[Idea] Reactor heat exchangers (replacement for CASUC)

I believe Alblaka recently mentioned that he wants to get rid of CASUC type reactors . I had an idea that might allow similar levels of cooling without (hopefully) a lot of extra work for the coders, since it might be able to re-use a lot of the energy net code.

Basically, instead of moving actual coolant around, you'd just use "heat units" (similar to EU).

At a high level, it would work like this:
- Attach a heat exchanger block to the reactor in place of a chamber. This would have a 16-slot inventory that accepts heat dispensers. Each one would allow the transfer of 128 HU (heat units) per reactor tick from the hull to the heat exchanger for a max of 2048/tick. The heat exchanger would be similar to an MFE or MFSU with a storage of 2048, and input and one output face.

- Create a coolant loop. The pipe blocks would work similar to cables, except you'd have extreme, high, medium, and low volume instead of voltage, capable of carrying the corresponding amount of heat units per tick (not exploding, just unable to carry more than that). It could require transformer analogues or just allow connecting smaller pipes to bigger ones. The loop has to start at the heat exchanger's output face, and return to an input face(s).

- Each section of pipe would dissipate 0.125 HU/t to air or stationary water, and 0.5 HU/t to flowing water, for 0.5-2HU/t per pipe section. Pipes carrying more than 128HU/t have a chance to evaporate adjacent water (maybe not... might create a lot of CPU load). This means to reach current CASUC cooling levels you'll need huge cooling towers (flowing water, remember), just like real reactors, with at least 1024 pipe sections (or over 1/2 cubic chunk in the most efficient configuration).

- Any heat left in the coolant is returned to the heat exchanger, reducing the amount it can take from the reactor next cycle.

- Charge the cooling system. The heat exchanger interface would have a slot that accepts cooling cells. It would take one cell for each pipe segment connected to the system. (no, you don't get the empties back) Any time you make ANY changes to the cooling loop, you'll have to repeat this step from the beginning, which makes it very expensive to fiddle with large reactors.

- The heat exchanger would reduce reactor output by 1EU (HV & EV pipes), 2EU (MV pipe), or 4EU (LV pipe) per reactor tick for every pipe section in the system to run the pumps. This translates to about a 10% loss on a 2kEU, 2kHU reactor.

Hopefully it would be possible to calculate the heat loss of the heat exchanger just once, and thereafter when it changes. That would mean that the only ongoing calculation is the heat exchanger block subtracting the loss number from storage, then transferring heat from the reactor.

= iron block
= nothing

= 4 EV pipe

= 4 HV pipe

= 4 MV pipe

= 4 LV pipe

= pump
= reactor chamber
= EV pipe

= heat exchanger

All of the heat numbers should be in reactor ticks (seconds).

The EV pipes don't need to be that expensive, simply because they're not any more effective at cooling than the smaller ones are. That recipe is already 18 iron per section - are you really going to make the whole thing out of EV pipe? That's 18,000 iron... If anything, it might need to be cheaper.

The bigger pipes are really just intended for going from the reactor to the cooling tower & back.

Quote from MJEvans

You're missed the point I was trying to get across; instead of advanced machines, the product of advanced machines (iridum plating) would be used. This conserves natural resources and instead uses un-natural resources (energy), but lots of them. Plus the recipe would produce multiple lengths of pipe (like RP tubes or BC pipes).

Well, it is a big, dumb chunk of metal pipe... it's supposed to use up natural resources. The balance I was going for was bigger pipes use lots of resources up front, but since it's easier to push fluid through a large diameter pipe, you get the ongoing benefit of less power leeched from your reactor. Or, you can use the smaller, cheap pipes that have much less of an impact initially, but have higher running costs. On a 2kEU reactor, the savings would be 30MEU per run using all EV pipe instead of LV. I would support the ability to toss pipes into a furnace and get 100% of your metals back, though...

A fifth type of pipe, maybe a finned LV or MV pipe that uses iridium plates as cooling fins to double heat transfer would be another alternative...

Edit: I also just realized that using cooling cells to fill the system (1000 cells that don't stack!?) is a terrible idea...

Maybe use normal water cells that give you 50% effectiveness (up to 1024/reactor tick), or liquid sodium for 100% effectiveness. You'd need a new machine to get the sodium coolant, though... Maybe a "mineral extractor" that takes any maceratable ore and empty cells, and gives you sodium cells and ore dust back (with a chance to only get n-1 dust instead of the normal amount).