TRIUMF

�

ANNUAL REPORT

SCIENTIFIC ACTIVITIES

2005

ISSN 1492-417X

CANADA’S NATIONAL LABORATORYFOR PARTICLE AND NUCLEAR PHYSICS

OPERATED AS A JOINT VENTURE

MEMBERS: ASSOCIATE MEMBERS:

THE UNIVERSITY OF ALBERTA THE UNIVERSITY OF GUELPHTHE UNIVERSITY OF BRITISH COLUMBIA THE UNIVERSITY OF MANITOBACARLETON UNIVERSITY McMASTER UNIVERSITY

SIMON FRASER UNIVERSITY L’UNIVERSITE DE MONTREALTHE UNIVERSITY OF TORONTO QUEEN’S UNIVERSITYTHE UNIVERSITY OF VICTORIA THE UNIVERSITY OF REGINA

SAINT MARY’S UNIVERSITY

UNDER A CONTRIBUTION FROM THENATIONAL RESEARCH COUNCIL OF CANADA DECEMBER 2006

The contributions on individual experiments in this report are out-

lines intended to demonstrate the extent of scientific activity atTRIUMF during the past year. The outlines are not publications

and often contain preliminary results not intended, or not yetready, for publication. Material from these reports should not be

reproduced or quoted without permission from the authors.

iv

CYCLOTRON OPERATIONS DIVISION

INTRODUCTION

Beam delivery for the year 2005 was quite success-ful with a total charge production of 722 mAh deliv-ered to the high current proton lines (BL1A, BL2Aand BL2C4). This was a 5% increase from last year,edging out the previous record of 720 mAh set in 2002.Of these, 434 mAh were delivered to BL1A for mesonproduction and 130 mAh (20% more than last year)were delivered to BL2A for the production of radioac-tive ion beams (RIB) in ISAC. This was a record yearfor strontium-82 production in the 2C4 solid target fa-cility (STF). A total of 55.0 Ci were produced from158 mAh in 163 days compared with 48.5 Ci producedfrom 134.3 mAh in 137 days in 2004. Eleven natural ru-bidium targets were irradiated in 2005 compared with9 targets in 2004. However, the cyclotron was availablefor only 86% of the scheduled time, down 6% from lastyear, mostly due to an rf resonator water leak.

A new BL2A tune, which produced the large spotsizes required by the ISAC high-current targets, wasaccomplished by the correction of an error in theprogram calculating the transfer matrices. The BL2Abeam stability was greatly improved by a pulser-adjustment program, developed by the Controls group,that greatly reduced the number of spurious trips andsignificantly improved the quality of RIB productionin ISAC. With this beam stabilization loop enabled, itwas possible to extract 275 μA for 4 hours without en-countering any problems. When the loop was disabledit was possible to extract 296 μA at a duty cycle of97.3% for 2.4 hours with greater than 60% transmis-sion and ∼1% tank spills without any thermal prob-lems. This is a new high-current record for TRIUMF.

The annual downtime of 827 hours was almost dou-ble that of last year with rf problems costing 584 hoursor 70% of the total. Most of this (508 hours) was due toa resonator water leak leaving the remaining 76 hoursof rf downtime (about 2 hours per operational week)continuing the excellent behaviour of last year. Thebulk (85%) of the remaining 244 hours of cyclotrondowntime was more or less shared equally by the usualsystems (beam lines, services, site power, beam trips,vacuum and power supplies).

During the winter shutdown the major beam lineactivities were various tasks on the 1AQ15 quadrupoletriplet, the repair of vacuum leaks at 1AVA8 and theM15 slit drive, and the repair of an M9BQ3 shortedcoil. The many other meson hall activities are coveredin the Beam Production and Primary Beam Lines sec-tions. An earlier decision had postponed the large M20refurbishment work to 2006, but this did not offer muchrelief as some of the remaining tasks proved to be very

difficult in terms of dose and time. For example, repairof the Q10/11 quadrupole doublet was not originallyanticipated and it developed into a warm-cell opera-tion.

When the lid was first raised, Safety surveyorsfound an area of the tank to be contaminated with7Be. Although not entirely understood, the evidencepointed to it being caused by an overheated extrac-tion foil. Overheating may have resulted from higherbeam densities associated with beam shadowing tech-niques or from the lower heat emissivity of the BeCufoil holder. A 3 m radius was flagged off inside of whichrespirators and additional protective clothing were re-quired for any tank work. The same precautions wererequired for work on the extraction 1 probe when re-tracted from the tank.

Major tank activities included work on extractionprobes, periscopes, slits, flags, wire ways, correctionplates, vacuum pumps and gate valves. A four-yearprogram of replacing the copper chore pads with fibre-glass pads, which were protecting the cooling circuitbellows from vibrating, was completed. The details ofthese activities are covered in the appropriate groupreports. Major work in the vault included the installa-tion of additional safety railings, the inspection of trim-and harmonic-coil junction boxes and the replacementof the phase probe and all lower correction plate cables.

During start-up it became apparent that the vac-uum was not as good as it should be and subsequenttests indicated a water leak in resonator 3U6 (quadrant3, upper resonator 6). This caused a three week delayin the beam schedule while the repair was being made.Another surprise that came on start-up following theresonator repair was a high-temperature indication onresonator 3L1. Back-flushing efforts failed to dislodgewhatever may have been restricting the coolant flowbut, in mid May, some small perturbations in the res-onator supply pressure that resulted from manual op-eration of the three-way valve after an rf trip appar-ently cleared the cooling channel and no subsequentproblems occurred.

The major activity in the fall mini-shutdown wasthe work on the 1A triplet, which was completed aheadof time because of an early start, and the testing ofvarious back-flushing techniques, where it was foundthat general chemical flushing was so successful that itwas not necessary to dig down to isolate and flush thefew individual coils known to overheat. This not onlysaved much time and dose, but held promise that anyinterventions required in the future would have similarsavings.

To offset this good news, a CuALCW water leak

197

was isolated to M20Q1. An additional week was re-quired to uncover the area, repair a Hansen fitting,and recover it again. This still left BL1A available nearthe beginning of the start-up period as the preferredoperational mode to tune the cyclotron and the leastdisruptive way to bring on the other proton lines. Itis worthwhile to note that following the fall shutdownthe cyclotron was available for 95% of the scheduledtime.

Except for the 508 hours of downtime due to theresonator water leak, the upgrade activities describedin the rf section of this report were responsible forthe continued reliable operation of the rf system. Ma-jor activities included the overall refurbishment of thefilament power supplies, a new stainless steel hairpininductor design, increased energy discharge capacityof the HVPS varistor disks, commissioning of a highpower rf coaxial switch to redirect rf power either to thecyclotron or a resistive load, and a complete overhaul ofthe transmission-line matching section. To complementthis work, the RF Controls group made good progressin the high VSWR protection system, the spark detec-tion/recovery system, the new HVPS crowbar system,and the rf booster control system.

A new intercepting probe for the TRIUMF cy-clotron capable of measuring phase and time structureof the circulating beam was designed, manufactured,and installed in the tank during the winter shutdown.Signals from the tank nonintercepting capacitive phaseprobes were inspected and signals from all four probesare now equalized and exactly out of phase for eachpair. New electronics that is capable of extracting reli-able dc current information from a narrow band-widthrf signal with the amplitude proportional to the beamintensity over a wide dynamic range was built for theBL2A capacitive probe.

The vacuum systems on the beam lines were mademore reliable by upgrading to O-ring metal seals, re-placing pumps and refurbishing gate valves. On thecyclotron side much MRO work was done, but the ma-jor activity was tracing down the water leak on theresonators that developed following the winter shut-down. The cyclotron vacuum/cryogenic system contin-ued to operate well this year. The main beam down-time was due to B-20 electrical problems and inflectorcryopump gate valve failure. These two problems willbe addressed in next year’s winter shutdown.

The cusp ion source, injection line, and inflec-tor/deflector systems continued to operate well for thepast year with uptime in excess of 99% of the sched-uled time. The Prompt Radiation Safety Trip Sys-tem was revised to include redundant Safety CriticalGamma and Neutron Monitors. The system was in-stalled, tested and successfully commissioned in Febru-ary and March. New 24-V power supplies for the ISIS

interlock control system were purchased and installed.The restored first chamber of the 1 MeV test cyclotronhas allowed testing of cusp ion sources by ISIS, Nor-dion and Dehnel Consulting. Dr. Yong-Seok Hwang, avisitor from the Department of Nuclear Engineering,Seoul National University in South Korea, completedan excellent study of the ion source characteristics.

The most significant new development by the Con-trols group was the work on beam stability for BL2A.This work resulted in an important improvement forISAC. The Central Control System (CCS) ran well dur-ing 2005. As recorded by the Operations group, the lossof scheduled beam time due to CCS faults during theyear was 3.4 hours. This low level of downtime was pro-vided without compromising important hardware andsoftware advances.

Operational Services, which includes Remote Han-dling, Magnet Power Supplies, Electrical Services andMechanical Systems, continues to play a vital role inthe reliable operation of the cyclotron. Their servicesextend to the whole TRIUMF site and are appreciatedsite wide.

BEAM PRODUCTION

Beam delivery for 2005 was good with a total chargeproduction of 722 mAh delivered to the high currentproton lines (1A, 2A and 2C4), up 5% from last year,to edge out the previous record of 720 mAh set in 2002.This success was achieved in spite of a lengthy start-up after the winter shutdown caused by a resonatorwater leak, and was in part due to record charge pro-duction for BL2A and BL2C4. The performance of thecyclotron remained strong, operating typically with atransmission above 60% and nominal tank spills about1.5% of the centre region injected current. This wasnot as good as last year because of the cumulativeeffect of small vacuum leaks and a malfunctioning rfbooster. ISIS and cyclotron beam developments com-bined with skilled tuning by operators were very help-ful. Twelve weeks of shutdown left 6088 scheduled op-erational hours of which 5226 were delivered for anavailability of 86%, down 6% from last year, mainly dueto the rf resonator water leak. These totals, including281 hours used for development and tuning, are shownin Fig. 248. The two-week periods preceding shutdownswere dedicated to lower-intensity operation, typicallyinvolving beam delivery to ISAC, with the proton irra-diation facility (PIF) using BL1B and BL2C1 runningin parallel. BL2C1 was also used at 74 μA for ocu-lar melanoma treatments for nine patients during fiveproton therapy (PT) sessions.

As Fig. 249 shows, the total beam charge deliveredto meson hall experiments along BL1A was 434 mAhor 84% of the scheduled amount, similar to last year,

198

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52

0

25

50

75

100

125

150

175

200

225

250

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

5500

6000

6500

7000

7500

����

�����

��

WEEK

WE

EK

LY H

OU

RS

TO

TA

L H

OU

RS

SCHEDULED

DELIVERED

6088 hours

5226 hours

�high current operation low current operation

SHUTDOWN

WINTER

MIN

I-S

HU

TD

OW

N

RE

SO

NA

TO

R W

AT

ER

LE

AK

Fig. 248. Operational hours for 2005.

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52

0

5

10

15

20

25

0

50

100

150

200

250

300

350

400

450

500

550

600

650

700

750

WEEK

WE

EK

LY B

EA

M C

HA

RG

E (

mA

*Hrs

)

TO

TA

L B

EA

M C

HA

RG

E (

mA

*Hrs

)

Delivered 1A Charge

(434 mAh)

Delivered 2A Charge

(130 mAh)

Delivered 2C4 Charge

(158 mAh)

TOTAL Deliv Charge

(722 mAh)

1A Charge

(499 mAh)

SHUTDOWN

min

i-shu

tdow

n

low

cur

rent

del

iver

y

1A tr

iple

t

Res

onat

orw

ater

leak

Scheduled

Fig. 249. Beam delivery for 2005.

199

because of the triplet cooling problems described be-low. In addition to the BL1A charge, there was a record158 mAh delivered at 85 MeV to rubidium targets inthe solid target facility (STF) in beam line 2C4 for theproduction of a radiopharmaceutical generator and an-other 130 mAh (20% more than last year) delivered tothe two ISAC target stations for the production of ra-dioactive ion beams (RIB).

The beam stability in BL2A was markedly im-proved by regulating the beam injected into the cy-clotron via a feedback program between the extractedBL2A current and the ISIS pulser. Previously, thisbeam line had been subject to an unacceptable numberof over-current trips, most of which were inherent inthe method of stripper shadow-splitting of the internalbeam for simultaneous 500 MeV 1A and 2A extrac-tion. This led to instabilities in the intensities of theextracted beam, causing the average 2A beam currentto be significantly reduced from the requested opti-mal values in order to decrease the frequency of over-current trips. On the other hand, BL1A and BL2C4 atnormal operating average intensities could easily tol-erate the increased fluctuations such a feedback mech-anism would produce. Consequently, the pulser duty

cycle could be varied to obtain the desired stability forISAC without affecting the efficiency of beam produc-tion on 1A or 2C while at the same time producing asignificant improvement of efficiency, quality and relia-bility of RIB production. The three proton lines some-times exceeded a total extracted current of 230 μA,although 200 μA or less was more often the case forextended production periods, particularly when BL1Acurrents were limited to 100 μA because of the use ofgraphite targets at 1AT1 for the TWIST experiment.

