Nuclear reactor water supply ... with barrels?

[Stryke(r)]

So digging into building a big nuclear power plant (32 reactors) i came accross that a single pipe does only transport around 1200 units of water per sec (which is enough to supply 12 heat exchangers)....

a single barrel of water would contain 250 units of water. put that together with a blue belt, that would be 10 000 units of water per sec (given that you could increase the speed for unbarreling enough), this would allow for a lot more water on the same space (which in my case is curcial (i have a max of 8 tiles available a total of 12 required pipes)

And before you asked, i already looked into the how you could improve throughput on pipes and with a lot of pumps i would roughly be able to bring the pipes to a throughput of around 1600+ units per sec. However, barreling seems like a much better idea for throughput/space wise

Its going to be a 5 GW power plant, so i wanna clarify, that power for the barreling/unbarreling does not really matter much

[Killcreek2]

I have actually just finished a barrelled-water, bot-supplied, tileable, 2n reactor design.

With 4 speed3 modules in a single L3 assembler, it can unbarrel fast enough to keep 9-10 HX supplied. .

You will need a decent barrelling plant to keep up with 32 cores: 200+ barrels / second, at full power output, which would require 5+ full blue belts of barrels [plus at least the same number again to return the empties].

I'm using bots to deliver water to the HX: 1 bot can transport 1000 water per trip [4x250], almost enough for 10 HX to work for 1 sec at max capacity.

[Stryke(r)]

Originally posted by Killcreek2:

I have actually just finished a barrelled-water, bot-supplied, tileable, 2n reactor design.

With 4 speed3 modules in a single L3 assembler, it can unbarrel fast enough to keep 9-10 HX supplied. .

You will need a decent barrelling plant to keep up with 32 cores: 200+ barrels / second, at full power output, which would require 5+ full blue belts of barrels [plus at least the same number again to return the empties].

I'm using bots to deliver water to the HX: 1 bot can transport 1000 water per trip [4x250], almost enough for 10 HX to work for 1 sec at max capacity. [/quote]

well technically i would need to supply only 16 reactors, since the front end has close water access, while the backend doesn't. Technically 1 assembler should work for 12 exchangers (which is the lenght i use) but that would require a beacon so space will be really tight there - i guess bots should work better (just gotta hope my laptop won't melt cause of it)

I still think that its kinda odd tho, that a belt has higher liquid throughput than pipes (eventhough the resource cost is immense)

[Killcreek2]

Oddly, bots are easier on the UPS than field of belts are [even if you are heavy into underneathies]. The belt code optimisations got pushed back to 0.16, unfortunately.

[Stryke(r)]

Guess bots are the way to go... don't even have enough space for belts - lol

http://steamcommunity.com/sharedfiles/filedetails/?id=952164778

http://steamcommunity.com/sharedfiles/filedetails/?id=952164408

[Killcreek2]

Be careful with overloading your pipes ~ remember that flow rate [throughput] decreases with distance between producers & consumers, so to maintain high flow rates you will need repeater pumps / very short pipelines / multiple parallel piplines.

A line of 16 HX, requires 1648 water per second at full power output [per core with full 3x neighbour bonus].

Here is part of mine, though my turbines are all inline with the cores:
http://steamcommunity.com/sharedfiles/filedetails/?id=952175285

[Stryke(r)]

I'm still constructing (building the stuff required on yellow belt takes forever, but in order to get up a full 4 blue belt setup i would need more ore and miners which i did not have power for) - still gotta find a way to get my steam to all the turbines in an apropriate ratio, it should have enough crossings so i'm pretty sure it'll work out. But i propably should try to go for like only 10 HX long setups since the 100steam per sec they produce could easily get through 100 tile long pipes. A 16HX line up looks kinda difficult to do as a compact design

I guess for true infinite blueprints you gotta have a barrels and bots (unless you will to spam pumps everywhere).

How much robots do you use for your 8 reactor setup btw.?

[Killcreek2]

I'm using 3 racks of turbines [8-10 per] for each reactor output.

There are 200 bots in the barrelling depot, and 200 in the reactor array.

I have not stress-tested it yet at max output [will have to turn off my original 8-core to do so].
At half output [500~600MW] there are up to ~50 bots active in the reactor & up to ~100 burst-delivering in the barreling depot.
Note: Am up to robot speed inf L10 which makes a noticable difference in flight times.

Will likely adjust that number up/down later. Extra bots sitting idle is always better than not having enough bots available, right?

Here's a pic of my HX -> steam control tanks -> turbine racks, the turbines are too long to fit onscreen, so the BP view is of an 4-cores deep block:
http://steamcommunity.com/sharedfiles/filedetails/?id=952232433

[Stryke(r)]

Em, since you use pumps for exiting the tanks, why not just use a less "complicated setup" like this (i guess its intentional that you're missing 2.3 turbines for max output?):

http://steamcommunity.com/sharedfiles/filedetails/?id=952574631

Good to hear that it doesn't even take that many bots (i personally was guessing that i would need 1.5k bots for 32 reactors (30sec for both way trip) with speed bonus 4 and max cargo size)


However, there is one thing you definatly have to consider when using that type of water supply... - black starting that thing will be horror (audto disconnecting array of accumulators should work tho). a large of array of reactors will take some time to heat up while your machines will draw power for unloading water. Just a thing to keep in mind (personally would recommend a "manual restart required" situation, since IF (lmao typing this) you run out of power/fuel or anything, you would need to adjust your setup anyway - and having accumulator auto turn on the reactor grid again would in the worst case lead to no power in general.) I dunno how much your power draw is on the machines, but i can imagine, that on a large array of reactors it can easily go up to around 40-50MW

[Killcreek2]

Yes, the "missing" turbines are on purpose: Remember that the ends of the 2n design do not get full neighbour bonus, so would output only 12/16 HX worth of energy, thus require only 75% of the turbines for the 4 endmost cores.
So the "missing" turbines are actually already included at each end. [It is scaled for 16 cores, so 32 would require a few additional turbines, but there is room to easily add a few at the edges of the design, if needed.]

I wanted the design to be completely in-line with the cores, which meant compacting 3 turbine rows inside a 5-wide strip. Not easy, but I think it turned out rather well.
The actual arrangement does not matter much: Once the steam exits the control / buffer tanks you can send it anywhere.


The central reactor core & barrelling depot are actually on a separate double-redundant chemical & solar power grid, to avoid that exact slow start / death-spiral situation you described. :3
The barreling depot uses a lot of power, compared to the reactor fuel / water bot feeds, due to many speed beacons & modules used there. It draws 40-60MW atm, should be around 120-150MW when fully operational [feeding 32 cores].
The turbine racks are only connected to the main power grid, not to the reactor fuel / water feeds or barrelling depot.


The reactor heat-up time is actually not too bad, I was expecting it to take much longer to reach full steam output. There are only 5 heat pipes between reactor & the first HX, so steam starts to be produced very quickly, then ramps up production to max as the heat spreads down the heat pipes.