Error on multiple instances of ACE

Ever since I upgraded from 1.2.7 to 1.3.1, I'm having trouble on launching multiple instances of ACE simultaneously. It never happened before, but now I'm getting this type of errors on `stdout`:

`did-fail-load: -3 - - acehttps://0.0.0.0//OEBPS/Text/book-2.html?AXE_RUNNER=2 - true - 6 - 1`
`did-fail-load: -3 - - acehttps://0.0.0.0//OEBPS/Text/cover.html?AXE_RUNNER=0 - true - 4 - 1`
`did-fail-load: -3 - - acehttps://0.0.0.0//OEBPS/Text/book-1.html?AXE_RUNNER=1 - true - 5 - 1`

When this happens, I get a return code of 1, and no report is generated. Only some books of the batch are affected.

Background:
1) I'm running ACE as part of a calibre plugin (Check Books).
2) It is written in python and uses `subprocess.Popen()` to run ACE.
3) `stdout` and `stderr` are directed to `subprocess.PIPE`
4) Since `PIPE` has a limited size, I'm running ACE in silent mode: args = `'ace -s -f -o check_books_path -l user_lang book_path'`
5) The multiple instances of ACE are being launched using multi-threading: `threading.Thread()`

Workaround:

Temporarily, I've modified the code to rerun ACE on books that get this type of error. Running a second time usually solves the problem. Also, when running a single instance, the error does not happen.

My system:
- Version of Ace impacted by the issue: 1.3.1
- Version of Node used to run Ace: v19.8.1
- My operating system and version: Windows 10 22H2, 19045.3086