The annual downtime of 827 hours (Fig. 250) wasnearly double that of last year, with rf problems ac-counting for 584 hours or 70% of the total. Most ofthis (508 hours) was due to the resonator water leak,leaving the remaining 76 hours to rf on-line downtime(about 2 hours per operational week). This contin-ued the excellent on-line performance of the rf systemrecorded last year. The bulk of the remaining downtimebelonged (somewhat equally) to the usual systems:beam lines, services, site power, beam trips, vacuumand power supplies. Operational records and beam toexperiments for the year are given in Tables XXIII andXXIV.

Rad

io F

requ

ency

Beam

Lin

esSe

rvic

esSi

te P

ower

Beam

Trip

sVa

cuum

Pow

er S

uppl

ies

Ion

Sour

ce 1

Inje

ctio

n BL

Infle

ctor

Targ

ets

Mis

cella

neou

sC

ontro

lsSa

fety

Dia

gnos

tics

BL2C

Cor

r. Pl

ates

TNF

BL2A

0

20

40

60

80

100

0

20

40

60

80

100583.55

36.7 36.2534.3529.9

26.0521.6 20.9

9.5 6.8 5.85 5.75 3.4 2.7 2.3 1.4 0.35 0.05 0.05

X 10

TOTAL = 827.45 HOURS

SYSTEM

HOURS

Fig. 250. Cyclotron downtime for 2005.

200

Table XXIII. Operational record for 2005.∗

Scheduled hours Actual hours

Cyclotron off:

Maintenance 470.0 439.10Start-up 171.0 130.40Shutdown 1, 989.0 2,057.00Other 0.0 5.50

Cyclotron downtime 0.0 827.45Overhead 18.0 50.50

Totals 2, 648.0 3,509.95

Cyclotron on:

Development 117.0 130.50Cyclotron tuning/training 465.0 150.95Beam to experiments 5, 506.0 4,944.60

Totals 6, 088.0 5,226.05

Actual / Scheduled = 5,226.0/ 6,088.0 = 85.8% availability

Beam to experiments:

1A Production 4, 282.0 3,884.401A Development/tuning 18.0 24.351A Down/open/no user 639.0 465.95

1B Production 434.0 22.301B Development/tuning 0.0 5.251B Down/open/no user 133.0 542.35

Total 1A+1B production 4, 716.0 3,906.40

2A Production 4, 580.0 3,669.952A Development/tuning 94.0 53.102A Down/open/no user 832.0 1,221.55

2C1 Production/tests 1, 571.0 303.202C1 Development/tuning 0.0 9.602C1 Down/open/no user 423.0 1,223.70

2C4 Production/tests 3, 407.0 2,926.852C4 Development/tuning 2.0 0.752C4 Down/open/no user 103.0 480.50

1A Beam charge (μA h) 515, 840.0 434,105.002A Beam charge (μA h) 161, 616.0 129,670.002C4 Beam charge (μA h) 194, 130.0 158,283.00

∗ There was no BL4 production this year and the polarized source was not used.

201

Table XXIV. Beam to experiments for 2005.

Scheduled Delivered

Experiment∗ Channel Schedule # Hours μAh Hours μA h

614 M13 107 2,579 335,270 2,259.30 267,757614 M13 108 1,114 123,010 1,113.10 117,884744 M9B 108 383 42,110 373.55 39,149881 M15 107 66 8,580 61.75 6,318891 M15 108 64 8,320 61.30 7,530891 M20B 107 135 17,550 141.35 15,083895 M15 108 69 8,970 68.65 9,019895 M15 108 69 8,970 70.25 8,355917 M20B 107 133 17,290 137.35 17,559938 M15 107 287 37,310 242.45 31,862938 M20B 108 283 32,290 280.55 31,012939 M20B 108 173 22,490 176.30 20,943944 M15 107 127 16,510 80.40 4,462945 M20B 107 150 19,500 101.05 7,409945 M20B 108 146 14,600 140.95 13,331949 M9B 108 150 15,000 148.25 14,249953 M20B 107 137 17,810 128.90 17,015976 M15 107 92 11,960 96.00 11,990976 M15 108 219 23,970 219.25 23,482998 M20B 107 130 16,900 129.45 14,326998 M15 107 130 16,900 129.45 14,326999 M20B 107 141 18,330 129.90 14,6341000 M15 108 142 14,200 147.35 14,1731001 M20B 107 127 16,510 0.00 01001 M20B 108 127 16,510 132.30 15,5641002 M15 107 137 17,810 0.00 01002 M15 108 156 15,600 148.45 14,6661004 M15 108 146 14,600 140.95 13,3311006 M20B 107 150 19,500 113.55 14,8471006 M20B 108 156 15,600 148.45 14,6661012 M9B 108 142 14,200 147.35 14,1731013 M15 107 69 8,970 79.60 8,7651013 M15 108 58 7,540 60.45 6,5031015 M15 107 150 19,500 147.45 20,0951018 M20B 107 137 17,810 116.70 12,9301032 M15 107 141 18,330 129.90 14,6341033 M20B 107 150 19,500 148.80 18,0411035 M9B 107 287 37,310 242.45 31,8621035 M9B 108 133 13,300 134.40 13,6161038 M20B 107 150 19,500 147.45 20,0951046 M15 108 133 13,300 134.40 13,6161047 M15 108 58 7,540 62.05 7,2091049 M20B 108 133 13,300 134.40 13,6161050 M20B 108 142 14,200 147.35 14,1739999 M9B 108 306 38,400 309.55 36,697TBA15 M15 107 191 24,830 188.45 23,993TBA20 M20B 107 150 19,500 147.10 18,424ISAC† 2A2/3 107/108 4,580 161,616 3,669.95 129,670ISOPROD 2C4 107/108 3,407 194,130 2,926.85 158,283PIF 1B 107/108 434 0 22.30 0

202

Table XXIV (cont’d.)

Scheduled Delivered

Experiment∗ Channel Schedule # Hours μAh Hours μA h

PIF 2C1 107/108 1,229 0 281.20 0PT 2C1 107/108 342 0 22.00 0

∗ See Appendix D for experiment title and spokesman.† Total proton beam on ITW and ITE for all ISAC RIB experiments and tests.

Winter Shutdown

Shutdown activities got under way early in the newyear with the removal of some 60 large shielding blocksfrom the meson hall in preparation for a number ofBL1A activities. These included: repair of two waterleaks, replacement of insulators, and back-flushing thetriplet quadrupole 1AQ15; repair of vacuum leaks atgate valve 1AVA8 and at the M15 slit drive mechanism;maintenance of the 1AT1 and 1AT2 water packagesand the T2 beam blockers; repair of thermal switchesand water leaks at 1AQ9/10/11; replacement of mon-itors 1AM9 and 1AM8; repair of an M9BQ3 shortedcoil and refurbishment of the 1A exhaust filter hous-ing. An earlier decision had postponed the complexrefurbishment of the M20 muon channel until 2006.However, this did not offer much relief because someof the remaining tasks proved to be both dose and timeconsuming. For example, repair of the 1AQ10/11 dou-blet was not originally anticipated; it developed into awarm-cell operation.

BL1A activities got under way while the cyclotronwork was deliberately delayed for a few more weeksof cool-down. Before the lid was raised in mid Jan-uary, the maintenance of jack station 7 in the cyclotronvault, as well as resonator draining and vacuum tests inpreparation for rf chore-pad replacements, were takencare of.

When the lid was first raised surveyors found thetank to be contaminated in a fairly confined regionaround the west end of the south cryopanel and imme-diately below the 500 MeV extraction point for BL1A.Initial dry swipe readings of 500,000 cpm were reducedto less than 1000 cpm after two cleanings of the mostcontaminated surfaces. It was not considered worth-while to decontaminate the rest of the limited area. Ac-tivity trailed off sharply within a metre but there weresome associated high levels (∼10,000 cpm) along theL arm of the extraction probe. The contaminant wasfound to be Be7 and, although not entirely understood,evidence pointed to it being driven off overheated ex-traction foils before condensing on the colder surfacesbelow. This possibly, but not necessarily exclusively,involved one with a BeCu holder that was being tried

for the first time. Overheating may have resulted fromhigher beam densities associated with beam-shadowingtechniques or from the lower heat emissivity of theBeCu foil holder. Therefore it was decided to avoidthe use of the latter until the problem was better un-derstood. An area of 3 m radius was flagged off, insideof which respirators and additional protective cloth-ing were required. The same precautions were imple-mented for work on the X1 probe when retracted fromthe tank.

Once the lid was up the following tank work tookplace: the Diagnostics group aligned the LE1 track andinstalled a new LE1 probe head; serviced extractionprobe 2C; replaced both periscope prisms; rewired theNW periscope; emptied the foil buckets and loadednew foils in all three extraction probes; checked outslit and vertical flag drives; replaced the entire HE3probe and performed commissioning tests. The RFgroup cleaned correction plate (CP) wire way insula-tors and installed new covers; realigned the CP2 uppertray; repaired broken insulators on the CP4 lower tray;replaced chore pads in LQ2, 3 and 4; performed cen-tre region inspections; made hot arm tip adjustmentsto the #10 resonators and checked tank thermocou-ples. Operators drained and leak checked resonators;assisted with CP wire way work, tank inspections andperiscope work, and checked the CP continuity. TheVacuum group repaired leaks at the south turbo back-ing valve and cryo-pump 2 gate valve; checked the 2Aextraction probe housing thermocouple gauge and in-spected the tank seal. The Remote Handling group in-stalled and removed shadow shields, pumping port andspill-monitor covers and copper blockers in addition toassisting probes work, tank vacuuming, video inspec-tions and safety surveys. Other work included inflectorMRO and the adjustment of the position of a 2C exithorn copper blocker. One week after the tank work wasfinished (mid February), the lid was lowered for beamdelivery to BL2A while the shutdown work continuedon BL1A.

Work in the cyclotron vault included the installa-tion of additional safety railings on the BL1A/2C over-head walkway and on the stairway to the tank lid; theinspection of trim and harmonic coil junction boxes;

203

the replacement of active filters; the replacement ofphase probe cables and all lower correction plate ca-bles; the preparations for a new 2C4 monitor; the re-plenishment of gas in the vault profile monitors; thereplacement of 1VM2 signal wiring; maintenance onthe rf transmission line (capacitor station 2 modifica-tions, gasket replacement and leak repair at the cou-pling loop) and the replacement of O-rings in the vaultbeam lines.

Start-up

Several minor and one more important air leak (atthe X2A gate valve) were dealt with during the cy-clotron tank pump-down and bake-out but early inMarch it became apparent that the tank vacuum wasnot as good as it should be. Subsequent tests indicateda resonator water leak on the south side. Before raisingthe lid a quick check was made to confirm that beamcould be tuned through the cyclotron. The resonatorswere then drained and further tests determined thatthe problem was in the cooling circuit in upper quad-rant 3 (UQ3). The lid was then raised to pinpoint theexact location of the leak (through manual and remoteleak-checking). This eventually led to the discovery ofa bad weld partially hidden in a hot-arm cooling chan-nel of a new-style resonator segment. The segment wasthen removed from the tank, repaired by replacing thefaulty cooling panel and reinstalled. After recovery, in-jection took place and beam was quickly tuned throughthe machine at the end of March, about three weeksbehind schedule. The extra dose incurred by this in-tervention was about 30 mSv, bringing the shutdowntotal to 168 mSv shared by 112 workers. This was con-sistent with other long and involved shutdowns in re-cent years with the various groups working closely withSafety, planning their work carefully and including theuse of volunteers where possible.

Beam Schedule 107

Once the cyclotron had recovered, beam deliverygot under way, first to BL2A and 2C1 and then, afterthe completion of shutdown activities there, to BL1Aas well. One area of concern was an apparent blockagein the lines cooling the innermost section of the 3L1resonator segment as evidenced by high temperaturereadings (130◦C) for the electrode, skirt and rib ther-mocouples. Back flushing efforts had failed to dislodgewhatever may have been restricting the flow but in midMay, small perturbations in the resonator water sup-ply pressure caused by manual operation of the three-way valve (after an rf trip) cleared the cooling channelwith no subsequent problems. The cyclotron availabil-ity for this schedule was only 82% of the 4360 sched-uled hours largely due to the 508 hour turnover timefor the resonator water leak. (Once running, the avail-

ability was a decent 93%.) The cyclotron itself faredreasonably well although several small air leaks (onein the vicinity of the extraction 1 gate valve, one ata tank feed-through and a third, eventually repaired,in the 2A vault section) kept the south side vacuum atroughly double its usual value. Also, the rf booster wasnot yet fully commissioned after undergoing a controlsupgrade and was therefore not used. These two fac-tors combined with the increased demand from ISAC,boosting the circulating current at times to 230 μA, re-sulted in a somewhat poorer transmission and higherspills than usual. Downtime averaged 10 hours a weekfor the 25 weeks of operation after the resonator re-pair, with a couple of copper active low conductivitywater system leaks near the discharge side of one of thepumps as well as a premature B-20 change and a siteUPS failure being the main interruptions in an other-wise fairly steady production period. In addition to thehigh intensity beams produced for the proton lines asdiscussed below, there was also low current delivery toBL2C1 for both PT (three sessions, six patients) andPIF (for two weeks) to provide some cool down at theend of the delivery period. BL2C1 currents were lessthan 7 nA while BL1B saw limited use at currents ofless than 1 nA. The problem with tuning 2C1 at en-ergies above 105 MeV remained unresolved in spite ofthe belief that a repositioned exit horn blocker wouldbe the answer, so the users had to be content with105 MeV and lower.

