5^3 (fluid) reactor issue: Keeps overheating when Reactor Planner says it shouldn't

Hey guys, I've run into a frustrating issue regarding the 5^3 fluid-cooled nuclear reactor; it appears to overheat and melt parts of the pressure vessel/hull, even though the setup I'm running should be quite stable. I haven't stared at it to see it overheat, however periodically I'll go down to check on the reactor and see it in EU mode (shouldn't be) and a random piece of the reactor hull missing (hence why it's in EU mode); I can only assume from this observation that something is stopping the coolant from cooling and returning to the reactor via the Liquid Heat Converters, thus forcing the core temp up until part of the reactor hull melts away.

So setup is the 5x5x5 (5^3) nuclear reactor in fluid cooling mode, with the following configuration (using Mauvecloud's IC2x reactor planner)

01010C090C090C090C01010A140D140D140901010C0D0C0D0C0D0C01010A140D110D110901010C0D0C0D0C0D0C01010A140D140D1109

The planner shows that the internal configuration is quite stable, though the heat distribution changes somewhat over different simulations. From the reactor itself I've just got a simple Stirling generator setup converting the ~800Hu/s to ~400Eu/t. The 9 fluid ports, LHEs and Stirlings required are all grouped on the top of the reactor in a 3x3 format. I can take/grab pictures if visual detail is needed to help me troubleshoot.

At present I've got it running my factory so that if none of the machines are running and the storage unit (an MFE) fills up, the MFE is in empty/partially full redstone mode, with redstone connecting from the MFE to a lamp adjacent to the reactor redstone port which, once drained slightly, will turn on the reactor and attempt to fill the MFE back to full, before the MFE deactivates and waits to be drained enough to turn the redstone signal on again. I've also got luminators running off of a batbox filled directly from the rearward stirling generators ('rearward' being the side of the reactor opposite the side with the access hatch). Again, i can grab pictures if verbal descriptions are less than helpful.

If I were to hazard a guess as to the issue, (assuming all other things working) I'd say the problem lies in the storage unit setup, specifically how I'm activating the reactor when the factory is effectively 'idle'; there are some units (e.g. Thermal Centrifuges, Heat Generators for the Blast Furnace) that constantly require a trickle of energy; this continuously drains the MFE, but not quickly enough to continue the redstone signal before the reactor fills again, thus a sort of toggled effect takes place within the reactor. I'm not sure how this would adversely affect the reactor stop it from being cooled, as even if the MFE were completely filled, I was under the impression that the Stirling generators would have enough of an internal buffer to stop the reactor effectively 'backing up' and stopping coolant to run.

At this point I think the easiest solution would be to just install Nuclear Control and install a failsafe switch to shut the reactor off if and when the core temp reaches intolerable levels (what's melting point these days, 50%?), but I'd like to solve the problem without installing more mods if I can help it. (I'm using a homebrew modpack at present, quite tiny and trying to keep it that way.)

If anyone needs further info (or aforementioned pics of the Nuke setup), just ask. Any and all help, assistance and advice is greatly appreciated.

Quote from MauveCloud

As far as the reactor hull melting, here are some things you might double check (though putting 9 of the fluid ports together on one side may make it difficult to check the center one):
1. Do all the LHEs have a full set of heat conductors?
2. Do all the LHEs and fluid ports have fluid ejector upgrades? (and are they configured to eject to the appropriate sides?)
3. Are all the LHEs and Stirling generators oriented correctly to transfer the heat?

I made sure when I assembled it that I had all the requisite doo-dads (fluid ejector upgrades, heat conductors), and I'm pretty sure I've got the stirling generators and LHEs oriented correctly (the two machines should have the side with the orange square touching the other machine, such that the square for each respective set of machines is touching, correct?), but it never hurts to double check; which I did, and it's all fine, all machines and their peripheral devices appear to be installed correctly.

Quote

Also note that as long as the redstone port isn't directly touching one of the inner reactor blocks, the reactor will shut down when one of the hull pieces is destroyed by those effects.

I find that when it does switch back to EU mode that the reactor is still running, even though I've got the redstone port sitting on the upper third of the reactor hull (where it shouldn't be touching any part of the interior reactor chambers), it still continues to operate. This doesn't seem to be an issue (as I said, the configuration appears stable provided I've got coolant running around), but I am more familiar with your explanation mauve, and the fact that that's not happening here is somewhat disconcerting to me.

I'm running a mockup of the reactor internals atm just to check if it is truly stable (no offense intended Mauve, just covering bases), and if it's all good I'll have to have a think about how this is happening; my hunch is that the coolant's stopping somehow, likely due to a full MFE/generator buffers.

Actually, depending on how EU packets are transferred these days, it could be the setup between the MFE and the factory; I've got a MV-transformer in front of the MFE dialing down packets to 128 EU/p before getting sent to the CESU factory storage buffers on each side (well, 3 CESUs and a MFE chargepad, but all at 128 EU/p. Depending how the packets transfer, no matter how intense the factory's working, the MFE just may not be draining quick enough to take on the extra power generated as the reactor cools down.

Quote from MauveCloud

instead of using the redstone modes of the MFE to determine when to activate the reactor, you can set that to "None" and use a redstone comparator to get a redstone level based on how full the MFE is.

Holy Moses, you can actually use a comparator to send a different strength signal depending on how full it is? That's kinda cool (P.S., you learn something new every day :D). I might have to play around with that the next fluid reactor I make.

Quote from MauveCloud

I haven't checked to find out what Stirling generators do when the energy storage blocks they're feeding (and any internal buffer they might have) are full...

You're basically right Mauve; Once a Stirling Generator's energy buffer fills completely, it stops accepting heat (as it's got nowhere to buffer the converted energy to). If this happens for all the Stirling Generators on a reactor this basically causes a cascade failure (the generators can't accept heat, which in turn prevents the LHEs from cooling coolant, which in turn forces any further heat generated by the reactor into the core, thus causing my previous issue, the stages of a reactor going into meltdown).

As for my problem; I seem to have solved it, but it's probably not the best solution (by far). I added a splitter cable directly adjacent to the MFE, so that it only transfers energy once it's completely full; otherwise it waits to fill up. It's a bit of a counterintuitive solution; you'd think I'd want to shunt energy out as soon as possible, but when I was mucking about with it I noticed that without the splitter cable the RS signal from the MFE wasn't updating. However, once I added the splitter cable, the signal down to the reactor worked as intended/planned. Still don't have a clue what's going on, but it's gone though nearly a full set of fuel rods now without so much as a hiccup, so I'm not complaining.

I made an album with pictures and more detailed comments here if people are interested.