Why is my Bash loop printing each number twice?

First check which shell is actually running the script. Calling sh script.sh bypasses the shebang, and sh may not be Bash. Run it with bash script.sh, or make it executable and run it directly so the shebang selects the interpreter.

That distinction matters here because the script uses Bash-style features, including let. It does not by itself explain the duplicate output, though. Check what sh points to on your system and compare the output when you explicitly use Bash.

Add set -x near the start and capture the trace along with the normal output. The trace should show which command runs before each extra number appears. Check whether let resolves to a function, alias, or unexpected command in the shell being used.

The loop shown has only one explicit echo of the counter per iteration, so the duplicate lines need more investigation. The observed output alone does not establish their source.

You can simplify the increment to ((counter += 1)) when running under Bash. Shorter variable names would make the loop easier to inspect too. Neither change should be treated as a proven fix for the extra output.

Quote the filename expansion in touch, for example touch "core${counter}-temp_orders.csv". The braces make the variable boundary clear and the quotes protect the resulting filename. That is good practice, but the original expansion does not explain the duplicate numbers on its own.

One separate issue: after the startup flag becomes true, the outer while true loop keeps spinning without doing useful work. If there is no later work in the full script, remove that loop or exit after setup.