newly bought external got worse after reinitialize, is it going/gone?

[htind]

so I bought this WD drive (external, usb, 4tb, new) and once I started filling it up I noticed it was getting slow, a lot slower than smr drives should be. I ran chkdsk and it showed 13 pending sectors.

I ran the write/read test and sure enough, 13 pending sectors. so I tried to reinitialize it in hopes that would fix it. almost a week later the test was done, no pending sectors, health back to 100%.

I ran a short self check test just to be sure and it failed "because of read element". so I started a read test and, well, this is what I got so far:

20230408-044631_R_WDC_WD40NDZW-11BCSS0_WD-WXB2D91FUKV4_01.01A01-surface-full.jpg (567.73 KiB) Viewed 1621 times

Health 2023 04 08.jpg (88.72 KiB) Viewed 1621 times

any idea what happened, why did it get so much worse all of a sudden? what's your recommendation? would another reinitialize do any good?

I'd also like to thank you folks for the awesome tool and prolonged support.

[hdsentinel]

There is good chance that the USB cable, connection or the USB slot itself is the problem.

For such high capacity 2.5" drives, this is very common: generally things should work - but when the power from the USB port and/or the connection is not stable enough, it can cause weak/pending sectors any time, risking data corruption / data loss.
Yes, as you can see, the Reinitialise Disk Surface test fixed the problems - but if the original cause not fixed, then it is "normal" to expect new and new problems again. As they appear at random positions can indicate that the drive is generally fine, just something between the drive (the cables, connections, possible USB hub / extender cable or so) can cause the problems.

Not sure how the drive is connected. For example, front panel USB 3.0 connections on desktop PCs and/or passive hubs and/or USB extension cables frequently cause similar. I'd try a different USB 3.0 slot and/or a different (shorter) data cable if possible. Sometimes even the original data cable is not the best choice, we can usually find shorter and/or thicker alternatives to provide better power source and data transfer.

Personally for such high capacity USB 3.0 hard disk drives I like to use (and frequently recommend) an USB 3.0 hub with external power supply - as this can provide enough power for proper (and long-term) operation.

[htind]

thank you for your reply. things got a lot worse and I'm dumbfounded.

some information about my setup: I'm using three t440p laptops with identical configurations. I'm connecting the drives directly into their usb 3.0 ports (which use the same interface I believe). I have several external drives, 4tb and 1tb, all WD-branded, same model (elements and elements se)

spent the past month running read/write tests on all of them. following the test I'd fill them with data and md5 hash check its integrity. all but a couple drives checked out.

I disconnected drive #1 (the known "bad" 4tb I tried to reinitialize, the one I first asked about) from t440p #1 because I thought maybe the laptop's usb interface was faulty. replaced the cable with a known good one and connected it to t440p #2. started a fresh read test and got red blocks in the exact same locations as before.

and now I have a problem with yet another drive, one that passed the tests I just mentioned and thus I assumed was good.

drive #2 was already connected to t440p #2 and downloading some files when I connected drive #1 to test it. to my surprise, when I started HDSentinel, it showed drive #2 as having only 7% health remaining with over 1000 reallocated blocks.

thing is, drive #2 is brand new. I got it 2 weeks ago and immediately ran a write/read test followed by a full self check. no problems found. how is it possible that now, about a week after I started downloading those files, it's almost dead?

similarly, how it is possible that drive #1 was at 100% health after I reinitialized it but now the read test shows bad and damaged blocks? I ran it immediately after, so if the cable and/or interface were damaged, wouldn't the reinitialize fail as well?