Support - Uranium Power

[Fatmice]

Hello, you're welcome to leave suggestions in the sticked thread.

[netmc]

The mod is a lot easier to use than I thought... At least for the basic stuff. I mined a few stacks of ore and set up the full processing chain. I ended up with 1 of the highest level fuel assembly, and about half a chest full of depleted uranium.

I was able to pipe clean water from dytech directly into the reactor and power a few steam engines from it without much issue. I am pretty sure this is the least efficient method for heating the water, but it is absolutely simple. From what I gathered from this thread, pressurised water has a high heat potential and can be used along with the various heat exchangers to be able to run more steam engines.


Is there any type of guideline on when I should use the small, large or wall heat exchangers? Are the R and S models just mirrors of each other so you can place the one that works best with your piping without needing a rats nest of pipes to connect everything?

[Fatmice]

The mod is a lot easier to use than I thought... At least for the basic stuff. I mined a few stacks of ore and set up the full processing chain. I ended up with 1 of the highest level fuel assembly, and about half a chest full of depleted uranium.

I'm glad you find it not too difficult to dive into.

I was able to pipe clean water from dytech directly into the reactor and power a few steam engines from it without much issue. I am pretty sure this is the least efficient method for heating the water, but it is absolutely simple. From what I gathered from this thread, pressurised water has a high heat potential and can be used along with the various heat exchangers to be able to run more steam engines.

Running liquid straight through the reactor does work, and as you expected isn't the most efficient method.

Is there any type of guideline on when I should use the small, large or wall heat exchangers? Are the R and S models just mirrors of each other so you can place the one that works best with your piping without needing a rats nest of pipes to connect everything?

The small heat exchanger is used with the 3x3 and large heat exchanger is used with 5x5 reactors to exchange heat between pressurized water, running in a closed loop, with "vanilla" water in a once through loop. The Dytech "clean" water won't work well. You can switch up the size of the heat exchangers with respect to the reactor if you want as it is a matter of preference. The main difference between those two are the amount of hot liquid they can produce per second and therefore influence how many steam-generators you can hook up to them. The R and S models are mirrors so they will act the same. Just select one that fits nicely with the pipe orientation that you are going for.

The wall heat exchanger is for a more organic build. It has infinite exchange capacities so the amount of hot liquid produced is dependent only on how efficient you are with managing fluid flow. It works with only vanilla pipes, or at least the pipes in contact with it must be of vanilla type.

See the example build for some visual aid.

[Udav]

I have simplified your circuit power plant with Big Heat Exchangers without loss of capacity.
The inner loop.
1. I do not understand purpose a pair of Fast pumps, pumping Pressurized Water from the pipe into her very close to the inlet pipe of the reactor, so remove them together with a pipe between them.
2. Five Fast pumps, pumping Pressurized Water into the reactor superfluous. Suffice it to three of them.
3. Fast pumps, pumping Pressurized Water of the reactor is not required, I removed them.
4. Removing of Fast pumps, pumping Pressurized Water from Big Heat Exchangers into into the tanks did not affect the operation of the system.
5. Assigning a vertical underground pipe located before Fast pumps of reactor is not clear. Everything works without it.
6. If possible I replaced the surface pipe to underground.
The outer loop.
7. Fast pumps supplying water into the tanks are not necessary Offshore pumps enough.
8. Three Offshore pumps per inlet pipe is superfluous. Suffice it to two of them. But without the two pipes supplying water in each storage tank system will not do.

[Fatmice]

I'm sure some simplifications can be done without loss in some generating capacity. I built that setup mainly to demonstrate use of the new heat exchangers.

The reason for pumps in #1 is because of a slight liquid imbalance between the two tanks. Liquid seemed to rush more to one tank than the other. The reason for this is because of the way liquids are moved around in the liquid inventories, probably one of the tank has a higher index number so liquid go there before going to the other one.

[Udav]

About what I expected. But the imbalance is not so large to affect the operation of the system and pumps create an imbalance in the opposite direction.

