14 October 1998

National Physical Laboratory

 

Attendees:

John Makepeace, NPL (Chairman)

Simon Woods, NPL (Secretary)

Ian Adsley, AEAT

David Anstee, Canberra

John Barton, Birkbeck College

Roger Benzing, Imperial College

Peter Clark, BNFL Instruments

Chris Dale, NNC

Stephen Day, Stanger Science and Environment

Sarah Dean, IRAS

Tony Day, VLA

Jim Desmond, BNFL

Julie Eccles, AEAT

George Elder, BRADTEC

Jim Fitzgerald, J F Computing

Mike Froggatt, BNFL

Jon Hartley, Somerset Scientific Services

John Hemingway, University of Liverpool

Elizabeth Howie, British Energy

Michael John, BNFL

Steven Judge, AEAT

Jonathan Kemp, Babcock Rosyth Defence

John Lees, BNFL

Richard Lukey , VLA

Ian McGregor, Rolls Royce

Desmond MacMahon, Royal Naval College

Bob Major, NNC

Fiona Miller, UKAEA

Martin Phillips, BNFL

Jennifer Poveda, King’s College Hospital

Graham Raw, AEAT

Bill Rogerson, BNFL

Martin Rushby, AWE

Paul Sharp, Canberra

Bob Sharrock, URENCO

Matthew Simpson, DRPS

Steve Waters, Hammersmith Hospital

Sandra Williams, British Energy

Mike Youngman, NRPB

Simon Jerome, NPL

John Keightley, NPL

Michaela Ciocanel, NPL

Piers de Lavison, NPL

Andy Pearce, NPL

1. Welcome

Those present at the forum were welcomed by Julian Hunt, Head of Centre for Ionising Radiation (CIRM) at NPL. He commented that this was the 4th new club to be formed through CIRM with funding for the secretariat provided by the National Measurement Policy Unit of the UK Department of Trade and Industry (NMSPU, DTI). He asked the users to remember that this was “their forum” with NPL simply providing the facilities. He then briefly described the current PFI initiative at NPL to provide new laboratories and outlined the new 3 year programme within CIRM.

2. Chairman’s Opening Remarks

John Makepeace welcomed the attendees and again stressed that the forum should be user led with the emphasis on joint effort in solving problems. If a problem were diverse and required further input, then a small working group of those interested members could be set up to address the issue.

3. Draft Terms of Reference

The draft Terms of Reference are attached to these minutes and were accepted with the following discussions.

3.1 Aims

The use of “UK” should be taken to mean “of benefit to the UK”. John Makepeace commented that care should be taken not to unnecessarily focus on Europe to widely.

3.2 Activities

“Traceability to national standards” was queried. The UK is not explicitly mentioned here as national standards may not sometimes be those of the UK. It was decided that initially GSUF meetings would be held on an annual basis. Attendees were generally happy that a register of interests and facilities within the UK be established, though a small concern was expressed over possible use of the list by manufacturers. It was pointed out that probably all of those people with an interest in the GSUF were already on the manufacturer’s mailing lists, however the situation would be monitored.

3.3 Membership

This is limited to the UK as the forum is funded by the UK NMSPU, DTI. However there should not be a problem in inviting suitable non-UK guests should the need arise. The emphasis of the forum should be on helping UK users. Desmond MacMahon commented that he believed involving the instrument manufacturers was a good idea.

3.4 Management

John Hemingway queried whether the forum could be held at venues outside of NPL. John Makepeace replied that this was a possibility if it could be done without major expenditure. Ian Adsley believed that holding the forum at NPL was preferable due to its more central location and relative ease of access. This latter view seemed to be the overall feeling of the forum.

3.5 Fees

No comments were made.

3.6 Liability

Simon Woods is to confirm that the wording is acceptable to the NPL legal experts.
Action: Simon Woods

Of the forum’s terms of reference in general, Ian Adsley enquired as to how much NPL resource was available. John Makepeace replied that the Secretariat is funded, but no additional monies were set aside for actions arising. However, if an important, justifiable programme of work were to be identified, then NMSPU could be approached for further funding.

4. Users Discussions

In order to gauge the range of interest areas of those at the forum, John Makepeace requested the attendees indicate the normal levels of activity being used.

