Down time Annual report 2002

Down Time Linear Accelerators.
University Medical Centre
Utrecht, the Netherlands
Annual Report 2002

J.G.M. Kok & associates

Dept. of Radiotherapy, University Medical Centre

Heidelberglaan 100, 3584 CX Utrecht, The Netherlands

j.g.m.kok@umcutrecht.nl

J. G .M. Kok

kok@radth.ruu.nl

March 5, 2010

Introduction

This is the fourteenth (2002) annual report on the down time of the UMC linear accelerators. Down time of the accelerators is registered to get an impression of the performance of the machines. It enables comparison between different machines and also keeps record of performance over the clinical life of the linac. The produced figures are less suitable for comparison between different radiotherapy departments. The major reason for this is: 1 Several definitions of down time are used and it is not always clear which one. 2 Reliable down time figures depend on accurate administration. The accuracy of this administration is not always warranted.

Definition:
In general there are two definitions of down time. From the patients point of view and for machine performance. For the patient it is important to be treated every day without long delays. Let us say if one linac breaks down but the patient can be treated on an other linac (which for example was just undergoing service). In this case treatment is not inhibited and from this point of view there is no down time. On the other hand it is useful to have an impression of the performance of the linac itself. How many times does it break down? How long does it take to repair? Does one particular part of the linac give trouble? (for example MLC) etc.

A point of discussion is if maintenance must been seen as down time. From the patients point of view the linac is not available so it would be down time. On the other hand when a linac is purchased it is known that maintenance must be carried out at regular intervals to ensure safe and trouble free operation.

Administration:
We make extensive use of the debug terminal (The debug program written by Niels de Graaff is free available on our Internet page: http://www.radiotherapie.nl/linaceng/index.html.) All errors are registered automatically and at the end of the month the debug files are edited and the number of interrupts are counted. This is multiplied by a standard time factor. When we replace let us say a light bulb we use a text editor available in the debug program to administrate the time we used.

The down time definition we use is:
Down time is the time when a linac can not be used for patient treatment due to electrical, mechanical or physical faults during normal working hours. (This implies that maintenance during working hours and repairs in the evening or at weekends are not registered as down time).

Down Time of our six linear accelerators

	U1	U2	U3	U4	U5	U6	Aver.
1989	1.7 %	5.7 %			7.8 %		5.1 %
1990	3.6 %	2.8 %		4.8 %	2.3 %		3.4 %
1991	0.6 %	2.3 %	1.0 %	1.5 %	3.2 %		1.7 %
1992	1.1 %	1.6 %	1.4 %	3.8 %	1.5 %		1.9 %
1993	0.8 %	4.3 %	1.4 %	0.7 %	6.5 %		2.7 %
1994	1.6 %	0.6 %	0.7 %	2.1 %	0.6 %		1.1 %
1995	1.6 %	3.1 %	0.4 %	0.6 %	2.1 %	0.6 %	1.4 %
1996	1.2 %	1.9 %	0.8 %	2.9 %	0.7 %	0.5 %	1.3 %
1997	2.0 %	5.4 %		1.7 %	1.6 %	4.6 %	3.1 %
1998	1.4 %	1.7 %		1.8 %	1.4 %	1.0 %	1.5 %
1999	1.4 %	0.9 %	5.7 %	6.5 %	0.6 %	2.1 %	2.7 %
2000	3.1 %	1.1 %	1.4 %	3.4 %	2.0 %	1.2 %	3.0 %
2001	2.9 %	4.9 %	5.5 %	3.2 %	2.0 %	0.9 %	3.2 %
2002	0.6 %	1.8 %	1.5 %	3.4 %	1.2 %	1 %	1.6 %

The average down time over 14 years is 2.4% $\sigma$ 1.1

Equipment:
U1: mc5037 1987 SL25-15 from 2002 mc5815 SL$i$ + MLC + $i$View GT.
U2: mc5047 1987 SL25-15 from 2002 mc5766 SL$i$ + MLC + $i$View GT.
U3: until 1996 SL75-20 from 1999 mc5504 SL$i$ + MLC + $i$View GT.
U4: mc5087 1991 SL25-20 + MLC and Iview.
U5: mc5074 1990 SL25-15.
U6: mc5289 1995 SL25-15 + Iview.