[netmc]

I am finally at the point that I am playing with the pressurised water and closed loop system. Since I am playing with dytech, I am using the wall heat exchanger. With one loop and a fast pump, I am generating about 22MW of power. I can't get any more water through the vanilla pipe to the steam engines beyond that. Doubling the loop and using the other two outlets on the reactor will bring me to 44MW. This it still far shy of the 144MW the reactor is supposed to be capable of producing.

Is there a better way I can create more loops? I feel that I am missing something critical to generate more power from the system.

[Fatmice]

Hello,

The 5x5 reactor produces 144 MW of heat. You can not produce 144 MW of electricity from that amount of heat . The most you can produce from the 144 MW of heat is about 55 MW of electricity. This is how it is in a real reactor. Now notwithstanding the real reactor, Factorio pipes has a throughput limit of 240 units/s at 0 distance, i.e. direct connection. So, suppose you were to have a steam-generator that consumes 240 units/s of pressurized water, you would still only make 84 MW at most, which is still far short of the 144 MW of heat produced by the 5x5 reactor.

As for getting more heat out, you will have to exchange heat between pressurized water -> pressurized water since you are using the wall heat exchanger. My example build has a setup that produces 55 MW using wall heat exchangers.

[netmc]

Hello,

The 5x5 reactor produces 144 MW of heat. You can not produce 144 MW of electricity from that amount of heat . The most you can produce from the 144 MW of heat is about 55 MW of electricity. This is how it is in a real reactor. Now notwithstanding the real reactor, Factorio pipes has a throughput limit of 240 units/s at 0 distance, i.e. direct connection. So, suppose you were to have a steam-generator that consumes 240 units/s of pressurized water, you would still only make 84 MW at most, which is still far short of the 144 MW of heat produced by the 5x5 reactor.

As for getting more heat out, you will have to exchange heat between pressurized water -> pressurized water since you are using the wall heat exchanger. My example build has a setup that produces 55 MW using wall heat exchangers.

Ahh. I keep messing up the heat output for the electrical output.

I had a couple recommendations from reddit, and I did a bunch of testing and came up with this deign. http://imgur.com/oYI5P3n

(All of the following is with Dytech included.) All pumps are the fast pumps. I added a tank on the steam engine water as a buffer, and reversed the pressurized water flow so it will run opposite the steam engine water. Normally, by the time the pressurized water reaches heat exchanger #8 it is already close (or under) 100 degrees, but since the incoming water is only 15 degrees, I am able to get a little bit more heat transfer between them and guarantee the output water is always 100 degrees. Without that 8th heat exchanger, it would sometimes drop a bit when all the steam engines ran at full throttle.

I am not sure if I am exploiting the higher pipe/engine capacity with Dytech, but with my setup, I only have 6 x 4.7% fuel assemblies installed, and I am able to run a sustained 50.4MW of electricity using Mk II steam engines. I thought for sure I would need additional fuel assemblies to reach this output. Does this mean that I still have more headroom to run a lot more steam engines off this reactor?

I tried using the large heat exchangers (which now work with Dytech clean water by the way), and found them a bit on the slow side. The wall heat exchangers appear to work much faster. The block heat exchangers seem to be a bit strange. It only seems to drop the pressurized water temperature about 25 degrees, but output the steam engine water at a variable rate. It appears that the faster the water leaves the heat exchanger, the lower the temperature drops. So, if you can completely fill the pipe feeding the steam engines, the slower output rate of the water would guarantee the max temperature. Running the water through two heat exchangers doesn't actually benefit anything. The temperature of the final output is still the same. I take it that this is a limitation of the factorio magic liquids and assembly prototype you are using?

[Fatmice]

Ahh. I keep messing up the heat output for the electrical output.

I had a couple recommendations from reddit, and I did a bunch of testing and came up with this deign. http://imgur.com/oYI5P3n

(All of the following is with Dytech included.) All pumps are the fast pumps. I added a tank on the steam engine water as a buffer, and reversed the pressurized water flow so it will run opposite the steam engine water. Normally, by the time the pressurized water reaches heat exchanger #8 it is already close (or under) 100 degrees, but since the incoming water is only 15 degrees, I am able to get a little bit more heat transfer between them and guarantee the output water is always 100 degrees. Without that 8th heat exchanger, it would sometimes drop a bit when all the steam engines ran at full throttle.