< 0.4 Bq/g	12
0.4 - 100 Bq/g	9
100 - 10 kBq/g	11
> 10 kBq/g	7
All	12

Jonathan Kemp queried whether the lower limit of 0.4 Bq/g was too high. John Makepeace replied that specific problems with very low level measurements should preferably be referred to the NPL Environmental Workshops (contact Mike Woods, NPL).

4.1 Nuclear Data

The recent publication of an evaluated set of decay data in the new Radiochemical Manual (RCM) was noted. Ian Adsley commented that the data was available on CD-ROM and that AEAT hoped that there might be provision in software to offer a route to interfacing. This would require permission to extract and use the data from AEAT - Ian Adsley will investigate if this is possible.
Action: Ian Adsley

It was noted that, with the breadth of interest at the forum, users could look at the data for the radionuclides they were interested in and report back to the GSUF, via Simon Woods, if there were any major changes between the values that they normally use and those in the RCM. If there were then, as the data is referenced, the problems could be investigated (by Simon Woods in conjunction with Alan Nichols (AEAT)).
Action: All
By e.o. April

Of 20 people present at the forum who use software analysis packages, only 3 used the data libraries as supplied. This somewhat surprised the manufacturers present. David Anstee commented that if a single data set were to be adopted then it would have to be a single, definitive recommended data set. The manufacturers would have problems if, e.g., a data set was constructed of half JEF data and half ENSDF data; it would still not be clear as to which data library is the more correct and where data should be used from.

It was generally agreed that, somehow, consensus on a single data set to be used must be achieved, though the way forward is unclear.

4.2 MDAs

The most widely used formula for calculating MDAs is that of Currie (1968). However there are many more about, so it is unclear as to which to use and when. Jim Fitzgerald commented that this is a fundamental problem and depends on what one is trying to say, i.e it is heavily dependant on the requirements of the end-user.

John Hemingway informed the forum that one of the MDA options in GammaVision is the German DIN standard and asked whether this should be taken up as the recommended formula and why there is no European agreement on the matter.

There are further problems in important aspects such as bidding for tendered work. Unrealistically low detection levels may be quoted, thus securing work, but often no definition of how these levels were reached is given.

The forum felt that this was an important, complex area which needed further discussion and investigation outside of the present meeting. A sub-group was thus formed to address the issue. Parties expressing interest were

Tim Sanders (AEAT via Ian Adsley)
John Hemingway
Simon Jerome
Weston Case (Nycomed-Amersham via Steve Judge)
Jim Desmond
Julie Eccles
Roger Benzing

(Note: As with all forum sub-groups, participation is not restricted to those expressing interest at the time. If anyone else is interested, then they should contact Simon Woods (e-mail: simon.woods@npl.co.uk, fax: 0208 943 6161) who will keep them informed of developments.)

Simon Woods to liaise with interested parties and provide meeting and contact facilities.
Action: Simon Woods

4.3 Millennium Compliance

It was agreed that this is a problem. Jim Fitzgerald pointed out that while software/hardware may not seem compliant e.g. does not use 4 figure years, some will undoubtedly still work after 2000, it is simply that the manufacturers have decided not to continue support for the products. David Anstee commented that this was practical as there is too much hardware to fully test and certify as millennium compliant. Users should not be afraid of seeking advice from the manufacturers of systems, who have indicated that they will be happy to advise on specific problems.
Any users experiencing problems with Millennium compliance, or finding solutions, should contact Simon Woods who will disseminate the information to other forum members.
Action: All
Action: Simon Woods

4.4 Peak Fitting

Michael John queried as to whether the fitting of singlet peaks was worthwhile. Whilst no definitive answer was forthcoming from the forum, Desmond MacMahon commented that the actual peak fitting normally does not pose much of a problem, it is the determination and fitting of the continuum background that causes most uncertainty. In particular, automatic software selection of peak boundaries, followed by variation, should be avoided if possible.

