Just to add an update based on my most recent long ride, hopefully of interest, and at least this way I won't forget
It may be much too much detail, sorry. I should also try to pull this into a graph, rather than lots of words. Something for later perhaps!
On Sunday I did a long ride (but certainly not long in comparison to the PBP which also started then!): total duration ~12h20 riding time ~10h20.
I took usual basal the night before (9U Abasaglar) and reduced morning bolus to 3U. I took a small bolus of 3.5U Novorapid when I got up (0715) though in hindsight I should have taken more. I had a cup of white coffee (~5g CHO) around 0730, blood sugar had started its morning rise (~12mmol/l), faffed around getting food ready and working out what to wear, did another blood test (~14mmol/l) and eventually had another cup of coffee + small bowl of porridge (5+30g CHO) at 0745 and then was rolling down the road at ~0810.
After an hour of Z1/Z2 exercise, my BG was >22mmol/l, not ideal, but I've been there before. I rode for another 30min ever hopeful it would start coming down on its own, but it sort of oscillated around but didn't drop below 22 mmol/l (I note this was all via XDrip+ talking to a libre2, so it's likely it wasn't quite this high). After 1h30 I decided enough was enough and I should drag it down, not least because I really needed to start eating at some point as otherwise I'd not have any appetite later on when it became more important. I took 1U Novorapid @ ~10am and continued on my way.
At ~1125 my libre was reading 10mmol/l, so I started eating - 20g of carbs (Bourneville brunch bar).
At this point the calibrated curve I was following diverged from what I think my blood glucose actually was, based on some finger prick measurements I did later in the day. At about 1215 my calibrated numbers said 7mmol/l, but I think it was probably around 9mmo/l. I ate another 10g of carbs (2x Nice biscuits, they were good), and then set about munching on some granola from a bag, which was another 30g CHO over the next 30min or so.
I arrived for lunch at ~1320 with BG of 8 (tho probably 10) mmol/l, had a cup of coffee (5g CHO), and a single slice bacon sandwich (~20g CHO) then another cup of coffee (5g CHO). Stopping is troublesome as the carbs are no longer being used so what you eat raises your blood sugar, but if you take bolus as I would have done in the past for slightly shorter rides, I end up going really low on the way back home, which isn't much fun as you need to keep eating constantly.
So this time I accepted I'd probably run high afterwards and planned to keep the stop short. In the end it was ~1h15 (I was chatting), but I eventually restarted on the homeward leg around 1435 with BG of 12 (15) mmol/l. It continued to rise for another 15min (peak of 15 (17) mmol/l), then started coming back down. Shortly before 1500 I started eating with BG 9 (10) mmol/l. I had 3 batches of 2 Nice biscuits over the next 30min with BG hitting a low of 7 (9) mmol/l around 1630 and then starting back up to a high of 10 (12) mmol/l at 1700 then starting back down.
The next cycle was similar (lasted ~an hour, with 30g CHO, min BG of 6(7) mmol/l @ ~1745 rising to 10 (12) mmol/l @ ~1810.
The cycle following that was also similar, however I'd upped the pace a bit now as I could see home was vaguely close, which may have resulted in a downward trend. I ate 30g CHO at ~1855 hit a low of 4 (6) mmol/l @ ~1915 then a peak of 7 (9) mmol/l at ~1950.
Had I been riding further I'd have eaten some more, but as I felt fine I accepted the downward trend from then on and arrived home @ ~2030 with a BG of 4 (4) mmol/l.
I had a glass of milk when I got back and my blood sugar started going up about 30min later. I wasn't sure whether I'd go low or high post-exercise, so had my bolus at the same time as food, I didn't need to bother, there was no difference to normal in my response to bolus insulin, so I ended up running a bit high after supper and then coming down to around 10 mmol/l at bed time. I took 0.5U Novorapid to fix that (I'd normally have taken more but was concerned I might go low), took my normal basal (9U) and went to bed.
I then dropped to ~3.5mmol/l and sat there solidly all night. I must have been just above the alarm threshold as I don't think it went off. My dawn phenomenon the following day was very small/almost non-existent and I had slightly higher insulin sensitivity than normal for my boluses - I'd take the normal amount and need to eat more than expected (this is as far as my analysis has gone), but by that evening everything was back to normal, including dawn phenomenon the following morning.
That was very long, sorry.
To summarise: 30g/h while riding seems reasonable with a morning basal reduction to ~1/3 of normal. I can eat that quite happily and felt fine riding - there is a net energy deficit but it's only a single day and it's easy to make it up in the evening/by eating more for lunch (but faster). I should take more bolus in the morning to cover dawn phenomenon, this will probably be fairly easy to handle as there's no issue with forcing food down so early in the day if I get it wrong. It may also be a case of nerves as this was my longest ride to date. Practice makes perfect, ideally not being so high tho!
While I have previously commented that I don't need to reduce basal the night after exercise, it appears to depend on what you're used to. For really long rides it looks like I should indeed do this as the reduced morning dose is not enough of a reduction.
I think lunch stops should be shorter. To fix the BG rise I'd need a reasonable amount of bolus (2U+), but then that would clobber me later in the day. I could see myself taking 0.5U, but that won't help the rise after lunch and will just result in me going (a bit or a lot, hard to know) lower later on. I should probably try to eat a bit more too.
In terms of eating later on in the ride (when you don't really feel like it), I will have to try setting myself an alarm to remind me to eat something every 30min, rather than every hour, as this might make for less pronounced oscillations (tho they weren't bad) and also make it easier to keep eating).
I can't find the data now nor remember quite when, but while riding up a rather steep hill, I did note that my BG had gone up more than I'd expected, which made me think that my liver must be recharging its glycogen and now releasing it for me (this was well into the ride). I can't for the life of me recall when this was though, so take it with a pinch of salt. Perhaps it will become clearer when I finally get around to plotting a graph of BG vs pedal power, OTOH I may just have been day-dreaming