Counter-flow is always best for heat exchange.

I am not sure if I am exploiting the higher pipe/engine capacity with Dytech, but with my setup, I only have 6 x 4.7% fuel assemblies installed, and I am able to run a sustained 50.4MW of electricity using Mk II steam engines. I thought for sure I would need additional fuel assemblies to reach this output. Does this mean that I still have more headroom to run a lot more steam engines off this reactor?

Is the Mk II steam engines more than 100% efficient? 6x4.7% => 13714286 J/s ~ 13.7 MW is how much heat is being output from the six fuel assemblies. The only way that you can run 50.4MW on 13.7MW of heat is if the MK II steam engines violate thermodynamics. Currently with version 0.5.3, you can actually extract up to 144 MWe from 144MWq since I took out the heat handicap. But still, it is strange how you only need 6x4.7% assemblies.

I tried using the large heat exchangers (which now work with Dytech clean water by the way), and found them a bit on the slow side. The wall heat exchangers appear to work much faster. The block heat exchangers seem to be a bit strange. It only seems to drop the pressurized water temperature about 25 degrees, but output the steam engine water at a variable rate. It appears that the faster the water leaves the heat exchanger, the lower the temperature drops. So, if you can completely fill the pipe feeding the steam engines, the slower output rate of the water would guarantee the max temperature. Running the water through two heat exchangers doesn't actually benefit anything. The temperature of the final output is still the same. I take it that this is a limitation of the factorio magic liquids and assembly prototype you are using?

Liquid out for the assembler-type heat exchanger is fixed units / s. Remember that it has two inputs and two outputs. For it to work well, you will need to know how much liquid is consumed then make sure that you use less than the amount that the assembler-type can output. You also want to drain the pressurized-water from the output and shunt it back to the reactor as fast as possible. Temperature variation in the liquid output will be a bit more than the wall-type due to heat averaging by Factorio code. As for temperature drop, it depends on the state of the liquids. Then again, it also depends on the consumers. If your steam engines are more than 100% efficient, you will consume less liquid on the steam-engine so overall temperature drop is less on the reactor side. There is also the deal with clean-water. I'm surprised the assembler-type heat exchanger works with clean-water at all because I haven't added its thermodynamic parameters. Maybe DyTech modifies my constants.lua on their end?

Running assembler-type heat exchangers back to back gives no benefit because the liquid products should already be at their predetermined temperatures. By the way, the wall-type heat exchangers also don't really benefit from more than 4-5 per lane as having more simply makes a better temperature gradient. The final temperature will not change significantly with 4-5 wall-type heat exchangers.

[netmc]

I am on version 0.5.4 of Uranium Power.

I've checked the lua files for the current Dytech steam engines, and here are the fluid uses for the Mk I and Mk II

Code: Select all

Mk I
effectivity = 1.1
fluid_usage_per_tick = 0.08

Mk II
effectivity = 1.0
fluid_usage_per_tick = 0.16

The effectivity for the steam engines actually goes down as the tier goes up. Looking through the water-fluid resource prototype, Dysoch has set the pressure_to_speed_ratio to 0.6, the flow_to_energy_ratio to 0.885 and the heat_capacity to 1500J for water. Dirty water is set with only the difference of the heat_capacity set to 1000J. I'm guessing that this differs from vanilla. I found the following in the mod discussion forum on the vanilla steam engine.

{
type = "generator",
name = "steam-engine",
fluid_usage_per_tick = 0.1,
}

{
type = "fluid",
name = "water",
default_temperature = 15,
max_temperature = 100,
heat_capacity = "1KJ",
}

0.1 water/tick * (100 degrees - 15 degrees) * 1 kJ/degree = 8.5 kJ/tick
8.5 kJ/tick * 60 tick/s = 510 kJ/s = 510 kW

Dysoch's math probably works out fine for his being the only method for generating and consuming the hot water, but breaks down once you get other methods involved. Reducing the fluid_usage_per_tick from .1 to .08 along with the increase of heat capacity in the above formula (and multiplying by the 1.1 efficiency) gives 673.2KW (which matches the in-game Mk I numbers exactly).

