Accelerator modifications consist of moving the quadrupoles closer to the interaction point and inserting the Roman pots into the region of the electrostatic separators.
There are several issues that must be considered before moving the quadrupoles
including the effect on Tevatron operation, shielding at DØ,
and the method for mechanically supporting the 
quadrupole.
The effects on the lattice are fairly minor and appear to be beneficial.
As mentioned earlier, moving the quadrupoles necessitates a few
percent change in the gradients in order to properly retune the Tevatron.
This is an acceptable change as the gradients remain below the limit
of 140 Tesla/meter for a modest two-thirds of a meter shift in the quadrupoles.
This change also allows the gradient on the 
quadrupole to be
set at an acceptable level, even with 1 TeV operation, which is not the
case for the dispersion-free lattice with the quadrupoles in
the current position [37].
The technical aspects of moving the magnets and
supporting the quadrupole must be studied in detail.
The location of the quadrupoles will have an impact on
the design of the extra shielding that will be implemented in Run II to
reduce backgrounds to the muon system. In this context, DØ engineers
have studied shielding design
with the quadrupoles in the nominal position and moved one meter closer
to the interaction region. The studies show that the
shielding can be just as effective in either scenario.
They also show that supporting the quadrupoles can be done
while maintaining or even improving the current deflection
of the quadrupole without a major redesign. This can be
accomplished by reinforcing and lengthening the shelf that
extends off the main girder that currently supports the
quadrupole. A final study and cost estimate should be prepared
by the Beams Division engineers, but preliminary results are
promising.
The modifications to the separator bypass also require a detailed Beams Division engineering study. It does not seem to be a problem to add the pots as long as a sufficient vacuum is maintained. To this end we plan to have vacuum valves on either end of the pots as shown earlier in Fig. 9, so that the separator vacuum and cleanliness are not compromised by the installation of the pots. It would also be easy to remove a station during a short access (and replace it with a spare or a standard section of beam pipe) in the unlikely event of some problem with the pots.