Down time per month in 2002

2002	U1	U2	U3	U4	U5	U6
January	2.4 %		1.4 %	2.2 %	5.4 %	0.3 %
February	2.0 %		1.4 %	3.9 %	0.8 %	0.4 %
March	1.0 %	(15.2)	0.9 %	0.4 %	0 %	0.2 %
April		2.0 %	3.2 %	1.3 %	0.8 %	0.2 %
May		4.6 %	0 %	2.3 %	1.5 %	1.1 %
June		2.1 %	1.6 %	1.3 %	0 %	0.1 %
July		2.7 %	1.6 %	1.6 %	0.9 %	6.5 %
August	(22)	1.1 %	1.2 %	5.5 %	0.6 %	1.4 %
September	0.3 %	1.5 %	1.0 %	1.2 %	0.2 %	0.3 %
October	0.7 %	0.4 %	1.1 %	7.5 %	2.7 %	0.4 %
November	1 %	1.3 %	2.2 %	0.8 %	0.8 %	0.3 %
December	0.5 %	0.9 %	1.9 %	11.1%	0.2 %	0.4 %
Average	0.6 %	1.8 %	1.5 %	3.4 %	1.2 %	1.0 %

In 2002 the U1 and U2 were replaced by new Elekta SL$i$ linacs. During there first month of clinical life they suffered quite a lot from growing pains. Because the down time figures of the first month are not representative for the performance of the linac over the year they are not used to calculate the annual average.

major events in 2002

U1

mc5037 1987 uptil March 2002.
From 19 August 2002 mc5815.

January: Power supplies HTCA.
March: Decomissioning MC5037

August: 19 start clinical use MC5815
August: Contactor H LT (Diode
placed over Gantry rotation brake.)
October: Plateau rotation clutch.
November: Water temp servo.
December: HHC 1 cable.

U2

mc5766 2002.

February: Start clinical use.
February: IMRT 2R 2T errors.
March: Lots of ``Contactor H LT'' problems due to ROC card in ICCA.
April: Table height, end switch low.
April: Thyratron fuse.
May: Iview GT
June: Y1 coarse potmeter.
June: Table height gear box and
potmeter replaced.
July: Table height Z drive replaced.
July: Application crashed twice.
August: Height control, Table
control unit.
November: Iview GT glass fibre.
December: Warming up problems.

U3

mc5504 1999

January: Drive belt diaphragm.
Focus 2 fuse.
February: IMRT 2R 2T errors.
April: table top lat. brake
June: Broken MLC video cable.
June: Fire connector realoff post.
July: MLC radiation damage to
the reflectors is reducing reflection.
July: Fuse 26 Volt motor drive
August: X2 drive belt.
August: Iview GT.
August: Thyratron fuse.
September: Sec. Filter Control.
Oct.- Dec.: IMRT MLC problem 7931.

U4

mc5087 1991.

January: Frequent MLC camera problems.
February: Table electronics contaminated with urine.
MLC PCB's replaced.
April: MLC leaf reflectors replaced.
May: MLC various problems.
May: sec. Filter mon.
May: Wedge fuse.
July: Thyratron replaced.
July: Fuse wedge.
August: Ionisation chamber replaced.
October: Gun replaced.
December: Vacuum leak bellow.

U5

mc5074 1990

January: Vacuum problems target slit.
January: Table height clutch replaced.
April: X2 fine potmeter replaced.
May: Microswitch Coll lockt, replaced.
July: Table height clutch replaced.
July: Thyratron replaced.
August: Table height clutch replaced.
October: Vacuum problems target slit.
October: Vacuum pump replaced.

U6

mc5289 1995

May: Bicca strip replaced.
July: Magnetron replaced.
July: Water cooler, compressor
failed.
August: Light field alignment.

conclusion

2002 has been a interesting year from the linac engineering point of view. The installation of two new accelerators equipped with new techniques like the slit-less flight tube and the fast track magnetron were accompanied with the introduction of IMRT treatments. The new machines are performing good but nevertheless the $i$View GT systems caused a lot of head ache. ( At the end bad opto fibres were to blame.)
For the fourth year in a row the U4 is performing less good than the other machines. The problems on this machine were not caused by one particular item. The MLC however consumed a lot of engineering attention. This was also the case for mlc of the U3. It is likely that the MLC problems are caused by the use of the 18MV photon beam, introducing deteriation of the reflectors and damage to the camera.
In general the down time in 2002 was at an acceptable level.