4.44 Peaks on high backgrounds appear to be a particular problem with a 15% - 20% underestimation of their area. As with other technical problems, the forum felt that the best route would be to e-mail/fax their problems/solutions to Simon Woods who would distribute the information as appropriate.
4.45 Action: All
4.46 Action: Simon Woods

5. Invited Talk : Validation of Gamma-ray Spectrometers Within QA Systems
Steven Judge, AEA Technology

The validation of systems was described with particular reference to the constraints placed upon pharmaceutical companies by the US Food and Drugs Administration. During isotope production, various measurement techniques are employed such as ion chambers, liquid scintillation and gamma-ray spectrometry, however, it normally is the latter which is used as the final check, thus strong QA procedures must be in place.

As validation can take in the order of 3 months, it must be done in advance of measurement; time pressures do not allow on-going validation.

For system validation, a sound overall knowledge of its capabilities must be demonstrated. These capabilities cover:

• Specific - can the system detect the radionuclides in question?
• Linear - are measurement parameters, e.g. dead time, well defined?
• Range - is the system proven to cover the activity range required?
• Accurate - are measurements traceable to national standards?
• Precise - does the system demonstrate reproducibility?
• Robust - are results independent of the operator?
• Detection Limits - what are they and how are they defined?

There then follows the documentation, which must be clear, written, appropriate, of verified accuracy and approved. At Nycomed-Amersham, 7 levels of documentation are generally used:

• Overview - i.e. introduction.
• User requirements - specific measurement purposes.
• Response - description of system to meet user requirements.
• Installation qualification - components delivered and installed.
• Operational qualification - completion of tests required by the FDA.
• Process qualification - demonstration of robustness.
• Approval - by named, qualified personnel.

The above lead to a systematic approach to system validation. Such an approach does lead to various benefits which can be summarised as:

• Passes audits.
• Saves time later.
• Projects can be easily handed over.
• “Sleep at night”.

Q. How often does a system have to be validated?
A. The Response at each end of the energy range is checked daily. The system has to be revalidated, at least in part, if new components or software are purchased.

Q. Is this the case if only a small component is changed?
A. Revalidation may not be necessary. What is important is to demonstrate the system is controlled in an overall manner.

It was suggested that the invited talk for the next forum should be on sources of nuclear data; where they are derived from and how they interconnect.

6. User Discussions Continued

6.1 Standard Spectra

It was noted that the NPL spectra produced for the testing of analysis software were still available. Problems may occur with the set 1 spectra in .chn format; a PC fault caused the time information in the files to become corrupted. This has been rectified and anyone experiencing a problem should contact Simon Woods who will issue repaired files.

Users of the spectra were asked, if further spectra were to be produced, what features they would like to see in them. A number of suggestions were made.

1) The time of collection and the reference date should be different. This will prevent spurious identification of short lived radionuclides and enable decay corrections to be checked.

2) Calibration and test spectra should have equivalent statistics. If, say, the calibration has low statistics and the test spectrum has high statistics then a small bias in the analysis can be introduced.

3) Spectra containing short lived nuclides were thought to be useful.

It was noted that the spectra demonstrate that major problems still occur when true coincidence summing is present. In general, software can’t correct for this summing effect. However, it would be of use to users if more analysis programs were to indicate that summing might be affecting specific peaks in a spectrum.

The spectra were thought to be useful. However one should remember that the report does not constitute a rigorous statistical analysis. If a user would like such, then the 1995 IAEA test spectra would be of more interest. (Blaauw et al, Nucl. Instr. And Meth. A387 (1997) 416, the spectra themselves are available at http://iriaxp.iri.tudelft.nl/~rc/fmr.html) Any further feedback on the NPL spectra should be sent to Simon Woods.

6.2 Field/Mobile Measurements

Ideally, samples are taken to the gamma-ray spectrometry laboratory for analysis. However, when large amounts of material are involved e.g. 1 m3 of concrete, field surveys have to undertaken. At present there is no approval mechanism for such measurements, current Environmental Agency approval is done on a case to case basis.

It was felt that a more formal scheme would be desirable, possibly utilising ANSI or IEC documentation. A forum sub-group was formed to investigate possibilities. Those at the meeting expressing an interest were

Ian Adsley
Bob Major
Michael John
Desmond MacMahon
Matthew Simpson
Peter Clark
Julie Eccles
Simon Jerome
Martin Phillips
David Anstee

6.3 High Count Rate Spectrometry

Martin Phillips briefly described the measurements he undertakes on flux monitoring samples where the activities involved are 100’s MBq/g. There is pressure from the licensing authorities to participate in some form of intercomparison, thus there is a need to identify laboratories with similar problems and co-operate. Anybody interested should contact Martin Phillips at Magnox Electric, Berkeley (Tel: 01453 812101).