BL1A ran for 2624 hours or 88% of the scheduledtime and received 302 mAh or 79% of the scheduledcharge. The low numbers can be mainly attributed tooverheating problems with the 1AQ14/15/16 triplet,which caused around 180 hours of 1A downtime aswell as a day of cyclotron downtime over the courseof three separate back flushing procedures. BL1A cur-rents were kept somewhat lower than the 130 μA sched-uled, partly as a conservative measure with respect tothe cyclotron vacuum but primarily due to the BL1Atriplet quadrupoles that were gradually detuned to suc-cessively lower values to keep them from overheatingwhile the 1A beam current was correspondingly low-ered to limit the associated beam spill. On the positiveside, the 1A vacuum remained very good throughoutthe schedule, thanks in part to new style seals installedin 1AVA8 during the winter shutdown.

BL2A ran for 2684 hours or 78% of the scheduledtime, at currents up to 65 μA for significantly extendedruns that contributed the bulk of the 85 mAh totalcharge. Beam was generally available on demand. Thelower availability reflects those times when RIB wasnot needed due to target problems or changeovers ortuning using stable beam from OLIS. This was the bestperformance BL2A has had so far both in charge and

204

stability, thanks in part to the use of the pulser feed-back program described above. However, tuning wasstill touchy at times due to the competing constraintsof a larger beam size to ensure target longevity ver-sus a smaller beam size to maintain cooler collimatortemperatures.

BL2C4 ran for 2088 hours or 81% of the scheduledtime receiving 111 mAh or 77% of the scheduled chargeat currents around 60 μA. Delivery here requires thatat least one other high current beam line is running sooccasionally there was a holdup. However, the beamline and the STF ran quite well with one exception,namely the signs of a possible small target rupture asevidenced by cooling water conductivity, activity andkrypton content as well as a slight excursion of an as-sociated air monitor. There was no indication of a leakupon examination and processing in the Nordion hotcells although an anomaly in the target fabrication wasnoted. This process was reviewed before approval wasgiven to resume operation.

Fall Shutdown

A cool-down prior to the fall shutdown wasachieved by the scheduled early termination of BL1A(switching to 1B), a reduction in the current to BL2Aand a switch from 2C4 to BL2C1 for low current PIFoperation. The busy 1A triplet maintenance schedulegot an early start due to early finish of BL1B oper-ation. There was no requirement to raise the lid butthere were plenty of jobs to fill the time allocation.

The Diagnostics group checked the water flow of thewater-cooled probe, installed the new 2C4 harp moni-tor, gassed vault and 1A monitors and loaded extrac-tion 1 foils. The Vacuum group leak-checked the 2Afront end, replaced turbopump 2AVP1 and replacedO-rings at 2AVM1. The ISIS group checked out skim-mer 3 and replaced cryo-pump 210. The Plant Ser-vices group rebuilt the south CuALCW pump. TheRF group installed and tested the new hairpin in PA4and checked the capacitor station 2 in the vault base-ment. The Controls group installed and tested a newUPS. The Remote Handling and Beam Lines groupscontinued the 1A triplet work and installed new flowmeters. The Targets group worked on the 1AT1 and1AT2 water packages and the Power Supplies groupfixed a water leak in the main magnet power supply.

Most jobs proceeded as planned except for the 1Atriplet work which was completed well ahead of sched-ule for two reasons. One was the early start as men-tioned above. The other resulted from testing differ-ent back-flushing techniques and finding that generalchemical flushing (straight water would not work) wasso successful that it was not necessary to dig downand isolate and flush the few individual rogue circuits

known to overheat. This not only saved time and dosebut held promise that any such interventions requiredin the future would have similar savings. To offset thisgood news a CuALCW water leak up to 60 l/h wastraced to M20Q1 and an additional week was requiredto uncover it, repair a Hansen fitting and cover upagain. This still left BL1A available near the begin-ning of the start-up period to tune the cyclotron. Thisis the preferred mode for cyclotron tuning. It also pro-vides the least disruptive way to bring on the otherproton lines. So start-up was fairly smooth and beamproduction continued as described below.

Beam Schedule 108

The cyclotron availability for this period was 1648hours or 95% of the 1728 scheduled hours, a very goodfinish for the last twelve weeks of the year. The aver-age weekly downtime was halved to 5 hours, the bulkof which was fairly evenly distributed among the RF,ISIS and Services groups as well as site power distur-bances. It should be noted that the rf behaved ex-tremely well with an average weekly downtime of only1.5 hours. The rf booster was eventually tested andready for use but soon developed a water leak at itsfeed-through, rendering it inoperable for the remain-der of the year. The previously degraded tank vacuumwas further challenged by a growing air leak in theBL1V section that was beginning to seriously compro-mise cyclotron performance. The problem was locatedon a vacuum seal at the beam blocker and could be re-paired (at a cost of 1.5 mSv) allowing the resumption ofnormal operation. The cyclotron tune was fairly goodwith a typical beam transmission of 62% and nominaltank spills around 1% of the total current extracted.

BL1A ran for 1260 hours or 97% of the sched-uled time, at an average current of 105 μA for a to-tal charge delivery of 132 mAh or an impressive 97%of that scheduled. However, the currents were for themost part kept reasonably low for a number of reasons.Not long after start-up the initial higher current wasreduced to 100 μA due to the early use of a graphitetarget at 1AT1, when a misalignment on the newly in-stalled, 4 mm low ladder precluded the use of berylliumtargets there. The previous ladder was swapped backand currents were raised to 120 μA for two weeks beforebeing reduced again to 100 μA for the final few weeksof the year’s operation. The temperature of the trou-blesome circuit of 1AQ15 started to climb and wouldnot have remained below the trip point much longerwithout requiring a back flush. By reducing its currentand retuning BL1A the quadrupole remained usableuntil the end of the schedule.

BL2A ran for 986 hours or 88% of the scheduledtime, at currents up to 70 μA during a significant ex-

205

tended run that delivered a record charge of 31.8 mAhaccumulated on one of the SiC targets. The total de-livered charge this quarter was 45 mAh, again aidedby the use of the beam stabilization program. Tuningwas made more difficult toward the end with the lossof the final scanning wire monitor but constraints onbeam size were achieved using the halo monitor.

BL2C4 ran for 839 hours or 99.5% of the scheduledtime receiving 47 mAh or 93% of the scheduled chargeat currents around 60 μA. Some time was spent inves-tigating a second suggested target blip which turnedout to be related to water conductivity problems whichwere becoming somewhat chronic until some relief wasobtained by changing the resin. Also, tunes toward theend of the year seemed to suffer a little from widerspots due to aging extraction foils.

BL2C1 was used for proton therapy (2 sessions, 3patients) as well as for PIF experiments for a total ofabout 3 weeks of operation. The tuning problem athigher energies was resolved when a similar problemcropped up at lower energy. The problem had beencaused by a shorted polarity switch in the combina-tion magnet power supply. The switch was replacedrestoring the ability to run the higher energy 116 MeVtune for PIF.

BL1B had been scheduled for PIF, however, dur-ing some tuning subsequent to 3 hours of initial op-eration at 225 MeV, there was difficulty finding thebeam after a power glitch. In the process the extrac-tion probe was dipped too low into the beam and wasdamaged and rendered unusable due to overheating, asconfirmed during an inspection several days later. In-dependently the probe itself had already been sched-uled to have most of the damaged parts replaced in theupcoming shutdown. PIF irradiations at higher ener-gies (200 to 500 MeV) had to be suspended but someusers were able to work at the lower energies offeredby BL2C1.

Just before the Christmas holidays there was a two-day development shift after which most cyclotron sys-tems were turned off. The year ended as it began withthe removal of around 60 large shielding blocks fromthe meson hall in preparation for the next round ofshutdown activities.

Apart from the many tasks associated with beamdelivery and shutdown work, individual operators wereagain involved with several other important activitiessuch as fire alarm and card access system improve-ments, site emergency planning, AutoCAD drawings,training, equipment repair, computer and console up-grades and maintenance, beam transport modelling,coordination of software improvements and more.

BEAM DEVELOPMENT

Cyclotron

Cyclotron beam development focused on increas-ing TRIUMF’s extracted current for beam productionfrom its existing level of ≈200 μA to ∼300 μA, on sta-bilizing the BL2A beam current extracted for ISAC,and on achieving/maintaining high quality tunes forroutine operation.

In 2005 realignment work of the centre region cor-rection plate was completed. Begun in 2002, this workwas aimed at reducing beam-induced heating on thecorrection plate scrapers. The plates in upper quad-rant 2 were surveyed, found to be tilted with respectto the median plane, and realigned. Afterward, whileextracting 298 μA, all scraper temperatures stabilizedbelow 50◦C, only slightly above their ambient beam-off temperature of ≈40◦C. This was a very importantachievement in the effort toward reducing beam losseson indirectly cooled electrodes in the centre region,as required for the higher intensity beam goals of cy-clotron refurbishment.

ISAC prefers to operate with currents close to, butnot above, the target’s limit. Thus small fluctuations inthe extracted current may produce over-current trips.In order to avoid sudden changes of thermal load onthe target, a slow ramp-up of the beam-pulser dutycycle is introduced delaying as much as two minutesthe restoration of the required conditions of the pri-mary beam. After this, the time required by the tar-get to reach its stable production regime for RIB isnormally at least as long. Fluctuations of the primary-beam current were substantially reduced by installinga BL2A/ion-source-pulser feedback loop. If the BL2Acurrent increases, then the loop compensates for theincrease by changing the setting of the pulser to de-crease the duty cycle of the injected beam and viceversa. This loop is now being used for routine beamproduction and has significantly decreased the numberof trips.

ISAC’s proposed BL4N extraction line, which wasbased on an increased total cyclotron current require-ment up to 400 μA, was removed from the 2005–2010five-year plan because of budget reductions. As a re-sult, in 2005, development concentrated on optimiz-ing the tune for 300 μA needed as soon as ISAC be-gins operating with 100 μA. Much of this work wasdone at high equivalent currents with less than fullduty cycle, because during most of the developmentshifts one or more of the extraction lines was unavail-able for use as a beam dump. Still, valuable experi-ence was obtained, and when the three high-intensitybeam lines became available in October, TRIUMF’sold total extracted current record of 275 μA set in 2002

206

was broken. 298 μA was extracted at 97.2% duty cy-cle for 2.4 hours with 60% cyclotron transmission. Theduty cycle was then reduced so that the stabilizationloop could be turned on, and 274 μA was extractedfor 1.4 hours. As shown in Fig. 251 the extracted cur-rents were fairly stable, and turning on the feedbackloop improved the stability of the current deliveredto BL2A. Although these results are promising, fur-ther refinements of the tune and/or the hardware maybe required before we will be able to reliably deliver300 μA to experiments for weeks or months at a time.

Development time was also used to diagnose andcorrect problems that arose during beam production.Routinely at the start of each development shift, a com-plete set of diagnostic scans is done and existing prob-lems are corrected before proceeding with the sched-uled development. Figure 252 shows one such problemthat was corrected. The HE1 probe is scanned whilethe time-of-flight between cyclotron injection and ex-traction is recorded as a function of the HE1 radius.

Fig. 251. Graph showing simultaneous extraction of high-intensity currents versus time with and without the BL2Apulser stabilization loop.

Fig. 252. HE1 V cos φ scans showing how the isochronismwas improved near 400 MeV by adjusting the trim coils.

This yields a measure of the product V cosφ. The dipnear 400 MeV, indicating poor isochronism, was cor-rected by adjusting the trim coils.

The development time allocated to eliminate in-stabilities was particularly important. Early in 2005an instability causing inflector trips was eliminated byadjusting the beam steering at the inflector entrance.In the fall a persistent intermittent instability causingBL1A target-protect monitor trips was eliminated byadjusting the centre region correction plates and trimcoils. Although the solutions appear simple, problemssuch as these are often difficult to diagnose/solve andfrequently require a large amount of development time.

Primary Beam Lines

In 2005, a new BL2A tune was developed for ISAC.The old tune produced a dense 1 mm × 2 mm spotwhich melted holes in the target. The new tune hasbeen operational since April, 2005 and produces a lessdense 12 mm spot (“size” being here defined as 4 timesrms). It has successfully delivered 70 μA to ISAC’s westtarget station over long periods of time. Post-mortemanalysis revealed no localized melting of the target. Akey factor in developing the new tunes was reportedlast year – the discovery of an incorrect descriptionof the beam transport from the stripper foil throughthe fringe field of the cyclotron to the beam line. Sincethis discovery, we now achieve very good agreement be-tween measured and calculated beam envelopes. Thisis shown in Fig. 253.

