has passed since last fuel insertion", under full load fuel will be inserted every 200s like clockwork and under less than full load the fuel will be inserted once the system starts cooling down.
A combinator takes the steam level and outputs T=1 on the red wire if the tanks are less than 80% full.
Another combinator works as a counter, which counts up to T=1 then halts. (This counter is on the green wire, and receives T=1 from the constant combinator)
The fuel inserters are set to read hand contents, when they pick up a fuel cell they output a fuel=1 pulse, which is multiplied by -6000 and fed as T into the timer, causing it to count up from -12000T to T = 1.
The fuel inserters are linked to both the red and green wires and are set to only insert on T = 2, which is only true if the counter has halted at T=1, AND the fuel level is low.
This setup does not have much allowance for a partial fuel outage - that is it won't wait until fuel is available for both reactors before inserting. I work on the assumption that you'll have some kind of fuel availability monitor which makes a godawful noise using a speaker if the fuel is going to run out. Also, even if it starts in an unbalanced state it will synchronize fuel insertions in the future so no special startup procedure is required, throw fuel into the chests or reactors or whatevs, it doesn't matter.
I believe this principle (essentially upgrading a dumb reactor with a steam measuring tank and some smart circuitry) would scale up quite well. Reactors in long 2x rows do put out as much as twice as much energy per fuel cycle, though they also need twice as many heat pipes and heat exchangers which largely balances that out - the only which doesn't is the reactors still having the same heat capacity. I believe that (extra) steam storage tanks should generally not be required for larger nuclear setups under reasonable loading (i.e. at least 30% base load). Using double rows of heat pipes instead of a single row would also be a way to provide ample heat buffering, and it's a nice compact solution as heat pipes provide twice the heat storage density as steam tanks (3x3 heat pipes = 5GJ, 1 steam tank = 2.4GJ). Ultimately whether or not a reactor will successfully buffer a full heat cycle at any load is not a problem which can be solved simply by maths, you need to run the reactor at a given load and see if the reactors reach 1000 degrees - but my experience is the heat transfer is usually fast enough that under reasonable loads no heat is wasted.
Re: 160MW Nuclear Setup, no waste w/ minimal storage tanks
Posted: Wed Aug 01, 2018 8:31 pm
by JackGruff
I've been using this for about 60+ hours in a game, seems to work very well.
Could you please make a 4 reactor version?
Re: 160MW Nuclear Setup, no waste w/ minimal storage tanks
Posted: Thu Aug 02, 2018 10:30 am
by zOldBulldog
This seems like the perfect "initial nuclear" setup, compact and easy to build.
I usually make a 2 reactor setup first, then scale up and build an 8 reactor (no real reason, it just seems to work for me, and I am very ignorant in the details of efficient nuclear design). This might be ideal for my first setup.
A couple of questions:
- Does this design conserve fuel when not needed? Or are more steam tanks essential for that?
- What happens when demand drops below the 30% level (around 50MW) you warn about? Just a brownout that can be easily covered by backup steam/solar power and then it reengages when demand goes up? Or does it shut down and require a manual restart?
- Speaking of that, is there a manual start sequence, or does this design start on its own?
Re: 160MW Nuclear Setup, no waste w/ minimal storage tanks
Posted: Thu Aug 02, 2018 11:47 am
by mrvn
zOldBulldog wrote:- What happens when demand drops below the 30% level (around 50MW) you warn about?
Reactors don't stop when there is no demand and simply heat up to 1000°C and then waste the fuel. So to be efficient your reactor design has to draw out all the heat each fuel cell creates and store it either in heat pipes, in steam tanks or use it. If you assume a minimum 30% load on the reactor that means you don't need to provide heat/steam storage for those 30%. Means you can build the reactor smaller.
But if you don't use those 30% then the reactor simply reaches 1000°C and waste the fuel. Nothing should break down ever due to that.
Re: 160MW Nuclear Setup, no waste w/ minimal storage tanks
Posted: Thu Aug 02, 2018 12:20 pm
by zOldBulldog
mrvn wrote:
zOldBulldog wrote:- What happens when demand drops below the 30% level (around 50MW) you warn about?
Reactors don't stop when there is no demand and simply heat up to 1000°C and then waste the fuel. So to be efficient your reactor design has to draw out all the heat each fuel cell creates and store it either in heat pipes, in steam tanks or use it. If you assume a minimum 30% load on the reactor that means you don't need to provide heat/steam storage for those 30%. Means you can build the reactor smaller.
But if you don't use those 30% then the reactor simply reaches 1000°C and waste the fuel. Nothing should break down ever due to that.
So, nothing wrong then to using this design for an early plant that gets replaced soon after, except for some fuel waste. Just go to a design that "fully converts unused fuel to stored steam and then shuts down until the steam is consumed before restarting" when upgrading to the larger design. Correct?