Your mod is expecting a water heat value of 8.5kJ/tick in the steam engines, and with the Dytech overrides, it is 10.2kJ/tick. So I guess Dysoch is breaking the rules of thermodynamics.

As far as the "clean-water" working with your heat exchangers goes, internally Dytech just calls the fluid "water". Dysoch is using an override to change the label, and making the off-shore pumps output "dirty-water" instead. My guess is that this keeps compatibility with various other mods and vanilla as the fluid is simply called "water" internally, and as such, other mods don't have to add any compatibility patches to work with the Dytech changes.

Even with all the changes that were made and the overabundance of heat for the Dytech steam engines, it was still a pain getting it all set up and working. At this point creating more fuel assemblies it just a matter of mining, since the processing time for creating them is rather quick. I'm using RSO, so I have a rather large deposit of uranium and fluorite available to me.

[Fatmice]

Hm, I meant to say 0.5.4.

So Mk II engines uses 0.16 fluid/tick or 9.6 units of fluid/s. The heat_capacity of a fluid has an affect on the maximal power output of the steam-generator in so far as to overheat it. This means its energy_buffer 0.16*1*85000*16/15 = 43520/3 J. If overheated, then it will produce 870400 W or 870.4 KW. Nominally it will only produce 816 KW with 100C "water." If you feed it fluid with a higher heat_capacity than "water" then it simply uses less of that fluid to fill its energy_buffer and at the same time overheat the generator.

Now regarding DyTech clean-water, I have to say that's pretty underhanded with the override. So it is faking a name of "water" when it is actually "clean-water," but otherwise its heat_capacity is 1.5 KJ/Degree? I have to say my fly-swatting stick seems pretty bothered. Maybe I should write a function to catch DyTech and enforce its clean-water to be clean-water.

But more importantly, I still don't see how you are running what you're running and still be stable. Mathematically, since clean-water has a heat_capacity of 1.5 KJ/Degree, you are getting 0.5 KJ/Degree for free given that the algorithm is being fooled into using 1KJ/C for its calculations. I counted 48 MKII steam-engines in your setup? If running flat out, that's 9.6x48 = 460.8 units of fluid/s => 460.8*1.5*85 = 58.752 MW where 19.584 MW of power is magical. Since the 6x4.7% fuel assemblies can only output 13714286 J/s, 58.752 MW - 19.584 MW - 13.714 MW = 25.454 MW unaccounted for

What's worst, I don't see how you can possibly produce more than 3816*48 = 39.168 MW from what you provided of the MK II steam-engines.

How long can you keep it at peak production? Do you observe a drop in the temperature of the pressurized-water in the reactor over time?

Ugh, you broke my mod. I demand a recourse.

[netmc]

Now regarding DyTech clean-water, I have to say that's pretty underhanded with the override. So it is faking a name of "water" when it is actually "clean-water," but otherwise its heat_capacity is 1.5 KJ/Degree? I have to say my fly-swatting stick seems pretty bothered. Maybe I should write a function to catch DyTech and enforce its clean-water to be clean-water.

Actually, he just relabeled the vanilla water fluid as "clean water" and changed it's heat value to 1.5kJ. It is still referenced as "water" inside of all the game tables.

But more importantly, I still don't see how you are running what you're running and still be stable. Mathematically, since clean-water has a heat_capacity of 1.5 KJ/Degree, you are getting 0.5 KJ/Degree for free given that the algorithm is being fooled into using 1KJ/C for its calculations. I counted 48 MKII steam-engines in you setup? If running flat out, that's 9.6x48 = 460.8 units of fluid/s => 460.8*1.5*85 = 58.752 MW where 19.584 MW of power is magical. Since the 6x4.7% fuel assemblies can only output 13714286 J/s, 58.752 MW - 19.584 MW - 13.714 MW = 25.454 MW unaccounted for

What's worst, I don't see how you can possible produce more than 3816*48 = 39.168 MW from what you provided of the MK II steam-engines.

How long can you keep it at peak production? Do you observe a drop in the temperature of the pressurized-water in the reactor over time?