Spill-monitor trip limits are set at around 1 nA/malong the beam line. At highest current running, we arethus sensitive to halos at the 10−5 level. The observedspills are around this level and are consistent with largeangle Rutherford scattering from the carbon extractionfoil. In order to minimize the spills, the transfer matrixelements M12, M16, and M34 of the new tune betweenthe stripping foil and the entrance of the ±15◦ dipole

Fig. 253. Measured and calculated BL2A beam envelopes.

207

on BL2A were all made small because the beam pipedecreases from a diameter of 4 in. to one of 3 in. at thedipole entrance. Although this proved helpful, for highcurrent it would be desirable to eliminate the halos byinstalling a collimator in the vault near the entrance ofBL2A. Additional studies will be needed to determinefeasibility and design configuration of the collimation.

Plans for 2006

Commissioning BL2A for 100 μA operation will bethe primary goal of beam development during 2006.Cyclotron development will concentrate on stabilizingthe extracted BL2A current and on identifying thesources of the instabilities, such as power supplies re-quiring additional regulation etc. while BL2A develop-ment will concentrate on finding tunes with acceptablespills that produce the required spot sizes on the tar-gets.

As in previous years, some time will also be devotedto high-current tests and to maintaining/improvingproduction tunes.

RADIO FREQUENCY SYSTEMS

RF Operation

The total cyclotron rf downtime for the year was76 hours, excluding the 3 weeks required following theshutdown to find and repair a resonator water leak.Downtime due to sparking was reduced by a factor oftwo from that of the previous year, accumulating only19 hours in total. Crowbars caused 13 hours of down-time with the remaining 44 hours caused by varioushardware failures. These are described in detail below.

As in the previous year, a great deal of activity wasdevoted to enhancing overall reliability. A number ofimprovements were made in various areas from poweramplifiers (PAs) to resonators.

Power Amplifiers

A thorough inspection and overhaul of most com-ponents in the PA cabinets took place during the shut-down. The replacement program for water-cooled fila-ment adapters was completed for all PAs. All eight fil-ament power supplies were overhauled. Of these, threeolder supplies were upgraded with new transformersbecause they were underrated for output voltage. Alleight PA screen power supplies were overhauled as well.

Five spare 4CW250,000 tubes, three new, one re-built and one old, were tested in PA3 and PA4. Allof them showed good and reliable performance. Inputmatching to all PAs was significantly improved.

Tube emission tests were performed in PA1, PA2and PA4. The main conclusion drawn from the mea-surements is that there was no emission degradation forthe entire tube lifetime in the PA operational regimein use.

The PA1 hairpin shorting plate showed very poorrf contact, which caused overheating. The original me-chanical design was found to be deficient; eroded anddeformed hairpin copper pipes did not allow proper rfcontact to the shorting plate. As a temporary measurethe rf contact was improved by using a 3 mm thickaluminum wire. At the same time, a new PA hairpininductor prototype was designed, built and installedinto PA4 (see Fig. 254). From an rf point of view thisis a replica of the old inductor, although it featuresa stainless steel copper-plated body, improved anodecontact electrodes, and a modified tuning (shorting)plate. This new unit has demonstrated excellent per-formance, but was found a bit too complicated to fab-ricate. An updated hairpin version was subsequentlydeveloped in which the expensive anode boxes wereeliminated. The new design was optimized through ex-tensive use of the HFSS 3D simulation package. Signif-icant simplification of the structure and suppression ofsome parasitic rf modes were obtained. New hairpinsare being fabricated and are scheduled for installationin 2006.

Anode HVPS

Many HVPS varistor disks were in bad conditionand required a thorough cleaning and polishing. An en-tire circuit was upgraded. Each of four varistor stackswas expanded from 24 to 30 disks, increasing the totalenergy-discharge capacity by 25%.

The continuity test circuit for the tube cathodeshunts was upgraded with test pulse generators in-stalled for each PA tube. Many modifications weremade to the crowbar driver front-end electronics, in-cluding crowbar triggering by Pearson current trans-formers and diagnostics added to detect the firing se-quence of the crowbar elements. There is also a newcrowbar counter that employs a binary output inter-face. All of these new signals are sent to the main con-trol system for remote display and data logging. In

Fig. 254. New stainless steel hairpin inductor during in-stallation in PA4.

208

addition, a signal from a HV resistive divider wasadded to the main control system for remote high-voltage reading, and a Hall-effect transducer was in-stalled to measure the HVPS total output dc current.

RF Coaxial Switch

The rf amplifier system provides up to 1 MW ofrf power at the output of combiner #3 and is con-nected to the cyclotron through a 9 in. transmissionline. To simplify troubleshooting, a 4 port 9 in. coaxialswitch has been installed at the output (see Fig. 255).This gives us the ability to redirect rf power to eitherthe cyclotron or the dummy load. Originally rated to700 kW, this switch was upgraded to carry 1.1 MW byadding water and air cooling. The switch is mounted ona stand which features height adjustment and perma-nently mounted casters for relocation. After one year ofoperation all movable rf finger contacts were inspectedand found to be in very good condition.

Soda Solution Dummy Load

Our dummy load set-up was upgraded with a new1 MW heat exchanger. A new conductivity sensor wasalso installed to aid in the monitoring of rf matchingat the input port of the dummy load. The rf termina-tion head received a new vertical stand (see Fig. 255),allowing easy adjustment of the height of the load andmobility required to connect it to either the rf switchport or any output of the PAs or combiners. The com-plete system was tested at 800 kW of rf power for afew hours.

Transmission line

A complete overhaul and cleaning of all spark andburn spots was made on the transmission-line match-ing section. Six new aluminum inner-conductor jointsleeves were installed. These replaced old copper ones

Fig. 255. RF switch installed into the system.

Fig. 256. Burnt spot at transmission line inner conductorjoint.

that had failed in a few locations because of elec-trochemical corrosion (see Fig. 256). Also, eight newpolypropylene disk spacers have replaced old unreli-able insulators.

New centre conductor “T” junctions were designedfor capacitor stations 1 and 2. A set of fabricationdrawings was issued with installation planned for 2006.The new design features simplified assembly, improvedelectrical contacts and rationalized cooling circuit in-lets and outlets.

Cyclotron

The chore pad replacement program was com-pleted. These protect the bellows of the resonator cool-ing circuits from excessive vibrations. In the wintershutdown the last 30 resonators in lower quadrants(LQs) 2, 3 and 4 were serviced. Most challenging wasLQ3 because of significant radioactive contaminationin this octant. Because of the contamination the use ofrespirators was necessary.

Thirty-two new correction plate (CP) HV cableswere run from the power supplies to the tank lowerfeedthroughs and the old cables were removed fromthe vault basement. There was a plan to refurbish in-tank CP wire ways. However, during the shutdown itwas decided to limit this intervention only to cleaningthe ceramic insulators and covering them for protec-tion against multipactor discharge (see Fig. 257).

We continued the alignment of the vertical positionof the CP in order to reduce beam losses on the protec-tive copper scraper and thus eliminate the overheatingthat had been observed in the past. During 2005 theset in upper quadrant #2 was raised by ∼5 mm towardits nominal position. Reference measurements with re-spect to the median plane for the final locations of allCP scrapers are shown in Table XXV.

209

Fig. 257. Correction plate wire way insulators coated dueto discharge.

Table XXV. Vertical coordinates (in mm) of the correctionplate scrapers.

LQ2 UQ2 LQ4 UQ4

Inner radius −24.97 22.19 −22.74 26.54Outer radius −25.43 22.19 −23.94 30.06

The ceramic window of the coupling loop developeda vacuum leak at the end of 2004. Failed O-rings werereplaced. At the same time damaged rf finger-stockcontacts were repaired.

In the 2004 winter shutdown the hot arm tip (HAT)positions had been adjusted for resonator segments#10 in LQ2, UQ2, and UQ4, in order to compensatefor misalignment of the corresponding ground arm tips.These lack remotely-controlled motors because of inter-ference with the main magnet yoke. Following these ad-justments no visible improvement was observed in theleakage field pattern. Instead they caused overheatingof the attached resonator panels. Thus it was decidedto return these resonators to their average vertical po-sition in line with the rest of the HATs.

While the cyclotron was being pumped down fol-lowing the shutdown, a water leak was detected in thetank volume. This was located on the copper skin of the3U6 resonator facing the beam gap. The segment wasremoved from the tank for repair, which took place inthe remote handling area. The entire skin of the failed3U6 unit was replaced with a new type of resonatorsegment. The downtime incurred by this failure wasabout 3 weeks.

RF Booster

The rf booster was not used after the installation ofa new rf control system during the 2005 winter shut-down because of its unsatisfactory performance. Theanode blocking capacitor failed and was repaired twicebefore it was realized that there were deficiencies inboth the new control system and the HVPS configura-tion. These deficiencies were corrected.

A high standing wave protection circuit was im-plemented in the control system. Trip points in theovercurrent protection circuit and 480 VAC breakerwere readjusted. LC filtering and resistive current-limiting networks were modified. An SCR-based ramp-ing/switching circuit was also fabricated and installed.In addition, a spark gap was installed in the PA to pro-tect the blocking capacitor. These measures allowed areduction of the energy dumped into a PA spark from15000 J to ∼80 J.

RF Diagnostics

All 12 reverse power directional couplers installedon the 9 in. transmission line in the rf room were re-furbished to accommodate higher coupling. This newrequirement was set to satisfy the input capacity of thenew high VSWR protection unit to resolve low level re-verse power signals. The maximum coupling achievablewas increased from −80 dB to −65 dB (−60 dB for thecombiner waster loads).

The transmission line standing-wave protectionunit was connected to the main rf system in a testmode in order to adjust the trip-point thresholds forall 12 reverse power signals from the rf room transmis-sion line branches. Final commissioning of the systemis scheduled for 2006.

The data acquisition system for PA dc voltagesand currents was commissioned and calibrated. Thissystem helps daily performance checkups, PA trou-bleshooting and tuning, and data archiving to monitorlong-term trends. A number of tests were required tosuppress rf interference for Hall-effect current trans-ducers; eventually they were enclosed in aluminumshielding boxes. RF filters were installed for all powerand signal leads connected to the sensors. More thanfifty dc signals describing the PA status are now avail-able through the control system.

RF Support

Activities of the RF group in ISAC are reported inthe ISAC Section.

RADIO FREQUENCY CONTROLS

Cyclotron RF

Software and hardware to support the high VSWRprotection system was installed and tested. Further ad-justment to optimize trip points is required before thesystem can be fully commissioned. Further fine tuningof the spark detection/recovery system was completed.Debugging and testing of the new ISIS control com-puter and Firewire interface were completed. The newsystem replaced an earlier system that was obsoleteand prone to failure.

210

A new crowbar design was implemented on themain HVPSU for the high power rf amplifier to thecyclotron. This new design uses Pearson coils for sens-ing overcurrent transient conditions. The previous de-sign, which had been in use for the past 20 years orso, sensed overcurrent transients in the ground returnresistor and was more prone to generating false crow-bars because of noise on the ac supply from BC Hydro.Thus the new design results in less machine downtimeand higher availability because fewer false alarms aregenerated by sensing the actual HV supply to the rfHPA.

RF Booster

The rf booster power amplifier was found to havemajor phase non-linearities, which interfered with nor-mal operation of the IQ control system. As a result,a major redesign of the booster controls to a con-ventional amplitude/phase based system was under-taken. The resulting operating mode is similar to thatused for the ISAC-II superconducting cavities. Thebooster is started in self-excited mode, then tuned towithin 2 kHz or so of the fourth harmonic and am-plitude/phase locked to the cyclotron reference. Al-though testing and commissioning were constrained tobiweekly shutdowns, they were completed before hard-ware failures in the cavity tuning and transmission lineforced a halt in operation until repairs could be madeduring a shutdown.

ISAC-I

A system was designed and implemented to permitpretuning of the RFQ. This is a necessary prerequisitefor full turn-key operation of the system. The ground-work was also undertaken to begin replacement of thecommercial tuning-motor drive systems with a morereliable in-house design. Design modifications to im-prove the stability of the rf systems of the DTL tanksand bunchers were implemented.

ISAC-II

A major effort – not unlike that of starting up asmall manufacturing line – was required to producethe 20 rf control systems plus spares that this linacrequires. The differing requirements of controlling su-perconducting cavities together with the obsolescenceof a number of key components of the systems usedpreviously forced a complete redesign of the rf controlsystem to a very tight schedule.

A test system was developed to permit measure-ment of very low levels of microphonics in the super-conducting cavities. Any microphonics would generateFM modulation of the cavity rf, so the test systemconsists of a low noise FM discriminator in which the

output voltage varies according to the difference be-tween the instantaneous frequency of the cavity rf andthe reference rf.

Another system was developed to allow data collec-tion, logging, and diagnostics from the operating cry-omodules.

