The ripples come from the camera sensor. See my comment here
When you take a flat with a slit spectrograph, you are effectively taking a spectrum of the flat lamp. The ripples are from the wavelength response of the camera. These are also imprinted on the star spectrum of course so when you divide the star spectrum image by the flat they cancel out along with any other flat defects.
The ripples are caused by interference in the light, between different layers in the sensor (an etalon effect) and their spacing and severity varies from sensor to sensor. (Similar but much finer fringes are seen in high resolution spectra where they are caused by interference due to internal reflections in the sensor cover glass)