IF the pressure is as high as the cylinder construction suggests, AND the control valve is closed-center, AND if the control valve has little internal leakage, it's possible for the cylinder to act as its own intensifier and build a -150 pct pressure in the oil trapped in the rod end.
Put a pressure gage in the rod end. Make sure it has a damper, and a full scale range at least twice the nominal system pressure.
Where is the cylinder used? What is the application? The root cause could go back to the original design - not just the cylinder seal but knowing the big picture may help. Is it on an excavator, a wheel loader, etc?
After reading this post, I can give more ideas (considering the correct installation of the scraper):
Regarding ice. By the way, it's due to know where is working this cylinder. The scraper user, following the sketch (and beware, because, the housing is on the wrong side)this scraper is to be used when several mud, or ice can be expected.
Considering that this scraper is a 2 piece scraper, and both of them is just press fitted, it's very easy that, in case of ice formed close to the rubber part, it takes off on the first stroke.
Regarding pressure. Yes, by the way a scraper is not intended to withstand pressure (althougth lately many manufacturers are using double-scrapers as function of U-lip seal, and a scraper, but I think this is not the case here). As told before, the scraper here is a 2 piece encased wiper (being 1 of them made in rubber, and the other in metal), so, in case of just a minimal pressure received, it wouldn't seal, but neither it will damage the rubber-part of the scraper, unless the pressure is too high, so you need to check the closer seals.
Depending on what type of seals are used, the wiper could have been displaced from poor lubrication. It looks like an extruded seal/wiper, from poor installation practices, or perhaps the cylinder uses chevron type packings which were over-tightened.
Your drawing shows a single metal scraper, but your image shows extruded rubber. My guess is the wrong scraper was installed flush (abmormal compression) or at some point, someone thought two scrapers would work better and compressed a rubber one behind the metal one.
The scraper used is made of two part a metal housing and a rubber part. You can read the code in the sketch I provided (trelleborg brand).
Since I'm not the manufacturer of the cylinder by only the end user, I'm now wondering if they have installed a scraper different from what they have specified (altough to my eyes it looks correct).
I'd simply measure the gland and check with the scrpaer requirements. Also, it looks like a lot of extensibility so I would check if the rubber is swollen from compatibility issues.
It looks to me like the rod scraper was installed backwards and the wiping lip was compressed against the bottom of the step causing it to get hot and stick to the rod pulling it out.
The scraper lip usually sticks out beyond the metal can so material will fall clear when scraped off. If you install it backwards and drive it down until it stops the lip will be driven into the chamfer and gap between the rod and the head creating grip between the rod and the lip causing it to wad up and push out.