It's not actually 48 primary steam engines. It's only 36. I actually have 6 secondary and 6 tertiary steam engines, which only really run when my laser turrets fire. The turrets use lots and lots of power. Using your calculations above and adjusting for 36 steam engines, I end up with 44.064 MW as the theoretical maximum. I disconnected the steam engines and let my accumulators drain entirely. After I reconnected the power poles, the steam engines reached a max output of 43.8 MW before the accumulators fully recharged and my power output dropped to the normal load of about 15 MW. Based on the tooltip, Mk II steam engines can output a max of 43.2MW which is still within the thermal max of 44.06MW, so I'm not breaking too many rules. (The tooltip reads 1.2MW output for the Mk II steam engines, but based on the generator formula, it is really 1.224 MW for a total of 44.064MW. I'm sure there is some rounding going on.)

During my testing, I found that the Mk I pipe maxed out at 21 steam engines. It wasn't physically possible to get any more water through the Mk I/vanilla pipe. The steam engines would regularly run out of water briefly. For whatever reason, adding the storage tank to collect the hot water before feeding the steam engines let the system fully power the 21 engines without any of them running out of water. To max out the coverage area of the power substations, I placed 18 primary and 6 secondary/tertiary steam engines on each side. They only run intermittently so don't really affect the sustained throughput of the system. (They do look nice and symmetrical though.) If I ever exceed my primary engine output, I'll need to redesign this a bit, as I can't sustain the water throughput to the tanks to feed that many engines constantly.

I can only run at max power for about 45 seconds or so before the accumulator bank is fully recharged. During that time however, the temperature in the reactor stayed at 349 degrees, and the buffer tanks stabilized at about 3.8k water. Under normal use, the storage tanks have 4.9k water. (I disconnected the secondary engines for this test.)

Here is my factorio folder and save games should you want to load the world up for yourself. https://www.dropbox.com/s/7rumazaipdp0c ... o.zip?dl=0

[Fatmice]

Hm, that is the thing though, leaving it water internally but using a different heat_capacity is not a good thing. Since I can't dynamically load thermodynamic values from data.raw, the script is using the wrong values to do calculations and this leads to strange heat flow behavior. I knew this was going to happen sooner or later.

I'll take a look at your world to see if I'm doing anything derpy. I'm in the process of wrapping up things for 0.6.0 release and I rather not have a bug flying about that could have been swatted.

You won't need a new build. The new version has a different way of making power that uses less water overall. Of course, you can keep what you have if you are happy with it.

[Fatmice]

Okay, I checked out your world. I figure out why there was a discrepancy in the energy input. It is because I derped and forgot to include the reactor power factor in my calculations in the previous post.

A 5x5 reactor has reactor power factor of 175/18 and this is multiplied to the fuel assembly energy potential. The reason for the power factor is that larger reactors are more fertile in terms of neutron economy and so they tend to produce more heat per fuel rod's surface area. This feature is not yet realized to its full potential since fuel_decay is not yet implemented. The power factor will go a long way into attenuating the fuel assembly energy potential when they decay, which will cause them to loose their energy potential as more of them burn up.

Anyways, the reason you only needed to use 6x4.7% is because 6x2^2*175/18*60*1e6 = 133.33 MW. I replicated your build and you really only needed 4 of them when all steam-generators are running flat out. I also pushed your build very close to 60 MW. It is not stable and will fluctuate at around 59.1 MW.

Since DyTech changed the heat_capacity of water, it caused all steam-generator power production to change as well. The reason you're seeing water jumping around is due DyTech's to the change to pressure_to_speed_ratio and flow_to_energy_ratio. The first affects the rate of flow among fluid boxes with different base_level, think stairs, while the second one is the rate of transition through the same base_level based on temperature, i.e. hotter liquid flows faster.

All in all, DyTech's modifications did not break my mod too badly. I just need a way to detect DyTech-Core and enforce the heat_capacity for modified water accordingly.

[netmc]

Before you get to far along in detecting the dytech changes, you may want to look at what he has done with the new version on github that he will be released after factorio .12 is released. It appears to be an almost complete rewrite and code cleanup. Based on the effectivity changes listed for the steam engines, he may no longer be changing the base heart value for water. Also, he is adding in his own nuclear reactor.

