I guess that the volumes of UL95 sold in UK and french filling stations far exceeds UL 98 so it's possible that UL98 gets to be stale if the throughput isn't high.

I have had a similar stale fuel problem with motorcycles after winter storage, so now I drain the tanks down in the spring, fill with 5 litres of fresh 95 to get me to a petrol station and use the drained fuel in the mowers. I don't use it in the generator as it hesitates when coming on load on the stale fuel.

Then of course there are the changes to the fuel that we don't know about.

Looking back over a quarter of a century or perhaps longer, there have been changes to:

octane quality: loss of “two star” petrol in 1989-1990
lead in petrol: withdrawal of leaded petrol at the end of the twentieth century
boiling characteristics of petrol (also called volatility): progressive increase in the use of very volatile components in the fuel blend
use of ethanol in petrol: initially at 5% volume, but with the expectation that 10% volume, and possibly higher levels in the blend, will become a reality in the future.

www.fbhvc.co.uk/fuels#changing-nature-of-fuels

Edited by focussed on 18/01/2021 at 21:59

Puzzled as to why you left the second half of my sentence (...so I suppose some of the more modern octane raisers (or detergents) might also leave deposits.) out of your quote, which changes the meaning.

Clearly I'm not suggesting lead caused your problem. I'm suggesting whatever they put in instead of lead caused your problem.

Fuel ageing MIGHT be involved, I suppose, but a loss of volatiles is going to be a lot slower from an underground bulk tank than from a motorcycle (which also suffer here).

I grew up (to the extent that I did) believing that a higher octane than specified gave no benefit, had a lower energy content, and was likely to be harmful, due to deposit formation. This may not apply to all modern engines, some of which may "self tune" to exploit a higher octane, but I've never had one that could do that.

I've always thought the "supermarket petrol BAAAD!" thing sounded like an urban myth. Petrol is nearly a state monopoly here but there is a private refining operation and a couple of private distribution chains, which tend to be very slightly cheaper, with similar unsupported myths in circulation about them. Screeds of nonsense about this (and a bit of "the voice of reason" from that Ducked character. ).

tw.forumosa.com/t/petrol-bought-from-npc-is-watere...1

FWIW (almost nothing) my possibly problematic fill was from a China Petroleum Corporation outlet, which myth mongers say has better quality.

Edited by edlithgow on 19/01/2021 at 09:41

Pollution died down a bit around dusk so I had a brief fiddle with it, just to confirm it was sparking from the coil HT lead, which it was,

Might get one of those adjustable gap in-line spark testers.

This kind of thing

www.aliexpress.com/i/4000897090455.html

Irritatingly about a month ago I was getting push adds from my browser for a Made in Taiwan one that looked quite well made, but of course now I cant find any trace of it,

Ignition capacitors seem to be available still on Ebay from about 6 quid with free postage to the UK.

Postage to Taiwan, OTOH, ranges 25 - 40 quid, a lot when I don't know there is anything wrong with the original. If I can get one locally for under a tenner I will.

Replacement caps are reported to be unreliable, and I wouldn't know if there was anything wrong with that either.

MIGHT be possible to use a non-automotive cap, AFAICT ignition caps generally are in the 0.22-0,3 microfarad range and I suppose have to be able to handle about 500 Volts or so, I'll keep my eyes open.

Meanwhile it seems to have got better, which isn't actually all that reassuring since I just did a general clean up which I doubt was likely to fix anything, so I suppose its consistent with an intermittent cap failure.

Put in a loom bypass from the temperature switch back-probed to the fan relay, starting the engine several times for testing, without any difficulty. (I doubt a back-probed connection is sufficiently robust for long-term use, though, so I will try and make a break-out for the fan relay).

The alternative of opening up the loom and tracing the wire does not appeal.

Edited by edlithgow on 31/01/2021 at 04:38