Do NOT use the distillery recipe to get glyceryl. Use the copious amounts of ash and dark ash you've obtained (you did keep that, right?) in the electrolyzer to get carbon (it's cheap and fast).
By "glyceryl" you mean carbon?
But most importantly don't fuel your nitro-diesel production with nitro-diesel! Use an energy source that doesn't consume EU to produce, such as biogas (this is what I use) or steam (if you have a ton of solar boilers). Finally, only use LV machines to produce the fuel. If you're not making it fast enough, build more LV, don't upgrade them to MV/HV.
All this only means you'll be a bit less efficient in terms of EU production. I haven't calculated it but using MV machinery to make it, while increasing EU usage 2x, won't really make a significant dent in the amount of nitrodiesel your production line will make.
As a final note I use GT distilleries to produce my gasoline. Others prefer using Pneumaticraft. Unless Pneumaticraft can get you a better than 1:1 ratio of oil to GT gasoline, then it's not as efficient as the GT distilleries. I don't use PC for gasoline, so maybe someone else can confirm this. The PC method is definitely the "cooler" way to produce it
In my current world I started with PC distillery but only because I thought making a (near) fully automated production line with GT machinery would be a bigger hassle than I thought (note - I still don't have GT distillation tower). PC distillery wastes a ton of potential fuel compared to GT stuff and you'll also have quite a bit of leftover LPG that I couldn't find a use for even though the entire PC distillation system was running off from it.
In my current world I have a mix of LV and MV machines for the nitrodiesel production. I have three MV distilleries breaking down BC oil to sulfuric light fuel and MV machines to de-sulfurize it and to mix the light fuel with glycerly. Glyceryl line is fully LV. I *could* make the de-sulfurizing and final mixing of the fuel to be LV as well (I had leftover MV machines so I used them) but as the operations take relatively little EU (~4.8k for de-sulfurizing 6 buckets and 30k for final mixing to get 5 buckets of nitrodiesel), having those operations take twice the EU is neglible in grand scheme of things.
What I should do is to replace my 3 MV distilleries making sulfuric fuel with LV tier stuff as they're the biggest energy consumers in the entire production line (512EU per 25mB, 20k per bucket of sulphuric light fuel).
Using rough napkin math and assuming LV machines, it takes 4 buckets of fuel per one run of final mixing and 6 buckets of sulfuric fuel to make 6 buckets of light fuel per run. In other words, for every bucket of light fuel used to make nitrodiesel, final mixing takes ~7.5k eu per bucket of input, de-sulfurizing takes ~800EU per bucket of input. The initial oil->sulphuric light fuel is 20k eu/bucket, as I said. ~2.7x higher fuel usage than final mixing and 25x higher compared to de-sulfurizing.
Obviously, having everything on LV and just spamming more machines to get higher output would be the most optimal solution. Though that doesn't take into account cable losses. Transferring energy at low voltage to more machines will have higher loss than transferring energy at higher voltage. To get same throughput you'll need more machines, 4x more in fact. Obviously, the amount of total loss will depend on specific setup for the machines and can't be trivially calculated but going from LV to MV will definitely not be 2x increase of energy usage, it'll be less than that due to cable losses. How much less, I'm not sure exactly.
My entire system is bottlenecked by the three MV distilleries converting BC oil to sulphuric light fuel. It's also fueled by very same nitrodiesel made by the system (MV generators for MV machines, LV for the LV ones). It still produces more than enough excess nitrodiesel to have constantly running advanced miner 2 (with 2048 EU input), MV ore washing/centrifuge/universal macerators running constantly and also a 512EU EBF, though that one isn't running constantly as I don't have much to give it. Even having everything run on nitrodiesel I still have a ton of leftover.
"Use some other source of energy to make nitrodiesel" makes no sense in terms of efficiency, really. It makes absolutely no difference in what machines you spend your energy. You could run your nitrodiesel production on the same nitrodiesel you make and use biogas somewhere else. You won't be wasting any more (or any less) fuel by doing so.
Only excuse might be that if you happen to run completely dry of nitrodiesel you'd have to jump-start the production with something else. Using regular diesel generators with their 16 bucket internal tank, it'll take quite a while for them to run out and it's relatively unlikely for that to happen before you manage to feed the system with new raw materials.