CYCLOTRON PROBES AND BEAM LINEDIAGNOSTICS

Probe Developments

Substantial progress was made on the high-energyprobe (HE) refurbishment. The new HE1 replacementassembly was completed and commissioned in the lab-oratory. An endurance test of several thousand cycleswas made to ensure that there were no problems re-lated to extensive use, such as cable stretch, etc. Thenew HE1 head (Fig. 258) will provide 7 vertical fingersas opposed to the old probe head that provided 5 sig-nals. The probe will be installed in the cyclotron in thecoming winter shutdown (2006). Also, the HE3 probewas upgraded for improved mechanics and electronicsas will be described below.

Fig. 258. Photograph of the HE1 probe head assembly.

211

The group now makes use of the site-standard, com-mercial design software program “Solid Works”. Newconceptual designs are modelled and developed usingthis tool, and then turned over to the Design Office fordetailing.

A planned modification to the low-energy (LE)probe bearing subassembly for installation in the win-ter shutdown was abandoned due to interferences withan rf shield plate. The original recirculating ball-bearing system was cleaned up and reassembled. In thelong-term, replacement of the segmented tracks witha continuous track, which is now feasible with modernmachining tools, may be required. New LE probe headswere designed to lighten the load on the mechanics ofthe cantilevered head. Shown in Figs. 259 and 260 arethe new heads that will be installed in the 2006 shut-down. Figure 259 shows the design model with the up-per rf shield removed; Fig. 260 shows the main partsprior to final assembly. Thin insulating strips will be

Fig. 259. Solid Works model of the LE probe head.

Fig. 260. Photograph of the LE probe head assembly.

Fig. 261. The new wideband head in the cyclotron tank.

used initially to establish alignment and isolation ofthe segmented fingers.

A new intercepting probe head for the TRIUMFcyclotron was designed, manufactured, and installedin the tank during the winter shutdown (see Fig. 261).Mounted with new signal cabling on the old HE4 as-sembly, in the HE3 location, this probe is capable ofmeasuring phase and time structure of the circulat-ing beam. The probe, described in last year’s AnnualReport, has an extended radial range and a newly de-veloped rf strip-line pick-up for fast signal acquisition.New, semi-rigid cables were installed. It is expected tooperate with an average current up to 500 nA and abandwidth in excess of 1 GHz. The signals extractedfrom the probe are processed by a pair of diplexers inwhich low-frequency and high-frequency componentsare separated. The low-frequency signal is directed toour standard electronics for processing and providesboth dc current and a time-of-flight signal with a risetime of ≤0.1 μs. No noticeable problems were foundafter one year of operation.

Several hours in two beam development shifts inMay were dedicated to studies of the broadband func-tionality of the new probe. For the high-frequency out-puts a signal-to-noise ratio of about 4 at 250 nA aver-age current and 0.1% duty cycle was measured in thepresence of rf noise from the cyclotron (see Fig. 262).Time-structure as short as 1 ns was resolved. Althoughthe probe meets design specification and shows goodperformance, the signal-to-noise ratio is relatively low.An especially strong noise signal is induced in theprobe by the cyclotron booster. Additional effort willbe required to further filter out this noise.

In addition, signals from the existing tank non-intercepting capacitive phase probes were inspected.The replacement of cables performed during the shut-down produced positive results. Cables were adjustedfor equal phase delay and amplitude attenuation so

212

Fig. 262. Waveform of the signal from both high frequencyoutputs of the wideband probe (the two upper curves) andthe rf sine wave (fundamental) measured with a 350 MHzscope. The horizontal scale is 10 ns/div.

that rf noise would approximately cancel and the signalfrom the beam would approximately double. Electron-ics for these phase probes are planned to be developedduring 2006.

Probes MRO

The three extraction probes currently in use wereinspected and found not to require servicing. Thewiring of the NW periscope controls was replaced.New stock prisms were installed but they blackenedquickly due to radiation damage. Two different sets ofradiation-resistant prisms were procured; one of eachwill be installed in the 2006 shutdown.

The cyclotron probes use low-pass filters at thetank ports to block rf pick-up from being passed to thesignal processing electronics. The filter design was op-timized using OrCAD PSpice to provide the best risetime for the time-of-flight signal from the new HE1probe.

A new, 24 V dc ferro-resonant power supply for the264 level probes PIE box was assembled, installed inthe rack, and commissioned.

A flow-rate problem of the water-cooled probehas also been investigated. Although no solution wasfound, analysis of its history shows the flow is slow-ing by about approximately 10–15% per year. Furtherinvestigations are necessary.

Beam Line Diagnostics Developments

During the spring it was verified that the proto-type of the new pulse-shaper module for the cyclotron

time-of-flight measurement from the beam line 1A ca-pacitive probe was functioning well. A PCB layoutwas developed for the production version. The num-ber of channels was increased to two. Logic signalsfor the start, stop, and duration of the beam macro-structure were also provided. The first production unitwas assembled and installed during the Septembermini-shutdown.

Currently, the system provides an accurate mea-surement of the beam time-of-flight through the cy-clotron by comparing the arrival time of the beam inBL1A with the injection time at the ISIS pulser. Thedata are now displayed on the operation 0 panel andare included in the shift log. As an alternative to us-ing the rather qualitative oscilloscope display on theconsole, the digital data can now be used by the op-erators to accurately monitor and minimize the time-of-flight and consequently maintain good isochronousconditions in the cyclotron.

New electronics were built for the BL2A capaci-tive probe making it possible to reliably extract dcbeam current information over a wide dynamic rangefrom a narrow bandwidth rf signal with amplitude pro-portional to the beam intensity for currents as low as50 nA. An advantage is that the average current canbe monitored for any, even 100%, duty cycle. There-fore it does not require the extracted beam to show apronounced beam-off interval in the micro-structure.

An indirect goal of the project was to develop ex-pertise and hardware and software bases for “intelli-gent” front-ends that would be flexible enough to bewidely used in different types of monitors with littleor no modification. The approach selected is based onthe Analog Devices mixed-signal ADSP 21992 proces-sor. In the monitor that was developed, demodulationof the rf signal from the capacitive probe is providedby an Analog Devices AD8306 precision log detectorcapable of a dynamic range of 90 dB with analogue-to-digital conversion and temperature compensation per-formed by the DSP. The system was calibrated witha precision Marconi synthesizer and a look-up tablewas generated from the data and stored in the mem-ory to correct for differential non-linearities of the de-tector. A digital low-pass filter was implemented toremove any residual modulation of the output signalof the log detector. Although there are several possibleways of reading the data from the DSP, a 16-bit digitalI/O port was chosen because of the ease of interfacingwith the cyclotron control system. A CAMAC TPG604dual-input register card was installed for this purpose.The module will be installed in the winter shutdownof 2006.

A new multi-wire chamber monitor was designedfor 2C4VM2. New cables and electronics were pro-

213

vided. The monitor boards were tested in the labo-ratory and the monitor was installed in the Septembermini-shutdown and tested with beam. Precision andresolution were found to be very good.

Beam Line Diagnostics MRO

All vault and standard beam line monitors were ser-viced during the shutdowns. Repairs were made to thefollowing monitors: 1AM8 received a new gas pack, aspare unit was installed to repair a ferro-fluid vacuumleak on 1AM9, and new O-rings were installed on thesignal feedthroughs for 1VM1.

The 0794 interlock-logic unit, part of the TNF pro-tect system, was suspected of triggering false error mes-sages. The problem was eventually found to be due toa faulty CAMAC input register.

On BL2A, actuator bellows of the ITW (ISAC tar-get west) entrance monitor (wire scanner 2A2M19) de-veloped a vacuum leak in the fall. It is scheduled forrepair in the 2006 shutdown.

The ITE (ISAC target east) wire scanner 2A3M19exhibited a high level of noise. On investigation, itwas found that strong noise was generated during thescanner motion because of vibrations. After the me-chanics had been adjusted the noise was slightly re-duced. Then the speed was reduced from ∼0.6 s/swingto ∼2 s/swing and the signal-to-noise ratio improvedto an acceptable level. Similarly, the 2AVM2, 2AVM6and 2AM13 wire scanners were slowed down to 2 s oftravel time in order to improve the resolution of mea-surements. It was also found that the wire-scanner soft-ware had errors in the geometric calculation of profileposition. These were corrected.

Increases in proton beam power on the ISAC tar-gets has resulted in damage to targets when the beamdensity was too high. In order to minimize the scatter-ing upstream of the target, the original specificationsfor the target protect monitor intentionally excludeda total plate monitor required for density protection.Recent understanding of the beam line tune and tar-get parameters indicates that scattering is not an is-sue, so a new target-protect monitor was designed andbuilt to provide density protection against small beamspots with excessive power density. It will be installedin ITW during the 2006 winter shutdown. The ITEtarget protect monitor will be similarly upgraded inthe 2007 shutdown.

The BL1A Cerenkov A and B monitors at the1AT1 target were found to have been seriously dam-aged by radiation. The plastic of each monitor hadturned brown and was replaced. Also, the photomul-tiplier tubes of monitors A and B had to be replaced.Both monitors are currently operating well, althoughsigns of rapid deterioration of the light collection effi-

ciency of the plastics is observed, indicating a ratherhigh rate of radiation damage.

A noise problem manifested by the target protectmonitors (2A2M20 and 2A3M20) at low duty cycleswas investigated and was attributed to insufficient in-tegration of the signal by QSX/W amplifiers. The orig-inal amplifiers had a low-pass filter up to 1kHz thatwas sensitive to the time structure of the proton beam.These amplifiers were replaced, for both east and westtarget stations, with QSX amplifiers with a 100 Hzlow-pass filter.

VACUUM

Upgrades and Refurbishments

The vacuum system performed reasonably well dur-ing the year, despite the budget for new equipment be-ing reduced to a bare minimum. Unreliable behaviourof the B-20 cryogenerators was the biggest problemand caused several emergency interventions and re-pairs. This worsening trend clearly showed that refur-bishing the old units was now no longer sufficient andthat the time for purchasing a new cryogenerator hascome. It was recommended that a new, efficient, reli-able modern unit be identified. This unit would meetthe requirements for the majority of the vacuum-tankpumping; existing units would be used only as back-ups or spares. Several options for the replacement unitwere investigated and a request for the 2006/2007 re-furbishing budget was prepared.

In August, 2004 a new Vacuum and Cryogenicsgroup was formed. The existing group was reorganizedto include most vacuum and cryogenics responsibilitiesat TRIUMF. Substantial cryogenic activity took placefor ISAC-II and is described elsewhere in this AnnualReport.

MRO

Some vacuum problems related to air and waterleaks into the cyclotron vacuum tank were experiencedduring the year. After the winter shutdown the ulti-mate vacuum of the tank failed to reach nominal levels(as measured by the north and south ion gauges of thetank). Changing the B-20 cryogenerator and defrostingthe cryopanels did not correct the vacuum situation.Consequently, extensive leak checking was conductedusing a leak detector connected directly to the back-ing line of the south turbo-pump. Only one air leak wasdetected. This leak, on a port on the bottom of the vac-uum tank, was temporarily repaired and replacementof the port flange is scheduled for the winter shutdown.An investigation of the correlation between the pres-sure of the cooling water system and the pressure of thevacuum tank pointed to a water leak as the cause ofthe poor vacuum. This leak was eventually eliminated

214

by the replacement of a resonator segment.The cyclotron vacuum system continued to oper-

ate well for the remainder of the year. A B-20 elec-trical problem and the failure of the gate valve of theinflector cryopump contributed to most of the down-time. A B-20 electrical refurbishment is required; itsinstallation and the commissioning of the new systemis planned during the next winter shutdown.

A leak in the housing of the beam line 1A extrac-tion probe was detected in September. This leak causedextra load on the cyclotron cryopanel system and thusincreased tank pressure. However, the vacuum was ac-ceptable for beam production and repair of the gatevalve could be delayed until the upcoming winter shut-down.

Beam Line 1

The vacuum system of beam line 1 comprises thevacuum in beam line 1V in the cyclotron vault and inbeam lines 1A and 1B external to the vault.

The vacuum system of BL1V operated well duringthe year except for a leak caused by radiation dam-age during beam tuning. The elastomer O-ring was re-placed and its upgrade to a metal seal is scheduled.

During the winter shutdown the gate valve isolat-ing target 1 from target 2 on beam line 1A was refur-bished. The Buna-N gate seal was replaced and metalseals were installed on both sides of the valve. However,the Buna-N seal was damaged again by beam and willhave to be replaced again during the 2006 shutdown.

One of the Leybold turbo-pumps on beam line 1Afailed and was replaced. The backing pumps of theturbo-pumps and blower were serviced or replaced dur-ing the year. Replacement could be done during sched-uled maintenance.Beam line 2A

The vacuum system of beam line 2A had minorproblems. A few O-rings damaged by beam were re-placed. In 2006 some will be replaced with metal seals.A roughing pump in the vault section was removedfrom service and is scheduled to be repaired duringthe 2006 winter shutdown. A l/s turbo-pump failedand was replaced.Beam line 2C

The turbo-pump on beam line 2C failed and it wasnecessary to operate by allowing the cyclotron vacuumto pump the beam line. The failed turbo-pump will berepaired during the 2006 winter shutdown.

