Why should we refurbish CEBAF cryomodules?
CEBAF availability at the desired beam energy is limited by cold window and waveguide arc trips. These are thought to be caused by charging of the cold ceramic windows by field emission. Installing doglegs in the cold waveguide should eliminate this. Reprocessing cavities using the latest techniques should also provide useful gradient improvement. Also about half a dozen cryomodules have one or more cavities permanently bypassed due to hurricane damage, helium leaks, tuner failure or other problems.
Is the cavity material good enough to warrant reprocessing?
The specification of the material is the same as presently used. Actual received new material significantly exceeds those specs. Whether the number of defects is comparable to present quality is unknown since production processes have evolved over time. Many CEBAF cavities had quench limits due to material defects or Q-switches so it is prudent to be conservative when projecting gradient improvements. Field emission should be substantially improved with current processes however.
Do we still have the tooling?
Yes, at least one set of all major tooling was put into storage after CEBAF construction and preserved for just this eventuality. Some tooling will need refurbishment and some smaller items will need to be replaced or remade. Some changes may be desirable to take account of current assembly practices, e.g. tooling to test single cavities (as opposed to cavity pairs) in the VTA, or for CEBAF cavity electropolishing.
What does refurbishment include?
Refurbishment as currently proposed includes disassembly and reprocessing of the 5-cell CEBAF cavities using the best available processes (including the options of electropolishing and baking), reassembly with dog-leg waveguides, improvements in magnetic shielding and replacement of all wearing parts such as seals, tuner mechanisms etc. Reasonable efforts will be made to minimize backlash in the rebuilt tuner assemblies. Any defective parts or instrumentation will be replaced.
What is not included in refurbishment?
New cavities are not foreseen. If any are damaged or unusable for any reason they will be repaired or replaced by the best of the old cavities in storage. No changes to the HOM dampers are envisioned, or to the helium vessel or indium seals. A major redesign of the tuner mechanism is not included in this scope of work.
What performance gain can we expect?
We can reasonably expect an average gradient of about 12-13 MV/m, or ~50 MV per module after refurbishment. This is about double the average of existing modules and more than twice that of the worst ones. Some cavities may do considerably better, especially if electropolishing is successful, however we may not be able to take full advantage of those without new LLRF and higher power klystrons. The FEL03 module can operate at over 13 MV/m with beam with the same cell shape using existing LLRF and 8 kW klystrons, however the Lorentz detuning is large at this gradient and recovery after a trip requires manual intervention, as does SL21. We will have a better idea of the reach of refurbished modules after the first few are re-commissioned.
How much does it cost?
Restarting the process, including gathering up all the tooling and refurbishing or replacing missing items, reestablishing the chemistry, HPR and clean assembly processes for the CEBAF geometry, evaluating the electropolishing option and training or retraining the assembly crews is estimated to take about $150k in procurements and $125k in labor in the first part of FY05. Note that this is less than originally projected as all the major tooling has been identified and found to be in quite good condition. Preliminary estimates for module refurbishment add up to $250k procurements, $450k assembly labor, $50k engineering support and $250k for infrastructure, QA, management and testing for a total of $1M/module. The procurements and a small fraction of the labor may come from AIP funds, the remainder must come from the nuclear physics budget. A more detailed estimate is being prepared in anticipation of the FY05 budget cycle.
Is the cost of some refurbishment included in the 12 GeV upgrade?
No, the 12 GeV project assumes a stable base machine. There may have been some conservatism in the 12 GeV spec. to allow for a base machine of less than 6 GeV, but there are no project funds to upgrade or maintain the existing stock of cryomodules.
Why not do the whole 12 GeV upgrade by refurbishment?
While it is clearly advantageous to refurbish some number of cryomodules to achieve stable 6 GeV operation ahead of the upgrade, estimates consistently indicate that it is more cost effective to switch to building new modules in new zones once project funds become available and the stock of underperforming old modules is used up. The exact crossover point will depend on the level of success of refurbishment and the exact cost of new zones.
Where will it be done and by whom?
Cryomodule refurbishment will be done in the test lab by the usual cast of characters. Once SNS production is finished and the Renascence cryomodule is assembled, the assembly areas will be used mainly for CEBAF module refurbishment, with some other work such as the new FEL injector cryomodule, R&D tests and work for others. The areas originally used for cryounit assembly are now occupied by the electropolish facility and the CMM room. We are currently searching for alternative spaces. Candidates include the area presently occupied by the SNS 1 MW klystron, the space in the test lab annex presently used for vacuum assembly and the area outside the CMTF control room presently being used to store a large detector. Proposals to eliminate the test lab annex tech shop, apart from destroying an essential asset, will exacerbate the space shortage and limit or eliminate access to the downstairs assembly area and upstairs weld shop and parts processing areas.
When can we start and how long does it take?
The first module will be available to refurbish around May 05. This is consistent with the roll off of SNS work and availability of assembly technicians. The cavity production group will finish with SNS work earlier and can begin setting up the cavity reprocessing facilities and practice on old 5-cell cavities. The present request is for us to refurbish two modules per year in steady state, far less than the original CEBAF construction rate. The first refurbished module should be ready in the first half of FY06 and when that one is installed we will start work on the one that comes out. A detailed plan is being prepared in anticipation of the FY05 budget cycle.
When will we have stable 6 GeV?
According to Jay’s LEM predictions we should reach 6.0 GeV at 10 FSD/hr with Renascence plus two refurbished modules installed and all systems up and running. To provide “stable” 6 GeV, i.e. with some headroom, we should refurbish at least two more modules and consider at least one per year thereafter to keep up with attrition. The trip rate at 12 GeV will be no better than the base 6 GeV machine unless we assume some contingency in the 12 GeV, so the more modules we can rework the better for the long term physics program.
How high in gradient can we go with the existing LLRF system?
The existing LLRF system operates in the FEL at 13.5 MV/m and maintains stable gradient with 8 kW klystrons. The FEL tuners have practically no hysteresis so the tuning limits are set very close in. Depending on the Qext’s and the detuning and microphonics allowances the system might be able to operate even a little higher. With the original CEBAF cavities, HPA’s and tuners the system will run out of RF power before reaching the limit of the LLRF.
Why do we need a new LLRF system?
<>The existing LLRF cannot handle the high Q’s and large Lorentz detuning of the new-style cryomodules. It is also obsolete and unsupportable and the stock of spares is rapidly diminishing. A new system is needed just to maintain the existing machine and is essential to operate new style cryomodules to their full potential. A requirement of the new system is that it must be capable of running old-style modules and eventually will be retrofitted to existing zones as spare parts for the old systems are exhausted. A new system can also take advantages of the latest thinking in algorithm development, tackle microphonics through electronic damping and piezo feedback and give much better diagnostics of RF system health. And be more economically maintained.
Is it worth installing new LLRF or high power klystrons in refurbished zones?
Depending on how successful we are with cavity reprocessing it may be cost effective to populate refurbished zones with new LLRF and the best existing klystrons. In any case if old LLRF systems are to be replaced due to attrition it makes sense to put the new LLRF into the refurbished and best existing zones first.