Controlled reactor overheating

[ghkbrew]

1 hour ago, JRup said:

I can picture one of the following patch notes:

• Research station no longer works when entombed.

It'll be so said, but it should probably happen.

1 hour ago, JRup said:

1 hour ago, ghkbrew said:

you lose a ton of heat if it turns to fallout because the SHC of nuclear fallout is so low

I'm wondering if some oddball "crying crab" setup can be achieved with this in mind. But I digress.

I had the same thought . And when you factor in that fallout condenses at 26C (IIRC) a complete cycle of nuclear waste -> fallout -> nuclear waste deletes *massive* amounts of heat.

 

42 minutes ago, silentdeth said:

giving 12819.63 kDTU/s which finally surpasses normal operation. (by 0.976%).

Does that factor in the extra heat from the nuclear waste too?

BTW, I love all the math.  I've been too lazy to optimize it quite that much.  (And I'm secretly hoping they fix the "wonky" calculations so I don't have to)

[silentdeth]

14 minutes ago, ghkbrew said:

Does that factor in the extra heat from the nuclear waste too?

In all cases, I'm assuming to waste is cooled to 130C. Every ~70C cooling for the waste is enough to run a turbine.

You could probably go even more efficient if you gave it a 1 gram packet of water to start venting follow by 3 10kg packets to heat up.

[silentdeth]

Ok, I looked into the reactor changes. No in depth, but just fiddled a bit in sandbox mode. It looks like the reactor has changed from heating up 2.5-3kg (2.75kg avg) of fuel 2 deg C a tick, 10 a second, to heating up an average of 59.75kg of fuel 20 deg C a tick. 100 a second. They also changed to heating of the coolant to use, what looks like, a correct conduction equation. There are still some wonky calculations, but fewer. Finally, the reactor now overheat at ~2700C (The overheat gauge doesn't seem to work right atm). Why 2700? I dunno.

The consumption rate of the fuel, and the production of the waste seems to be the same.

What this all means, is that it is definitely more efficient to run the reactor hot. As a rough estimate, just based on the energy value of from the nuclear waste of running the reactor at 2500C (a 200C margin of error) you would produce enough energy run 33.9 steam turbines. ~40.3 including steam

For 'normal' operation, I estimated ~10.2 turbines including both waste and steam.

[ghkbrew]

48 minutes ago, silentdeth said:

They also changed to heating of the coolant to use, what looks like, a correct conduction equation. There are still some wonky calculations, but fewer.

That's my assessment too. But crucially, the heat transfer is balanced now, so no heat is created during conduction.  We only need to worry about the fuel when calculating heat output. Decreased coolant won't change it (until you cause a meltdown)

53 minutes ago, silentdeth said:

Why 2700? I dunno.

It's 3000K exactly (2726.85C). I guess, it's just a round number that gave approximately the right gameplay characteristics (eg time to overheat with no coolant)

 

[ghkbrew]

Practice and theory match up pretty well:

 

This is restricting input to 850g/s of 95C water.  Fuel stays in 2500-2600C range.  Though with the amount of variability I'm seeing I wouldn't run it quite that hot in survival. It just takes one temperature spike to cause a meltdown.

[silentdeth]

8 hours ago, ghkbrew said:

Decreased coolant won't change it (until you cause a meltdown)

Not 100% true. There is still a wonky calculation when you limit the coolant flowrate below the amount needed for normal operations. i.e. In the above example: immediately after venting vent, a 850g packet of water arrives, for 5 ticks everything is fine, it's ok, its a perfect energy balance, then comes another 850g packet of water and physics goes out the window as the packet mixes with the coolant using the wonky equation. Then another second passes and a third packet comes by and says, "Hey guys, is this where we are violating the fundamental laws of thermodynamics?" to which the coolant replies "Yup, come on in." It does, and another wonky calculations follows. Meanwhile physics is curled up in a ball in the corner; sobbing uncontrollably. 

 

3 hours ago, ghkbrew said:

Though with the amount of variability I'm seeing I wouldn't run it quite that hot in survival. It just takes one temperature spike to cause a meltdown.

Yeah, they are a lot more temperamental now. Before it would take it 50 seconds to go from normal 400C to meltdown. Now it would only take 23 seconds.

[silentdeth]

I did some more testing and math. It isn't as wonky as I thought at first, but still a bit weird. [For limited flow rate designs]

Heat transfer is limited by 1000 times the thermal conductivity of the coolant, times delta T. The delay between venting has also been increased to 4 seconds, but they are offset from the pipe packets (at least in my test world) such that the fourth packet is only in the reactor for 0.4 seconds before venting, then there is no coolant for 0.4s then the next packet of coolant arrives the next tick.