[Fatmice]

Oh, but he still does. It is now 2KJ/degree on his GitHub version. This is almost equivalent to my pressurized water, but whereas I don't touch vanilla values he willy-nilly modifies everything. I don't envy his position though as what he does tend to make some strange balancing issues. The doubling of water's heat_capacity means that all steam-generators will output at double rate, i.e. 510 KW -> 1200 KW. Power creep, anyone?

Regarding his nuclear reactor, I see they took a page out of my code as well. The difference is I will never implement reactor explosion by way of blowing up, which is on his trello but not yet implemented. I don't see meltdowns interesting and rather very melodramatic. I have thought of implementing an add-on that will allow manual control of the reactors, but that will likely be far into 0.8.x. With that add-on, you can either place the reactor into "autopilot" or manual to control its heat output by manipulating control rods, which affects reactor temperature, which if too high, can cause reactor damage, which leads to permanent shutdown, which destroys your fuel assemblies. The upsides is that with careful manual control, you will be able to overclock your reactor and run it past the thermal rating. The downside is that you will burn up those fuel assemblies much faster. The addon will also allow you to manually regulate steam production as well as overspeed your turbines and generators, which causes them to take damage and will need constant repair. Of course since it is an addon, you won't need to use it if you don't want the extra layers of complications. Isn't this more interesting than unrealistic reactor explosion with the ensuing radiation clouds that will no doubt cause performance issues?

Returning to the water issue, I don't see a good way to go about circumventing his changes. The only way is to detect if there is DyTech and upscale all of my liquid constants. Since there is no way to read data.raw and upscale dynamically, the values will be fixed and manually adjusted. The long term solution is of course on Factorio's end. They should make the power computation not be directly based on water's heat_capacity. What I could do on my end is to move away from steam-generator types and instead use accumulator types for the generators, heat_capacity be damned.

[Val]

How does one start the turbine in 0.6.0?

I connected everything like in the example, the reactor core is full with hot, high-pressure water, but no steam is generated.

Even in the example, if I just remove a pipe, let the steam to be used up, and then re-add the pipe, the production doesn't restart. The same happens if I rotate (or remove) the recirculation pump, and re-add it later. The cycle doesn't restart.

[Fatmice]

Hello,

Which pipe did you break? Could you show picture please?

If you broke nothing and just rotated the recirculation pump in Uranium_Mod-Example-16, then superheated dry steam should be produced again when the flow of pressurized water is reestablished. Steam generator will only produce steam when two conditions are satisfied: pressurised-water temperature > 280 and there is sufficient liquid in Cold Leg fluidbox.

[Val]

Hello,

I found out what happened. It seems i accidentally also rotated the "cold leg", which doesn't have any visual effect on the cold leg itself, but it breaks its connection to the pipes. The same thing can happen if I remove the pipe directly connected to the cold leg.

If I rotate the cold leg once, I can never get it back to its correct state, no matter how often I press Rotate. I have to dismantle and rebuild both the steam generator and the turbine.

There is only one visual cue about everything being connected properly: if it's correct, then the pipe directly connected to the cold leg shows a small window with water. Otherwise, even if there is plenty of water in the pipe, the pipe is opaque.

This happened to me when placing the buildings (or when removing the buildings from the example and placing them again).

Removing everything and placing them back a few times solved it for me, now the reactor works in my own game as well. In the example, I removed and rebuilt everything twice, after that, it worked again.

By just removing the pipes between the turbine and the cold leg and adding them back again, this "bug" can also happen, but not every time. When it happens, then the pipe directly connected to the cold led doesn't have any window. The only solution I found is removing both the steam generator and the turbine, build them back again, and repeat removing and rebuilding until it works.

At least now I can generate my own nuclear power, many thanks to this awesome mod!


Update: The reactor complex worked for about half a minute, produced about 1.8 MW (the demand was low) and after that it just stalled, the steam generator running out of steam, despite the reactor being full with 350 degrees pressurized water. There is still demand, I placed a few radar stations which were only powered by the nuclear plant, but still, I have to remove the steam generator, the turbine, and all elements which were placed together with them, rebuild them, and the power generation works for another half minute... after which the steam generator fails to generate any steam. If I repeat the process, the plant works for another short period before stalling again.