I would disagree with the sweeping statement that a scraper is hardly ever run before cement evaluation logging. Some clients will run a scraper before any cement evaluation log, and regard it as a best practice.
For some operators a scraper is always run before cement evaluation logging, unless there is a strong reason why it cannot be accommodated.
Generally speaking, the scraper run is recommended to ensure "clean casing" with no cement sheath, or patches of cement.
There are two main considerations:
1. the response of a cement evaluation tool can be adversely affected by any cement left on the inner surface of the casing
(Ultrasonic measurements are more likely to be affected, or likely to be affected to a greater degree, than a traditional CBL tool, but in my experience all the major logging providers recommend a scraper run regardless of the specific tool technology to be run).
If you are planning to run an ultrasonic tool such as USIT or CAST, they are very likely give better data and more accurate results in a casing that is free from cement coating, deposits, scale or corrosion.
I have never seen a complete log affected ie over the entire logged interval, but I have seen effects ranging from clear degradation of data (to the point of introducing uncertainty in the interpretation) right through to completely unusable or completely uninterpretable data over key sections.
2. Cement evaluation tools generally require strong and effective centralisation - mechanically you do not want cement-debris on the inside of the casing
Additionally a tool with a rotating sub can become fouled by cement left on the inside of the casing and then dislodged during the logging tool rih and log-up by the centralisers.
In many cases after a recent cement job, of course a scraper-run can be accommodated in the time that you might wish to "wait on cement" before logging anyway, as long as you remember to rent/mobilise a scraper.
Now, people HAVE run cement evaluation logs without a scraper run and not encountered a problem, but I would always recommend a scraper run as a best-practice. If it has been decided that a cement evaluation log is required, and if people are spending thousands of dollars for the logging crew and the log-data and perhaps several hundred thousand for the rig time, I assume that it is because they want a good quality log and the most best evaluation/interpretation possible.
3. There is probably a third consideration that has come to mind as I write this. Why are you planning to run the cement evaluation logs and what decisions will the results drive? Eg do you actually need a detailed cement evaluation and estimate of isolation or just a top-of-cement?
There may be other issues (eg a particular well's history), and I am sure there may be different viewpoints - but these are mine based on 38 years in the industry working with cement evaluation logs and data, from running logs , to designing programmes and interpreting the data (or being unable to interpret poor data) for end-users.