One other possible solution to the problem might be to seek UKAS accreditation in this area.

6.4 Low Energy Photon Measurements

The identification of gamma- and X-rays when they fall in the same energy region (50 - 100 keV) is a problem due to the X-ray peaks being multiplets. There is good software available to analyse either gamma-rays or X-rays but not a combination. There are further complications as measurements at this energy are very dependant on the type of source matrix.

The forum felt that NPL should recommend good measurement practice. As such a sub-group was formed to investigate the problems. Members expressing interest at the meeting were

Ian Adsley
Desmond MacMahon
Jim Fitzgerald
John Hemingway
Steve Judge

6.5 Detector Calibration Problems

Various problems with detector calibration were expressed.

• When using a mixed solution in a realistic matrix, e.g. filter paper, the Hg-203 may volatolise to a greater or lesser extent.
• Efficiency calibrations are not good below 120 keV for real samples.
• In close geometries there are summing problems with Co-60 and Y-88.
• Different types of efficiency fit give different answers.
• How should a user estimate their uncertainties?

As a solution to the Hg-203 problem Steve Judge commented that a mixed radionuclide solution containing Cr-51 instead of Hg-203 was available from Nycomed-Amersham.

The forum generally felt that, as a supplier of such calibration sources, NPL should investigate and try to quantify these problems, reporting back at the next forum.
Action: Simon Woods

6.6 Intercomparisons

Several members of the forum expressed a desire to have intercomparisons at higher activity levels than those in the environmental intercomparisons organised by NPL (by a factor of 10). In particular, interest was expressed in the Ra-226 and Ra-228 decay chains.

A questionnaire will be sent out to members of the forum asking what sort of intercomparisons they would find useful.
Action: Simon Woods

Michael John commented that a standard was needed on how to conduct intercomparisons, in particular to conform to the BS standard, BS ISO 5725.

6.7 Software Problems

Roger Benzing described a problem he had with the latest version of GammaVision, whereby the date and time of start of acquisition can change with a change in the PC time. Apparently this facility was included to facilitate networking of acquisition PCs across different time zones.

An effective solution to this would be to ensure that the PC time remains accurate using radio controlled clocks. These simply plug into the back of a PC, are synchronised with the Rugby time signal and cost less than £100. One known supplier is Galleon Engineering Ltd, whose products are available through Inmac, Borehamwood, (Tel: 0990 275 422), but there may well be others.

7. Concluding Discussions

Those present felt that the forum had been useful and was a success. Many problems had been aired and these topics could now be expanded on.

The next forum would be arranged for October 1999.

Simon Woods 13 January 1999

GAMMA-RAY SPECTROMETRY USERS FORUM - GSUF

DRAFT TERMS OF REFERENCE

Aims

The Forum shall aim to facilitate both the exchange of information about UK gamma-ray spectrometry facilities and measurement techniques and the efficient use of those facilities by the user base. It shall represent members’ interests in discussions and correspondence with expert bodies in the field of gamma-ray spectrometry.

Activities

The Forum shall:

• encourage, by all means which shall be deemed appropriate and practical, good practice in measurement, including traceability to national standards.

• hold regular meetings to discuss matters of common interest relating to calibration and measurement, including analysis techniques.

• maintain and disseminate a register of UK gamma-ray spectrometry facilities and services.

Membership

Members shall be representatives of UK establishments or organisations actively involved in measurements in the field of gamma-ray spectrometry, together with representatives of manufacturers of instruments and artefacts pertaining to such measurements.

Management

The Forum shall be managed by NPL, assisted by advisory committees drawn from and established according to the wishes of the members. Those responsible for the management of the GSUF shall take due account of views of the members. NPL shall provide the Chairman and Secretary of the GSUF.

Fees

There shall be no annual subscription but NPL shall reserve the right to make a charge for organising meetings, producing reports and providing services.

Liability

Neither NPL nor the GSUF nor any member thereof shall be responsible for any consequence arising from the activities of the GSUF.

October 1998