When it is at a low battery percentage, I would take a look at the details of what power details were negotiated and what the current current is.
Should be “[PD]”, max 5V 3A, and would be perhaps 1.5A actual current, maybe a bit more.
Have you tried removing the battery and reinserting it? Then doing several full discharge / charge cycles in order to recalibrate?
Have you tried doing a long (20 seconds) press of the power button?
I am actually in one of those precious moments where the lapdock is working so I won’t unplug to see how exactly power delivery works with the charger 
However my first thought was that the battery was dying so I removed it and inserted the never-used one I had ordered with the phone a few years back, and that didn’t improve things. I’ve done the very-long reset once or twice as well.
One question about the things in /sys/class/power_supply, I see a bq25890-charger-0 and a tps6598x-source-psy-0-003f (and the battery). What is the difference between those? I think I remember seeing the charger enabled (“ONLINE”) and the “source-psy” disabled (ONLINE=0).
/sys/class/power_supply/tps6598x-source-psy-0-003f/uevent before power-cycling (charger plugged in, but not working):
POWER_SUPPLY_NAME=tps6598x-source-psy-0-003f
POWER_SUPPLY_TYPE=USB
POWER_SUPPLY_ONLINE=0
POWER_SUPPLY_PRESENT=0
POWER_SUPPLY_CURRENT_MAX=500000
POWER_SUPPLY_VOLTAGE_MAX=5000000
/sys/class/power_supply/bq25890-charger-0/uevent:
POWER_SUPPLY_NAME=bq25890-charger-0
POWER_SUPPLY_TYPE=USB
POWER_SUPPLY_MANUFACTURER=Texas Instruments
POWER_SUPPLY_MODEL_NAME=BQ25895
POWER_SUPPLY_STATUS=Discharging
POWER_SUPPLY_CHARGE_TYPE=N/A
POWER_SUPPLY_ONLINE=0
POWER_SUPPLY_HEALTH=Good
POWER_SUPPLY_CONSTANT_CHARGE_CURRENT=1472000
POWER_SUPPLY_CONSTANT_CHARGE_CURRENT_MAX=1472000
POWER_SUPPLY_CONSTANT_CHARGE_VOLTAGE=4208000
POWER_SUPPLY_CONSTANT_CHARGE_VOLTAGE_MAX=4208000
POWER_SUPPLY_PRECHARGE_CURRENT=128000
POWER_SUPPLY_CHARGE_TERM_CURRENT=128000
POWER_SUPPLY_INPUT_CURRENT_LIMIT=500000
POWER_SUPPLY_VOLTAGE_NOW=2304000
POWER_SUPPLY_CURRENT_NOW=0
After power off, connecting charger (red indicator turns on), turning phone back on, the psy one says the same, and charger says:
POWER_SUPPLY_NAME=bq25890-charger-0
POWER_SUPPLY_TYPE=USB
POWER_SUPPLY_MANUFACTURER=Texas Instruments
POWER_SUPPLY_MODEL_NAME=BQ25895
POWER_SUPPLY_STATUS=Charging
POWER_SUPPLY_CHARGE_TYPE=Fast
POWER_SUPPLY_ONLINE=1
POWER_SUPPLY_HEALTH=Good
POWER_SUPPLY_CONSTANT_CHARGE_CURRENT=1472000
POWER_SUPPLY_CONSTANT_CHARGE_CURRENT_MAX=1472000
POWER_SUPPLY_CONSTANT_CHARGE_VOLTAGE=4208000
POWER_SUPPLY_CONSTANT_CHARGE_VOLTAGE_MAX=4208000
POWER_SUPPLY_PRECHARGE_CURRENT=128000
POWER_SUPPLY_CHARGE_TERM_CURRENT=128000
POWER_SUPPLY_INPUT_CURRENT_LIMIT=500000
POWER_SUPPLY_VOLTAGE_NOW=3744000
POWER_SUPPLY_CURRENT_NOW=0
POWER_SUPPLY_TEMP=275
I don’t see a PD anywhere…
For my Librem 5, this has been the “normal” behavior for several years now. I’ve never found an explanation as to why. But I can plug it in while it’s powered on, and then power down while it’s still plugged in, to make charging begin a few seconds later.
At first I thought it might be like a particular laptop of mine, needing a minute or so for the connected power to be updated in the settings and to “realize” it was charging. Unfortunately, it wasn’t the case.
And it has never been caught in the rain.
Oh that’s weird. Does it also take forever to charge?