ISIS

The cusp ion source, injection line, and inflec-tor/deflector systems continued to operate well for thepast year with uptime in excess of 99% of the sched-uled time. Approximately one-half of the 37.2 hours

of downtime was due to a failed insulator in the EinzelLens and a vacuum leak. Although ISIS personnel wereinvolved in other TRIUMF projects, high-priority ex-traordinary maintenance as well as a few importantprojects were undertaken during the past year.

General

The Prompt Radiation Safety Trip System has beenrevised to include redundant safety-critical gamma andneutron monitors. Each of these safety-critical tripfunctions requires three independent ISIS beam con-trol devices. To provide the additional beam-controldevice functionality we designed a system of four high-voltage relays to service the four vertical-bend elec-trodes as well as two beam stops. Each trip functionuses two high-voltage relays to switch two electrodesto ground potential and to insert a beam stop. Thesystem was installed, tested and successfully commis-sioned in February and March.

Approximately fifty new power supplies for the ver-tical section of the beam line have been ordered as partof our refurbishing program. These supplies are of thesame type as used in ISAC LEBT and will help reducethe overhead for spares. Last year we purchased two re-placement supplies for the inflector/deflector; this yearwe have ordered two more as well as a spare with re-versible polarity. It is anticipated that these supplieswill be installed in 2006.

The 24 V power for the ISIS interlock control sys-tem was traditionally supplied from the Controls groupthrough a UPS/converter system residing in ISIS. Thisunit is no longer serviceable and the Controls groupnow uses a site UPS feed for their system that doesnot have sufficient capacity to provide ISIS with thetraditional requirements. Therefore we have purchasedand installed our own 24 V power supplies for the in-terlock control system.

Ion Source

In order to provide the necessary ion source life-time for the revised beam schedule (maintenance onceevery 2 weeks) we have changed the ion source fila-ment diameter to 2 mm from 1.8 mm. This allows forcontinuous source operation between maintenance pe-riods. Unexpected leaks in the ion source led to thediscovery that an O-ring compression specification forthe filament rod was incorrect. This was temporarilyremedied, the drawing revised and new parts manufac-tured.

Injection Line

The injection line was leak-checked at the begin-ning of the winter shutdown. The most serious leak wasfound at the top of the vertical beam line. The leak is

215

small and the decision was made not to institute a re-pair at this time as this would involve removal of theVR1 section of beam line. Other small leaks were foundin the area immediately upstream of the vertical bendsection. This area was disassembled; the related opticswas removed and cleaned, reinstalled and aligned. Mi-nor leaks in the pepper pot and chamber 127 were alsorepaired.

Inflector/Deflector

The inflector/deflector and the high-voltage feedinsulators were serviced during the winter shutdowntogether with a complete replacement of the skimmercabling. The system ran well for the entire schedule al-beit with some sparking scan trips. Some developmenttime, as well as on-line subtle tune changes, greatlyreduced the number of sparks. Further investigationwill take place in the winter shutdown when the inflec-tor/deflector is removed from the cyclotron for servic-ing.

Development

We recently installed an adapter, gate valve, andstandard 8 in. ISIS box frame (borrowed from the I3beam line) at the end of the east-to-west section of theI1 beam line in preparation for the installation of anemittance scanner.

This year development was to focus on producing300 to 350 μA-equivalent tunes at higher duty cycleswith improved cyclotron beam losses. Because of otherdevelopment demands and limited scheduled time forsimultaneous extraction from three beam lines, onlyone shift was available for this work. However, someexcellent results were achieved as described in greaterdetail in the Beam Development section of this AnnualReport.

CRM

The restored first chamber of the 1 MeV cyclotronhas allowed the testing of cusp ion sources by ISIS, Nor-dion, and Dehnel Consulting. Dr. Yong-Seok Hwang,from the Department of Nuclear Engineering, SeoulNational University in South Korea, has completed anexcellent study of our ion source characteristics as re-ported at the International Ion Source Conference inCaen, France.

ISAC

ISIS has provided assistance to a number of ISACrelated activities during the past year. These includethe assembly of the diagnostic boxes and some diagnos-tics devices for the transfer line from ISAC-I to ISAC-II. We have continued to provide mechanical effort to-ward the completion of the ISAC-II medium-beta cry-omodules along with some of their diagnostics.

Standard optics chambers for the TITAN extensionof LEBT and the TITAN beam line have been pre-pared. We continue to assist in this work as necessary.

PRIMARY BEAM LINES

It had been assumed that the major task in thewinter shutdown would be the upgrade of the frontend of the M20 muon channel. However, as a result ofseveral problems that occurred on beam line 1A dur-ing the year it became necessary to address them firstand, consequently, to postpone the refurbishing of theM20 channel. To this end, shielding blocks from themeson area were moved to an area near the machineshop and other shielding blocks were rearranged to al-low shutdown access to the problem areas of the beamline.

Beam Line 1A

Winter shutdown

Magnet temperature trips had occurred atquadrupole 1AQ10 of the 1AQ10/11 doublet locatedbetween the 1AT1 and 1AT2 targets on beam line 1A.Further, both vacuum and water problems occurred in-termittently at this location. These were the first prob-lems tackled during the shutdown.

A video inspection showed corrosion on the 1AQ10base plate from water leakage near the lower thermalswitches and a broken spring eye on the south side ofthe 1AQ11 flange. It was hypothesized that because ofwater leakage in the area corrosion or rust were prob-ably causing a short in the thermal switch circuit. Be-cause repair of the magnet could not be done in-situthe doublet was moved to the remote handling warmcell.

With better access to the thermal switches, currentleakage was confirmed from a lower thermal switch tothe coil. Creative solutions were found for tooling toremove and reinstall and repair broken switches whilemaintaining the integrity of adjacent components. Thebroken flange spring on 1AQ11 was also replaced.

Significant effort and dose were required for thiswork; work in the warm cell alone required about 3.5weeks to complete. The dose accumulated for this andother beam line work is given in the Remote Handlingsection of this Annual Report.

Repair of quadrupole 1AQ9 was another prob-lem corrected during the shutdown. This quadrupole,which lies just downstream of the 1AT1 target, wasused primarily to steer π+ or π− mesons into the M11pion channel. Because of a thermal interlock problemin an inaccessible region it had been necessary to shutdown 1AQ9. As a result there was an increase in beamspill at a vacuum valve between the 1AT1 and 1AT2targets that caused radiation damage to the O-rings of

216

the valve. Work to correct this 1AQ9 problem was thenext job tackled.

When the bridge block covering 1AQ9 was removeda very fine coating of removable rust-like powder wasfound covering the quadrupole and surrounding ser-vices. Visual inspection of the underside of the bridgeblock indicated that the paint was disintegrating andsettling on the magnet – a spallation phenomenon thathad been observed in the M8 line during its decom-missioning. Careful vacuuming of the material in thetrench and magnet was performed allowing work toproceed in the area.

A faulty Hansen fitting was replaced to restore fullwater flow to the magnet. Repainting of the bridgeblock was accomplished using an epoxy paint.

Inspection of the thermal-switch lead junctions in-dicated significant corrosion that likely accounted forthe indicated fault. A solution was devised whereby apremade chain of thermal switches would clip onto thereturn water line of each circuit of the magnet. Thiswas done and confirmed by video inspection. Standardpower tests confirmed the magnet to be operational.

As reported last year, magnet overtemperatureproblems in the 1AQ14/15/16 quadrupole tripletdownstream of the 1AT2 target necessitated a retun-ing of the triplet to reduce the currents to the magnets.This resulted in the reduction of the extracted beamcurrent in order to lower beam spills downstream ofthe triplet. Flushing of this triplet was the final job ofthe shutdown.

The area was accessed for the removal of the waterinsulators to individually flush each cooling circuit ofthe 1AQ15. Flushing results were similar to those re-ported last year. As a bench mark, the 1AQ16 magnetwas exposed and one circuit was flushed. In comparisonto 1AQ15 this circuit was very clean with no change inflow rate before and after flushing.

The shielding above the 1AQ12/13 quadrupoledoublet was removed to inspect the area. No obvi-ous problems were observed. Video inspections of thecrumbling block on the upstream side of 1AT2 mon-ument showed no further disintegration of the block.Vacuum leak checking in the exposed area discoveredleaks only on monitor 1AM9. The beam line was ventedto replace 1AM9 and to remove valve 1AVA8 which wastransported to the remote handling warm cell wherethe transition flange O-rings were replaced with Helico-Flex metal seals. The O-rings on the valve gate andair cylinder were also replaced and the valve was rein-stalled in the beam line.

Beam production

BL1A performed well until late May when 1AQ14experienced magnet temperature trips. Again thebeam line had to be retuned to reduce the current in

the magnet. In late June the 1AQ15 water tempera-tures climbed to the point where the magnet currenthad to be reduced thereby reducing the beam current.Measurements on one circuit of 1AQ15 revealed a fac-tor of 3 reduction in flow rate. After flushing with waterand air 90% of the flow rate was recovered. A chelat-ing agent, which was purchased with the intent to flushthe coils, was used in an unscheduled intervention inearly July. It was mixed according the manufacturer’sspecifications and, when pumped through the circuits,it was found that the return filter was clogged with anundissolved abrasive used in the chemical agent. Chem-ical flushing was aborted and the circuit was rinsedwith water for a number of hours. Flow measurementsindicated a further 10% improvement. Similar resultswere obtained for a circuit of 1AQ14. A thermocou-ple was added to the 1AQ15 magnet for remote tem-perature monitoring. Temperature measurements on1AQ16 confirmed the magnet was operating normally.

After the July intervention the chelating agent wastested in our lab to ascertain an appropriate mixtureratio to water. The manufacturer states 6 ounces pergallon; our findings show that 3 grams per gallon aresufficient to clean copper and to remain in solution.

September mini-shutdown and beam production

In the September mini-shutdown the temperaturesof the coil circuits were measured for all three magnetsof the BL1A triplet. As expected, some circuits of the1AQ14 and 1AQ15 magnets were running at elevatedtemperatures. All three magnets were treated with thechelating agent for 4 hours with notable improvementsin water flow for 1AQ14 and 1AQ15. In addition, theexisting float flow meters were replaced with Proteuspaddle-wheel sensors.

After the mini-shutdown the beam line could berun at the nominal tune parameters.

Beam Line 2A

Beam intensity fluctuations at ISAC targets hadbeen a problem. A constant beam intensity is imper-ative in order that a constant temperature be main-tained at a target. Because fluctuation of the beamintensity in BL1A is not important at the level of afew per cent, a scheme has been implemented to keepfluctuations in the beam intensity in BL2A constantwithin a fraction of 1% at any BL2A current. Furtherdetails are given elsewhere in this Annual Report.

Another problem with beam delivery to the BL2Atargets was corrected this year. The original tune ofthe beam line provided a ±2 mm beam spot at thetarget. This proved to be too small and holes weredrilled through several targets. A revised tune was de-veloped to produce a beam spot ∼1 cm in diameter.

217

In the course of this development an error in the pro-gram that calculated the vertical portion of the ex-traction matrix through the cyclotron extraction fieldwas found and corrected. (See the Beam Developmentsection for further details.)

BL2A was surveyed and O-rings were replaced inhigh radiation field locations during the shutdown.During the year five radiation damaged O-rings werereplaced in scheduled maintenance days. The metalseals installed this shutdown at 1AVA8 will providea test case and, if successful, the BL2A O-rings will bereplaced with metal seals starting in 2007.

Beam Line 2C

2005 was a record year for the production ofstrontium-82 from the 2C4 solid target facility (STF),which was scheduled for 146 days. On 2C1, 35 dayswere scheduled for proton therapy (PT) and 36 daysfor experiments at the proton irradiation facility (PIF).

On 2C4 55.0 Ci of 82Sr were produced from156.9 mAh of beam compared with 48.5 Ci from134.3 mAh in 2004. Eleven natural rubidium targetswere irradiated compared with nine targets in 2004.The make rate of 0.35 mCi/μA-h remained constantwith beam currents of 55–60 μA and irradiations ofapproximately 20 days. The beam intensity could bemaintained at 55–60 μA because of the improved man-ufacturing of extraction strippers and of the larger ac-ceptance of the 2C4 beam line resulting from the re-alignment of the STF (see last year’s Annual Report).One of the improved, 0.200 in. wide graphite-brushstrippers was used for 75 mA/h. In previous years theaccumulated charge on an extraction stripper was lim-ited to ∼50 mA/h and strippers were replaced as a pre-cautionary measure during short cyclotron openings.

No work was done on the STF target assembly dur-ing the winter shutdown because of the anticipated re-placement with an upgraded STF and to limit person-nel dose. On the STF water package the ball valveswere replaced with globe valves and a metering valvewas installed on the water line for the ion exchangecolumn to achieve better regulation of water flow andof the water column height. Also, the 2C4 split platemonitor was replaced with a harp monitor and new ca-bling and electronics were installed. This monitor nowprovides more accurate beam position data at the en-trance of 2C4 downstream of the switching magnet.