Re: 160MW Nuclear Setup, no waste w/ minimal storage tanks
Posted: Thu Aug 02, 2018 1:22 pm
by BlakeMW
zOldBulldog wrote:
- Does this design conserve fuel when not needed? Or are more steam tanks essential for that?
It conserves fuel by exploiting the very large heat capacity (and perfect insulation!) of nuclear reactor components. Basically, the setup is allowed to cool down to nearly 500 degrees, then fuel is inserted which heats it up to nearly 1000 degrees, the stored heat runs the heat exchangers until it cools back down to nearly 500 degrees and that cycle repeats (note that reactor setups never lose heat to the environment, the only way it cool down is heat exchangers removing heat to make steam, thus heat storage enjoys perfect effeciency).
- What happens when demand drops below the 30% level (around 50MW) you warn about? Just a brownout that can be easily covered by backup steam/solar power and then it reengages when demand goes up? Or does it shut down and require a manual restart?
During a fuel burning cycle the reactors will sometimes reach 1000 degrees causing a small portion of the fuel to be wasted as there is nowhere for the heat to go. There is no danger of brownouts with this setup regardless of what you do. That was a deliberate design decision I made: I decided it was more important to avoid any possibility of a brownout than any possibility of wasting fuel.
However with this 2 reactor design, it's nearly impossible to get it to actually waste fuel (probably could by running it at 100% load, then abruptly cutting off all the load, but not under any realistic operating conditions). Scaling it up to 4+ reactors makes it more possible to waste fuel when run at low loads, though it still wastes very little compared with a logic-less reactor.
- Speaking of that, is there a manual start sequence, or does this design start on its own?
It's designed to be placable by blueprint and starts itself up automatically once there are fuel cells in the requester chests. But note that if only one chest has a fuel cell, only one fuel cell will be inserted and for that fuel cycle only one reactor will be running. Typically this shouldn't be a problem because it'll correct itself on the next fuel cycle 200s later, but if it is a problem just manually insert 1 fuel cell into the empty reactor - or put fuel into both chests before hooking it up to the power grid. The setup is highly resistant to operator error, if you randomly shove fuel cells into reactors it won't cause any serious problems.
Re: 160MW Nuclear Setup, no waste w/ minimal storage tanks
Posted: Thu Aug 02, 2018 1:47 pm
by mrvn
You can also just add a few extra heat pipes or steam tanks (without wires) to get more buffering. Heat pipes should also allow you to extend this to 4 reactors without extra heat exchangers or turbines. They just have to be able to buffer 3 times as much heat. Makes it 3 times as fuel efficient.
Re: 160MW Nuclear Setup, no waste w/ minimal storage tanks
Posted: Thu Aug 02, 2018 1:56 pm
by BlakeMW
mrvn wrote:You can also just add a few extra heat pipes or steam tanks (without wires) to get more buffering. Heat pipes should also allow you to extend this to 4 reactors without extra heat exchangers or turbines. They just have to be able to buffer 3 times as much heat. Makes it 3 times as fuel efficient.
Yeah in the OP I mention that using double rows of heat pipes would be an effective way of scaling up the heat storage, and that heat pipes have about twice the energy storage density (per tile) as steam tanks.
Re: 160MW Nuclear Setup, no waste w/ minimal storage tanks
Posted: Thu Aug 02, 2018 3:18 pm
by mrvn
A note of using nuclear power with solar cells:
With solar cells your demand switches rather drastically from 0% to 100% in the extreme case. Stored steam can be used instantly to make electricity. Stored heat can be used to make more steam instantly too. But throwing in more fuel into the reactor takes a long time to heat up the reactor and heat pipes. Heat only slowly crawls along the heat pipe activating one heat exchanger after another. So you want to throw in fuel early, basically as soon as the steam tank drops from full. And in case that happens right when the sun is rising you need enough heat storage to store a full fuel cells worth of heat.
So maybe not just a double row of heat exchangers. How about 6 rows so the heat pipe connects to 4 points (2 each) on the reactors?
Re: 160MW Nuclear Setup, no waste w/ minimal storage tanks
Posted: Fri Aug 03, 2018 8:49 am
by BlakeMW
mrvn wrote:
So maybe not just a double row of heat exchangers. How about 6 rows so the heat pipe connects to 4 points (2 each) on the reactors?
In practice I've found that once the entire setup is at at least 500 degrees, heat spreads very quickly across it, so the steam stored in the measuring tanks, turbines and pipes tends to be enough.
