Until February 2003, the shutters located in the front end of all beamlines were closed for a few minutes before and during injection, resulting in some temperature variations on the monochromators and/or the first mirrors of the beamlines. This thermal transient resulted in an interruption in data acquisition at each injection, ranging from several minutes up to one hour. The shutters are now left open during injection, allowing the operation to be truly continuous. Thanks to the naturally long lifetime of the electron beam, top-up can be carried out at a very low repetition rate (typically once every 12 hours in the most demanded multi-bunch modes). This is beneficial for the most sensitive and/or advanced beamlines, since it reduces the necessity of gating the data acquisition at frequent injection times. A software counter, which can be used to synchronise the data acquisition, informs the beamlines of the injection schedule.

Before implementing the injection with front end (FE) open, a feasibility study was carried out to verify that it would comply with the radiation protection policy. The Euratom/96-29 recommends a maximum dose of 1 mSv/year for the category of personnel who are not exposed. The ESRF applies this recommendation to all ESRF staff. Taking into account a rate of 2000 hours of work/year yields a maximum dose rate of 2 microSv every 4 hours. The accidental steering of an electron beam pulse freshly injected from the booster into a beamline with its front end open would result in an integrated dose which would greatly exceed this limit. But a detailed study has demonstrated that it is impossible to steer a beam from the booster into a beamline, if a beam is already stored in the storage ring. Consequently, injection with FE open is only allowed if the measured stored beam before injection is greater than 5 mA.

Measurements were carried out to assess dose values outside the hutches of the beamlines during injection with front end open for different machine tunings and filling modes. They have shown that, under normal operation conditions, the injection losses, as well as the losses due to bunch cleaning, in all filling modes produce very low dose rates. Significant dose values were only measured with seriously degraded storage ring optics, resulting in a poor injection efficiency. Nevertheless, radiation monitors have been installed outside the optics hutches on all beamlines, fast enough to interlock the corresponding front end if dose limits exceeded the authorised level during injection (Figure 176).

 

Fig. 176: Ionisation Chamber monitoring the dose rate close to the hutch of a beamline.

 

Indeed, the measured dose around the beamline hutches is dominated by the background level in the experimental hall or in a few beamlines by the Bremsstrahlung generated by the residual pressure in the ID vessel. No significant dose values are integrated during the injections.

The demagnetisation of permanent magnetic material exposed to high-energy electrons was observed at the ESRF in 1993 on two undulator segments made with NdFeB material. Nowadays, the much tighter diagnostic and control of the losses during injection, by means of the network of neutron and Bremsstrahlung detectors, has made such accidents very unlikely for undulators with in-air magnets (magnetic gap > 11 mm). Consequently, the magnetic gap of undulators and wigglers is maintained unchanged during injection, thereby maintaining the heatload unchanged in the beamlines. In-vacuum undulators are built with more resistant material (Sm2Co17). When they are closed to a gap of 5 or 6 mm, even though the injection efficiency when optimal settings are achieved is normally close to 90-100%, it is expected that important accidental electron beam losses will take place from time to time in some of the in-vacuum undulators. For this reason it has been decided to limit the minimum gap to 8 mm (instead of 6 mm) for all in-vacuum undulators during injection. For the corresponding beamlines, this generates a maximum 30% heat load variation over a few minutes. At present we have insufficient information on the long-term effects on magnets subjected to such losses. In the future it is planned to reduce the 8 mm gap limit as low as possible following a detailed quantitative study performed on the ID6 machine beamline.