By the fall the design of an upgraded STF was com-pleted. All major mechanical components for the targetwater vessel and the target insertion were in hand. Amock-up and assembly of the STF had been set up inthe proton hall and on the proton hall roof to check re-liability and installation procedures. The water-cooleddouble window at the end of the 2C4 beam pipe had

to be redesigned because the window proved to be dif-ficult to manufacture and did not meet water pressurespecifications. The 2C4 beam pipe and diagnostics vac-uum chamber were also redesigned to accommodate ascanning wire monitor and a new target protect (halo)monitor. Components were being manufactured for in-stallation in 2006. By year-end, designs for the waterpackage and the PLC based control system had beencompleted and most parts had been ordered.

On 2C1, nine patients were treated during five pro-ton therapy sessions. PIF operation was partially com-promised because energies above 108 MeV could not beextracted, although the 2C extraction system had pre-viously been working from 65 MeV to 120 MeV. Dur-ing a winter shutdown assessment no blockages couldbe observed in the 2C exit horn. Later in the fall itwas discovered that the 2C combination magnet po-larity switch was shorted for magnet settings above96 MeV and that the exit horn intercepted beam whenthe extraction probe was tuned above 108 MeV. Oncediscovered, the fault was easily repaired.

CONTROLS

2005 was an active year for the Controls group.There were new developments and a large componentof maintenance. Perhaps the most significant of thenew developments was the work on beam stability forBeam Line 2A, which resulted in an important im-provement to the beam current stability. After beinggiven cyclotron development time to test and refine thesoftware, closed loop feedback on the ISIS pulser wascommissioned. This software now offers Operations anew tool for providing stable proton beam on 2A withless downtime for radioactive ion beam production.

The Central Control System (CCS) ran well dur-ing 2005. The loss of scheduled beam time due to CCSfaults during the year, as recorded by the Operationsgroup, was 3.4 hours. This low level of downtime wasprovided without compromising important hardwareand software advances.

A number of tasks had been scheduled for 2005and those projects progressed as anticipated. These in-cluded decommissioning the DSSI disk system in theProduction Cluster, continuing to replace old termi-nal servers, changing cyclotron device numbering fromoctal to decimal, and introducing Itanium based com-puters into the OpenVMS clusters. Some of these jobsare described in further detail below.

Central Control System Facilities

Developing software to provide beam-current sta-bility on beam line 2A was a major beam achievment.For maximum RIB target production, it is desirableto run the beam current very close to the over-current

218

trip set-point. Minor drifts or noise in the beam cur-rent may lead to the beam being shut off and the as-sociated downtime. An application has been developedand commissioned to monitor the extracted beam cur-rent as measured by the extraction foil and use theISIS pulser to quickly stabilize the current on beam line2A where stability is critical (at the expense of someinstability in the simultaneously extracted beam line1A beam, which is quite acceptable). Development andtesting time were difficult to obtain but once granted,the software was refined and put into production. Dis-play changes and reassignments on the operator’s con-sole were also made. Beam production with the stabil-ity program enabled is now a normal operating condi-tion.

Among the larger tasks was the investigation ofa new line of HP computers called Integrity Servers.These were explored for their potential use in runningOpenVMS in the CCS. They use Intel 64-bit Itaniumchips instead of the Alpha chips which are more com-monly associated with OpenVMS. In 2004 an IntegrityServer RX 2600 was borrowed from HP and set up fortesting. The unit was later purchased and integratedinto the CCS, although not initially at a productionlevel. A significant amount of the software infrastruc-ture from the manufacturers was originally at the beta-test or field-test level (OpenVMS FT 8.2, Multinet 5.1,and DMQ T5.01) but during 2005 these have been re-placed by production releases.

Within the TRIUMF-written control-system soft-ware there have been a number of developments. Alow-level software driver was adapted to the new hard-ware platform and then enhanced to support multi-processor computers. This software driver allows accessto cyclotron devices. All of the device tables and sup-port routines have also been ported. By the end of 2005almost all CCS software applications had been ported.Some porting issues arose but, in general, applicationsjust need to be recompiled and relinked. During thelast part of this year’s beam delivery the displays onthe cyclotron’s main console were running in produc-tion mode on one of these Itanium based machines.

The mini-shutdown in September was a busy time.One of the most important events was the final de-commissioning of the old DSSI disk system from theProduction Cluster. This disk system had run well formany years but had been replaced by a fibre chan-nel raid system (a storage area network). The mini-shutdown provided the first opportunity with sufficienttime to remove this disk hardware. The phasing outof the DSSI disk storage equipment on the Produc-tion Cluster was an ongoing job for more than 1 year(the DSSI system on the Development Cluster was al-ready decommissioned). The labour intensive part of

this task was moving the files to the newer fibre chan-nel disk system, reorganizing the files, and removingold and obsolete files. This large reorganizing job wasfinally completed during the second quarter. The diskswere dismounted in software but will not be physicallydisconnected and removed until the next shutdown in2006. During 2006 the fibre channel disk system willbe expanded to meet the increasing requirements fordisk space.

A major effort went into supporting the new ISISsafety trip devices that were introduced for the up-graded prompt radiation hazard system scheme. Thisproject involves high-voltage relays at the verticalbends in ISIS together with a variety of other equip-ment. Numerous hardware and software changes wererequired in the CCS. New devices were defined, dis-plays were developed, scans configured, and messagesset up. Because these changes involve personnel safetyand are essential for cyclotron operation, the work wasof high priority.

Support was provided to the RF group in a num-ber of areas. One example was a development involv-ing their vector voltmeters, a multiplexer, and a num-ber of related signals. Device definitions, logging, anddisplay pages were developed. This was quite an in-volved task. Another example was the creation of ascan to check that the rf booster is detuned when notin use. In still another development, the two existing XWindow terminals in the rf console were replaced by amulti-headed display computer, which provides betterperformance. An RFinfo display panel was developedwhich makes it easy to insert information from the dis-play page into the RF group’s electronic log (elog).This Web-based elog facility was ported from the ver-sion being used by ISAC and was made available tocyclotron groups as their needs arose. The RF groupwas the first to start using this facility as supported bythe Controls group.

There were serious problems with the site UPS andthe decision was made to move CCS equipment off thatUPS. This later led to the problem of moving all of thatequipment back to UPS. Some disruption was experi-enced but these procedures went relatively smoothly.A new site UPS will be installed during the 2006 shut-down and it is hoped that this will remove these powersystem issues.

During 2005 one of the development cluster com-puters shut itself down a number of times. The prob-lem was traced to a faulty network interface. Replacingthe motherboard rectified the problem, but during theinteraction with HP support services a new networkconfiguration was suggested. This new configurationwas set up, tested, and determined to be a significantimprovement. It uses two network ports on each com-

219

puter (the Alphas and Itaniums each have 2 connec-tions) with each connection going to a separate Eth-ernet switch and, in fact, to separate networks. Thereare multiple benefits. Without any software reconfigu-ration, OpenVMS clustering immediately senses mul-tiple networks and uses them to provide increased per-formance and redundancy. The original hardware fail-ure now will not cause a system crash. In addition, ourdisplay computers are also configured in this manner,something that is not possible for the aging X Win-dow terminals. The Operator display pages run usingX Window displays over the DECNET protocol andDECNET will dynamically switch over to the othernetwork if a network failure occurs. As a result, theoverall performance improved, the computer/cluster/XWindow display reliability increased, and the networkbandwidth doubled.

More display computers were installed during thisquarter. The aging VXT X Window terminals areslowly being phased out. A new display computer wasinstalled in the rf console and another in the main con-sole of the cyclotron. Additional display computers willbe installed during the next year.

There were a number of CCS hardware-related ac-tivities during the year of which a sample are reportedhere. For hardware diagnostics another computer waspurchased and configured in the development cluster.This provides more capacity to do on-line diagnosticsand testing. In an effort to replace some of the olderCAMAC crates, six crates were updated. In the on-going task of the installation of new crate-diagnosticequipment, more power monitors were added. Thecontrol-system-specific UPS hardware saw the 7.5 kWunit upgraded to 10 kW. A number of 24 V systemswere standardized and received new 24 V supplies andcabling. A variety of hardware faults occurred althoughnone caused major downtime. There were problemswith input gates, dicons, fans, power supplies, powermonitors, X Window terminals, and the like.

The year ended with a flurry of activities relatedto projects planned for the upcoming shutdown. Thesewere dominated by those related to the solid targetfacility upgrade.

Secondary Beam Lines

Increased use of secondary beam lines controls andthe limitations of the existing microprocessor have ledto the need to expand the computing power. After tak-ing measurements and considering the options, it hasbeen decided to break the existing CAMAC serial high-way into two highways and have a microprocessor driveeach branch. These changes may be done in the 2006shutdown if time and parts delivery allow.

Diagnostics for the secondary beam line system

have been enhanced through the development of anon-line crate-monitoring program.

Security on the secondary beam line servers hasbeen tightened.

Other Systems

A request from Nordion for beam-related data wasreceived. An application was developed that returnsdata to the caller via the Web server.

Support continued for the Vacuum group’s teststand. Devices were configured in the device databaseand added to the logging stream. In addition, a newXTPAGE was developed to display the data.

Proton therapy received some attention. An agingterminal server was replaced with a new unit. This in-cluded changing the software support from the LATprotocol to TCPIP.

The proton irradiation facility had numerous usersduring 2005 and several requests for new functionalitywere handled.

A request from the PET facility to have real-timeaccess to the data from air monitor 27 was met byallowing the trending data to be displayed locally.

OPERATIONAL SERVICES

Remote Handling

Cyclotron servicing

The Remote Handling winter shutdown activitiesfor the cyclotron included the usual interim removaland storage of copper blockers, tank peripheral person-nel shadow shield installation, remote vacuum clean-ing, as well as video and tank seal inspections. In addi-tion the 2C extraction probe and the LE2 low-energydiagnostics probe were both removed for servicing andreinstalled.

A new photographic documentation of the cy-clotron tank was begun with quadrants 1, 2 and 4completed and quadrant 3 nearly completed. This willprovide valuable information on critical in-tank details.

Toward the end of the shutdown a leak was locatedin the Upper Quadrant 3 resonator cooling system.Remote leak checking with helium was performed tonarrow this down to the upper #6 resonator segmentin this quadrant. The segment was remotely removedfrom the tank using the resonator handling trolley andmoved to the Remote Handling building basement. Theleak was pinpointed in the ground panel and the panelreplaced.

Routine inspection and maintenance were also per-formed on the cyclotron elevating system. The elevat-ing jacks and worm gear reducers at station 8 werereplaced with original units. Station 7 was removedand replaced with spare units. These components willbe rebuilt for installation during the next shutdown.

220

As part of the overall refurbishing program, sourcingbegan for new counters to replace the original 35 year-old plastic counters that are rapidly deteriorating dueto radiation damage.

Hot cells and targets

In the winter shutdown routine maintenance ofthe 1AT1 and 1AT2 cooling packages was performed.Both packages were drained and refilled and the ion-exchange resin was replaced, as was the target, collima-tor and, at 1AT1, the M15 permanent-magnet coolingcircuit filters. Also at 1AT1, the rubber pneumatic airline hoses supplying the target profile monitor were re-plumbed with radiation-hard, 1/4 in. diameter coppertubing. At 1AT2, similar rubber hoses supplying thetarget profile monitor as well as M9 and M20 beamblockers were replumbed with copper tube. The P1Aand P2A transducers (target cooling-water pressure,supply and return) were moved from the cooling pack-age into the beam line 1A access tunnel.

The old M8 medical beam line blocker solenoidvalve and air line were removed. The M9 beam-blocker vacuum seal, centre shaft and radiation dam-aged micro-switch wiring were replaced in the hot cell.

At mid-year the demineralizer water-flow trans-ducer for the 1AT2 cooling package was replaced be-cause of rotor and axle wear. The 1AT1-Mark2 targetwas transferred to the hot cell in preparation for thescheduled installation of graphite targets. The spareM20Q1 cooling-water jumpers were transferred fromstorage in the east hot cell to the warm cell wherethey were radiation surveyed to permit safe manualreplacement of the sealing O-rings. In the beam line,the M20 beam-blocker height-position micro-switch ac-tuator was realigned.

Developments on the graphite target included a jigredesign and new parts for more accurate placement ofgraphite material onto the cooled Saddle, new brazedand machined graphite segments and the first segmentsoft-soldered onto a saddle.

At the 2C hot cell, a total of nine rubidium produc-tion targets were installed and removed. One target,which had a cracked weld and elevated radiation lev-els, prompted removal of an unused target from the cellfor comparison to determine a cause for weld failure.Also, the 2C cooling package was routinely inspected,serviced and adjusted as required. The demineralizerwas exchanged to increase flow for optimized resistiv-ity, and the conductivity cell was replaced. ContinuedRemote Handling input was given to the solid targetfacility redesign project.

Beam lines

Winter shutdown began with remote removal ofthe 1AQ10/11 quadrupole doublet from beam line 1A

and its transport to the hot cell facility. The person-nel access shielding wall in the warm cell was utilizedfor hands-on access to replace faulty Klixon thermal-switches and to repair wiring on 1AQ10. A cracked ce-ramic circuit insulator on 1AQ10 was replaced, and acorroded and broken compression spring on the indiumvacuum joint of 1AQ11 was also manually replaced.Total personnel exposure for the 1AQ10/11 work wasrecorded at 8.02 mSv.