But of course if you were serious about night-nuclear you'd want to use a higher ratio of steam turbines to reactors since the turbines only have to run 30% of the time while the reactor can run all the time. The inability to directly measure heat would make it difficult to guarantee the whole setup is piping hot in preparation for the night - short of always inserting fuel. For night nuclear I'd probably go with enough steam tanks to last the night (to provide the extra turbines with steam), the combinator logic from this thread will work perfectly since it's designed to maintain steam tanks in a full state.
Re: 160MW Nuclear Setup, no waste w/ minimal storage tanks
Posted: Fri Aug 03, 2018 10:32 am
by BlakeMW
JackGruff wrote:I've been using this for about 60+ hours in a game, seems to work very well.
The pipe in the middle are to exchange steam, which is necessary because each heat exchanger is not really used by its direct turbine. It also allows you reclaims unused steam.
All pipes around this setup are useless, it was just make a test on my existing installation (160 pipes less).
This is a 480MW perfect math setup if I remember well :
4 reactors means 200% bonus each
So 40MW +80MW bonus for each reactor : 480MW
So you need 82.75 turbines (so 83) => I used 84 so it's pretty
and 48 heat exchangers !
You can place buffers at each corners and between left and right part of the setup.
I have a perfect 8 core setup with safety for pumps (separate electric network) if someone wants it. Pretty, but a bit massive.
Re: 160MW Nuclear Setup, no waste w/ minimal storage tanks
Posted: Mon Jan 28, 2019 11:24 am
by mrvn
The steam tanks seem badly placed with only one pipe leading in and out of them. When combing nuclear with solar power you could need twice the throughput from the tank compared to the heat exchangers.
While not designed for a full night-only nuclear setup I would suggest this: Place the first row of turbines directly after the heat exchangers. Only then have the row of pipes. The places where there is no second turbine can then be filled by a tank.
Note: For a full night-only nuclear setup you would need ~3 times the turbines and who knows how many tanks in a 4 turbines, 2 tanks, 2 turbines setup.
Re: 160MW Nuclear Setup, no waste w/ minimal storage tanks
Posted: Tue Jan 29, 2019 11:36 pm
by nafira
mrvn wrote: Mon Jan 28, 2019 11:24 am
The steam tanks seem badly placed with only one pipe leading in and out of them. When combing nuclear with solar power you could need twice the throughput from the tank compared to the heat exchangers.
While not designed for a full night-only nuclear setup I would suggest this: Place the first row of turbines directly after the heat exchangers. Only then have the row of pipes. The places where there is no second turbine can then be filled by a tank.
Note: For a full night-only nuclear setup you would need ~3 times the turbines and who knows how many tanks in a 4 turbines, 2 tanks, 2 turbines setup.
As said, this image is just an extract to show it, I'm not using tanks, nor external pipes, cause I didn't need it. It's an old design and it was in the middle just to save it.
I've two 8 core nuclear setup now (controlled by need of energy) and a lot bigger map
Re: 160MW Nuclear Setup, no waste w/ minimal storage tanks
Re: 160MW Nuclear Setup, no waste w/ minimal storage tanks
Posted: Mon Apr 08, 2019 10:30 pm
by zOldBulldog
I made my own chunk-alignable, tilable version using BlakeMW's brilliant design tricks of his 480MW power plant. Engineering is always built on the shoulders of giants.
BulldogNuclearPower.png (4.93 MiB) Viewed 51094 times
Re: 160MW Nuclear Setup, no waste w/ minimal storage tanks
Posted: Sat Jun 05, 2021 12:32 am
by untrust
536MW Square version - Here's my implementation
After starting a server with friends, we quickly needed a nuclear reactor.
Wanting to design one (almost) by myself I stumbled across this forum page, and loved the idea of using the circuit network and steam tanks to judge the reactor temperature.
However, 160MW was going to be too small, and the larger setups that do not use this feature did not seem very efficient, and if you are going to have a lot of reactors it makes sense to save fuel.
After some tweaking and designing in a lab world, I came up with this.
I will try to avoid repeating what has already been said, but we have been running a slightly more basic version of this design (Cleaned it up a bit for forum post!) on our game for about 100+ hours, and used about 800 fuel cells, being roughly a quarter of what we would have needed without a setup like this.
There was an issue with my initial design, however, as the rows of heat pipe and steam turbines were too close and so only one row of turbines could be connected to the steam buffer, causing the power to drop to around 360MW at the end of the "fuel cycle"! Even though it did this for only a second or so, it was enough to make me correct it for this post. Thankfully we didn't yet need the full ~500MW that the design we had provided.
I recommend building this on landfill in an ocean, mainly to make getting enough water to it easier and as a bonus, provides protection from many alien threats.
Often, when starting, the inserters will add an additional fuel cell. You can either remove this and put it back into the circuit network, or just let it get used up. All chests can be supplemented with non networked chests and roboports removed if needed. Just make sure that you don't run out of fuel! (Use a programmable speaker!)