I start recharging at about 40% and it usually takes 3 or 4 hours to get to 80%. From 80% to 100% requires another couple of hours, I think. I haven’t really clocked it lately, though, and also have only had it powered on for a few minutes at a time over the last few weeks.
You can get some idea of the charge trends by installing the Power Statistics app. You’ll have to use the Mobile Settings app to make it fit the window better. (Open the Power Statistics app, then open the Mobile Settings app, from which you enable the toggle in the “Compositor” setting.)
Can we see the first one when the battery is charging, preferably when the battery is at a lower percentage state of charge?
Example for me when it’s working …
POWER_SUPPLY_NAME=tps6598x-source-psy-0-003f
POWER_SUPPLY_TYPE=USB
POWER_SUPPLY_USB_TYPE=C [PD]
POWER_SUPPLY_ONLINE=1
POWER_SUPPLY_PRESENT=1
POWER_SUPPLY_CURRENT_MAX=3000000
POWER_SUPPLY_VOLTAGE_MAX=5000000
That’s connected to the Purism-supplied charger. So it has negotiated USB PD and specifically 5V 3A max. (Actual current, not shown by that output, will decline throughout the charging process.)
If it were connected to a random computer’s USB-A 3.0 port then the CURRENT_MAX would be 500000 and USB_TYPE would be “[C]”. Your output may show additional options for USB_TYPE but, regardless, the square brackets enclose the power protocol option that has been successfully negotiated.
There may be some merit in posting output from
sudo cat /sys/kernel/debug/tps6598x/0-003f/customer_use
which will be a 64-bit number in hex prefixed with “0x”.
Yes. I got caught in the rain once and it ruined the USB board. The repair cost $49.00 USD. If you contacted Purism, they might (should) send you a usb board to replace yourself.
Do we need a USB-C M connector that is a rubber lug for water-proofing purposes when the USB-C port is not in use?
Or avoid holding it upside down in the rain. 
No, my Librem 5 USA still charges fine in my region where rain occurs frequently, but I also keep it connected to a power bank when traveling outside.
I wondered about that too. Maybe put back in a pocket carelessly and hence in a pocket upside down? Perhaps the OP could shed light on that.
There is evidence that the answer might be “Yes” but it may be that a very specific and unfortunate set of circumstances are required to reproduce the problem. You’ve been fortunate. Two other customers may have been unfortunate. (Rain hasn’t been a thing here lately so I can’t say.)
Such a lug could also serve to exclude dust and other particulate matter, even bugs (of the animate kind).
I have a portable disk (USB) that comes with such a lug. I have a number of AC outlets where I use a similar type of thing but that is mainly for safety reasons (and they are typically plastic, not rubber).
Other handheld devices in my household are able to detect moisture in the charging port, such as the iPhone:
I would assume advice from Apple for this device could also easily apply to the Librem 5.
Sorry but I would think that that is a poor assumption.
I believe the Ingress Protection (IP) rating of the Librem 5 has been asked about a few times over the years but I have never seen any official claim e.g. result of actual testing in accordance with IP standards.
Probably Purism should sort that out. @JCS ?
The observation has been made previously in this forum that a high IP rating tends to correlate with poor repairability / replaceability. That is, it is easier to make a device water proof at 6 metres for 30 minutes if the customer can never open the damn thing up because there is nothing that you can repair or replace (and when a qualified repairer does open it up, the repairer will have to reestablish the water proofing and a diligent repairer would even waterproof test the repaired device).
It’s great that more recent models of iPhone have charging-port liquid detection but that tells you nothing about what the Librem 5 has.
You will note that Apple says
Splash, water, and dust resistance are not permanent conditions and resistance might decrease as a result of normal wear. Liquid damage is not covered under warranty
I assume that’s because the seals will perish over time. So eventually an iPhone will lose its IP rating. I would imagine that the speed with which the IP rating will be lost will depend on the extremes of temperature to which the iPhone is exposed.
I think the only solid advice in common is
Yes, I agree with your conclusion.
It is a great recommendation to include IP ratings (as well as a myriad of other certifications) as part of official documentation. I created an internal ticket for product & mechanical design attention.
Oh good to know thank you! I’ll contact support then.
I had put it in my rain coat pocket (indeed upside-down), which looked very much waterproof 
Anyway on the other end is the ear speaker and headphone jack, maybe I’d be writing how those no longer work since the rain if I had put it upside up instead!
Raincoats vary in water resistance technologies, and the only type that is truly waterproof are based using PVC.