At the end of January the 1AQ9 magnet, whichlies downstream of T1 target, was uncovered. Anopen thermal-switch circuit preventing operation of themagnet was repaired by installing an ingenious “clip-on” style redundant thermal-switch wiring to bypassthe open circuit. A water leak required replacementof the quick-disconnect coupling fittings. Opportunitywas taken to repaint the iron shield block situated overthe magnet. Personnel exposure for the 1AQ9 repairwas 3.07 mSv.

General inspection and vacuum leak checking wereperformed in the 1AQ12/13 area and in the 1AVAW2-to-1AT2 beam line section. Video inspection was com-pleted on a crumbling concrete block in this area inanticipation of its future replacement. This inspectionwork recorded 0.51 mSv in total.

The 1AVA8 vacuum gate valve was removed andtransported to the warm cell where it was rebuilt withradiation hard metal seals replacing rubber O-rings. Abroken vacuum flange compression spring on the as-sembly was also replaced. Assistance was given withthe replacement of damaged rubber O-rings in thebeam line. The total exposure for 1AVA8 work was3.01 mSv.

To complete the shutdown, the 1AQ14/15/16triplet at 1AT2 was uncovered, a solder repair made toa joint on 1AQ15, and a slow drip water leak at anothercircuit examined. Finally all flow rates of its coolingcircuits were individually measured, the circuits pres-sure flushed and rates remeasured. The 1AQ16 magnetalso had one circuit flushed and measured. Personnelexposure for the work was 3.72 mSv.

In May, after the shutdown, the magnets of theBL1A triplet underwent a second series of back-and pressure-flushing to improve deteriorating coolingflows. The 1AQ15 quadrupole, because of its higher op-erating current, was compromised by over-temperaturetripping, limiting beam line 1A operating currents.Further temperature “warning” trips from 1AQ14 in-stigated a universal pressure back-flush of both mag-nets again in June. This was completed by accessto TNF south without the removal of any shieldingblocks.

During early July elevated temperature trips onceagain required independent coil flushing of 1AQ14 cir-

221

cuit 1 and 1AQ15 circuits 5 and 11. Beam line shieldingwas removed for local access and chemical (PhoenixCu) and air/water flushes were performed on the in-dividual circuits. A thermocouple was mounted on the1AQ14 circuit 1 to monitor future temperature rise ver-sus thermal switch trips. In late July a cooling waterconductivity meter was installed on the 1AQ15 cool-ing return line to learn more detail regarding the con-stantly deteriorating cooling flow blockage.

In mid August the 1AQ14 and 1AQ15 quadrupoleswere once again universally back-flushed from the TNFsouth area on two maintenance days that occurred inconsecutive weeks.

Finally, during the September mini-shutdown thetriplet was universally flushed with the chemical so-lution at elevated pressures and temperatures. Signifi-cant improvement was noted in the individual magnetflow rates at this time. During this same shutdown awater leak in the M20Q1 magnet required shieldingblock removal for access. A leaking rubber seal ringon a quick-disconnect water coupling was observed.The entire cooling line was replaced with a refurbishedspare.

Other significant work was the design of the beamline vacuum box and services for the replacementM20Q1/Q2 magnet assembly scheduled for installationin 2006. An investigation also began on possible revi-sions to the 1AQ14/15/16 triplet magnet cooling ser-vices. Alternatives to the bellows/insulators that limitpressure boost of the cooling system are being evalu-ated.

Magnet Power Supplies

The year started with shutdown activities, one ofwhich was the inspection of power supplies and em-ploying corrective action as required. The brushes ofthe variable-voltage transformers were replaced. In therectifier cabinet of the main magnet supply the torqueof the heat sink of the rectifier clamp was verified andhoses were replaced.

A major project was the modification of the trimand harmonic power supplies for bay 1. These wererewired to bring them up to the required standards ofthe electrical code. Work proceeded to ready the powersupplies for the rewiring of the trim and harmonic sup-plies in rack 3. This will be done in the winter 2006shutdown.

Assistance was provided to resolve the over-temperature problems that had occurred with thebeam line 1A quadrupole triplet.

In ISAC, power supplies were installed and com-missioned for the components of the beam line fromISAC-I to ISAC-II. The power supplies for the mag-netic elements of the medium-beta accelerator (SCB)

were procured as were those for the beam line in theaccelerator vault of ISAC-II (SEBT).

A new polarity switch controller was designed andimplemented for the SCB and SEBT.

The group performed normal MRO activitiesthroughout the year as well as providing experimen-tal support when required.

The group leader was involved in the coordinationof the winter and fall shutdown activities. In addition,he was involved in the coordination of the activities ofthe various service groups involved in the installationof the components of the vault section of ISAC-II.

Electrical Services

It was a relatively quiet year on the site mainte-nance front. The only major exception was the failureof the main UPS unit. However, on the constructionfront we were extremely busy building the ISAC-II lab-oratory infrastructures. The ISAC work is covered ina separate section.

In all, about 43 entries of various importance werelogged in the engineering logbook. Of these, 28 be-longed to ISAC. Major site-related jobs completed in-clude replacement of an existing 5 kVA UPS with a10 kVA unit in the control room, refurbishment of theac distribution for the trim harmonics bay 3, the TR13radiation exhaust, and UPS to the rf room to servicecontrols critical to the reliability of the rf system.

Continuing engineering support was provided toTRIUMF users, Nordion and the ATG group. Majorchanges were introduced by the government regardingsafety regulations following the passage of a new SafetyStandards Act last year. One of the most significant ofthese that affected TRIUMF was the change from anannual permit to an annual operating permit. Underthe new regime our electrical permit no longer coverselectrical installations and equipment approval. Equip-ment approval is mandated to external agencies or toa special certification program that is administered bythe newly created BC Safety Authority. This programhas started but their policies are still evolving.

Design of the refurbishment of the rf ac distribu-tion was started and good progress has been made.Existing drawings were updated to reflect reality andthe single-line diagram was updated with future addi-tions and rationalized. The proposed design includes athree-level breaker distribution to allow maintenanceintervention on a single group of devices without shut-ting down the whole unit. As this is a large project, itwas decided that it was more efficient to prepare thespecifications and contract the engineering work to in-dustry. The new distribution is planned for installationin the winter 2007 shutdown.

Typical maintenance activities included servicing

222

lighting systems, motors and associated controls, airconditioning controls, panel boards and transformers,high-voltage switchgear, breakers and capacitor banks,and the fire alarm system. The annual inspection ofthe fire alarm system was carried out during the shut-down. Some problems with the ISAC-II fire alarm sys-tem that were related to electro-magnetic interferencewere eventually solved by inserting isolator modules onthe various addressable lines.

After many problems the main UPS unit eventu-ally died. Following an extensive search for the bestoptions a decision was made late in the year to installthe replacement UPS in early January, 2006. The newunit design is an N+1, “hot-swappable” model withsignificantly better reliability and better diagnosticsand monitoring functions. Four 10 kVA modules arecontinuously on-line to supply our 40 kVA load andone 10 kVA unit is in “hot” stand-by mode, ready totake the load should one of the operating modules fail.The other significant advantage is that the failed mod-ule can be replaced without shutting down the UPS.

Power delivery

Power delivery continued to be reliable and withoutmajor unscheduled outages.

Coordination with Powertech Labs Inc. continuedon a regular basis. It is important to acknowledge theexcellent cooperation of the staff of Powertech Labswho worked closely with our operation staff to mini-mize the impact of the activities of the high voltagelaboratory on our experimental program. In February,UBC replaced the metering units of the South Campussubstation.

For the first time in five years the monthly av-eraged, peak-power demand took a turn downward,dropping by 1.8% from 7,560 kVA to 7,426 kVA(Fig. 263). The annual maximum peak demand,reached in December (8,957 kVA), was a mere 0.5%higher than the previous year. The electricity com-sumption also bucked the previous trend, decreasinga good 5% from 58.4 GWh to 55.5 GWh (Fig. 264).The largest monthly consumption was recorded in Au-gust (6.0 GWh) and was approximately 3.4% less thanlast year’s maximum. These declines are indicative oflonger maintenance activities on site.

The power factor (PF) averaged over the calendaryear improved a mere 0.3% to 96.7% (Fig. 265). Theaverage load duration – the indicator of how well weuse our power demand – dropped a significant 6.2% to82.7% (Fig. 266). The drop is primarily the result ofthe February and March start-up activities.

It may be of interest to compare this year’s con-sumption figures to those of the year 2000. Thepower demand has increased 10.2% while the electric-ity consumption has been more contained with a 4.4%

Monthly Peak Power Demand

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

JAN

FE

B

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SE

P

OC

T

NO

V

DE

C

MO

AV

G

kVA

2005 2004

2003 2002

MONTH YEAR AVERAGE UNITS 2002 6834 kVA 2003 7358 KVA 2004 7560 KVA 2005 7426 KVA

Fig. 263. Monthly peak power demand – four year com-parison.

Electricity Consumption

0

1000

2000

3000

4000

5000

6000

7000

JAN

FE

B

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SE

P

OC

T

NO

V

DE

C

MO

AV

G

(M

Wh

)

2005 2004

2003 2002

MONTH YEAR TOTAL AVERAGE UNITS 2002 53.3 4.440 GWh 2003 56.4 4.702 GWh 2004 58.4 4.866 GWh 2005 55.5 4.627 GWh

Fig. 264. Electricity consumption – four year comparison.

Power Factor

94%

95%

96%

97%

98%

99%

100%

JAN

FE

B

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SE

P

OC

T

NO

V

DE

C

YR

AV

G

PF

(%

)

2005 2004 2003 2002

Fig. 265. Electrical system power factor – four year com-parison.

223

Load Factor

55%

60%

65%

70%

75%

80%

85%

90%

95%

100%

105%

JAN

FE

B

MA

R

AP

R

MA

Y

JUN

JUL

AU

G

SE

P

OC

T

NO

V

DE

C

YR

AV

G

LD

(%

)

2005 2004 2003 2002

Fig. 266. Electrical system load duration – four year com-parison.

increase. Despite the difference in demand versus con-sumption increase, the power factor edged up slightlyby 0.7%. This improvement can be attributed to theaddition of the capacitor banks in ISAC.

Mechanical Systems

The largest project during the last year was thedemolition of the existing BL1A exhaust HEPA filterhousing and the installation of a new filter rack sys-tem. Impressive in size, it is out of sight to most peoplebecause it is down the stairs of the south TNF. Theold polyflo instrument tubing was replaced with cop-per tubing. The second largest job was the new nuclearventilation system of the TR13 hot cell.

A new heat exchanger for the rf dummy load wasinstalled and new check valves were installed on anamplifier cooling circuit. Deficiencies observed in theannual fire sprinkler inspection were repaired. Othersprinkler work was the installation of a new pressureswitch and the repair of a leak in the chemistry an-nex. The failure of a fire sprinkler in the main officebuilding prompted a random inspection of other heads(although they have a nominal minimum lifespan of 50years).

After the winter shutdown assistance was givenin the flushing of the resonators. Vibration in the

CuALCW pump led to cracks in a discharge pipingweld. This problem was immediately addressed by in-stalling a vibration-isolation spool and, later, by apump overhaul. A portable deionized-water resistivitymeter was provided to the Beam Lines group to aid intheir investigation of the cooling problems of the BL1Aquadrupole triplet. A dissolved oxygen sensor was alsoprovided so that additional data could be obtained.

Other activities included the installation of a newcooling fan in the Machine Shop boiler room, the startof the extension of the CuLCW system into the mainoffice building from the new underground connection,and the fabrication and installation of new manifoldsfor cooling water for the M9 meson channel and for theM9 vacuum system. New air conditioning was providedfor the Thaumaturgy room and the AHU-1 of the mainoffice building was overhauled. A new exhaust systemfor the kitchen fume hood was also installed. The sinkand hot water tank in the MHESA basement darkroomwere relocated.

A design review was held for a new M20 VA1 valvelifting jig. Engineering assistance was provided for anew fume hood and related speed-up of fan AH5 forthe CP42 vault and to the ISAC-II warm helium pipingjob. Cost estimates were provided for the air condition-ing of the top floor of MHESA and for a new routerroom. The latter was eventually relocated to the pro-ton hall and the air conditioning job was begun withcompletion expected in the new year.

In ISAC construction of water, air and vacuumpiping continued for the transfer line from ISAC-I toISAC-II and in the accelerator vault. A new, dry vac-uum pump exhaust line was run from ISAC-II to thenuclear ventilation system of ISAC-I. The air condi-tioning programming of room 242 was modified toavoid a deadband situation. A concept design and costestimate was provided for air conditioning the pro-posed ATLAS computer centre. An upgraded transferpump for deionized water was installed. The oxygendeficiency system was put in service and underwentsome development.

Maintenance was provided for building services: airconditioning, piping, and the new DDC. A new aircompressor was installed on the ISAC-II dry (duo-action) sprinkler system, and a compressed air line inroom 164 was relocated to allow for cryogenic piping.

224