Also the heat transfer to the packets is strange. First packet works fine, when the 2nd packet arrives each packet transfers heat with the fuel separately, each with their own 1000xTCxDeltaT limit. Then the two packets mix together and become the same temperature. The next tick, they still transfer heat with the fuel as two separate packets, even though they are the same temperature. This means that the rate of cooling of the fuel is increased. Each packet, as it arrives, is treated the same way until you end up with 4 packets each transferring heat from the fuel.

This makes valve based limited flow designs somewhat less efficient than shutoff based designs. I haven't done the math to see how much the difference is yet (waiting to see if they improve the meters), so I can't say if it is worth the effort to focus on shutoff designs or not.  

[silentdeth]

I did some more testing and and found shutoff based limited coolant designs do not work. At a delay of 1s green 2.8s red, the reactor has too much cooling. Anything above 2.8s red causes the reactor to meltdown.  This didn't agree with the math, so I did some more digging. And it turns out that the heat transfer cap is (kgs of coolant)*1000xTCxDeltaT. I am not sure how this works with fractions of kgs of coolant as with valve limited designs. It's also possible that something maybe worked out with meters instead of shutoffs, to produce packets smaller than 10kg. More testing is required.

[Acrious]

I guess I don't understand how to get the reactor 'running hot', the most efficient, (and aesthetically pleasing) I could come up with is my 13:1 Mushroom Cloud Reactor but I cant get the steam to go above 200c with 13 turbines sucking up the heat... ~16.57kw minus the 3.42kw cost of aquatuners and centrifuges... I could squeeze out more kw by running the turbines plenty warmer... But I kinda' like my chilly mushroom...  How do I run my reactor 'hot?' (I tried adding a flow restriction valve and going down to 900g/s water but all my turbine power outputs went down)

 

[Acrious]

On 4/30/2021 at 2:06 PM, ghkbrew said:

Practice and theory match up pretty well:

 

This is restricting input to 850g/s of 95C water.  Fuel stays in 2500-2600C range.  Though with the amount of variability I'm seeing I wouldn't run it quite that hot in survival. It just takes one temperature spike to cause a meltdown.

My attempts restricting water flow to the reactor ended up with all turbines going down in power... not really sure how you're getting your results... how hot is your nuclear waste pool? even with restricted water flow my nuclear waste didn't come out 'hotter' 

[ghkbrew]

On 5/8/2021 at 2:22 PM, silentdeth said:

And it turns out that the heat transfer cap is (kgs of coolant)*1000xTCxDeltaT. I am not sure how this works with fractions of kgs of coolant as with valve limited designs. It's also possible that something maybe worked out with meters instead of shutoffs, to produce packets smaller than 10kg. More testing is required.

Yeah, I saw this in the new heat transfer code.  There's a transfer cap that is proportional to smaller of the two masses (fuel or coolant, but that it will nearly always be the coolant since they increased the reaction mass).  I'd guess with the delayed 10kg packet designs the coolant is only staying in the reactor for 1 tick before venting so it can't take advantage of the faster heat transfer from having more mass.  Your idea to have a small packet before the 10kg might be effective here.  1g to cause venting then 10kg to sit in the reactor and absorb heat for ~3s.

 

15 hours ago, Acrious said:

My attempts restricting water flow to the reactor ended up with all turbines going down in power

That doesn't make sense.  The heat produced by the fuel is un-affected by the amount of coolant. And the heat produced by the nuclear waste should increase with the average temperature of the fuel.   The nuclear waste is steadily produced in the reactor at whatever the current temperature of the fuel is and then vented when 100kg have been produced.  With a restricted coolant input you quickly start producing nuclear waste hotter than it's 526.85C vaporization temperature.  Are you forming fallout?  The conversion from nuclear waste to fallout deletes massive amounts of heat because of the difference in SHCs.

15 hours ago, Acrious said:

how hot is your nuclear waste pool? even with restricted water flow my nuclear waste didn't come out 'hotter' 

The pool itself is about 200C, but the waste is being produced in a super-heated state.  The trick is to make sure the drips of nuclear waste immediately merge with a large tile of existing "cool" nuclear waste and merge temperatures.  That way you never have a tile of nuclear waste above 526C but you still get all the heat from the ~2500C waste produced by the reactor.

 

Edit:  I'm attaching the save file for this reactor if anyone wants to play with it.

The Lab - New Reactor.sav

[silentdeth]

No changes from testing branch to live. Recommended flowrate for running hot is 867gm/s this will lead to a temperature range from low 2300s to low 2600s Celsius. I don't know what causes such instability in the temperature. I've only seen it happen at higher game speeds. I think it may be that the venting offset from the flowrate is not constant.