held at the National Physical Laboratory
Thursday 7th October 1999

Present:
J   Allen		Marconi Marine, Barrow-in-Furness
M Armishaw		AEA Technology, Winfrith
N Baghini		Imperial College, Ascot
D  Bartlett		National Radiological Protection Board, Chilton
P  Beeley		DNST, Gosport
R  Bosley		AEA Technology, Harwell
J   Caunt		John Caunt Scientific, Oxford
N  Chapman		Imperial College, Ascot
P  Chard		Harwell Instruments
M Cooper		AWE, Aldermaston
A  Corns		DERA RPS, Gosport
S  Crotch-Harvey		Blue-C Ltd, Barrow-in-Furness
I   Dalton		Marconi Marine (Barrow)
T  Daniels		National Radiological Protection Board, Chilton
C  Dumper		DERA Aquila, Bromley
C  Dyer		DERA, Farnborough
J   Ferguson		University of Lancaster
N  Hawkes		Harwell Instruments
M Healy		University of Cranfield
S  Hughes		AWE, Aldermaston
M Joyce		University of Lancaster
S  Marriott		Marconi Marine (Barrow)
B  Mason		Centronic, Croydon
D  McClure		National Radiological Protection Board, Chilton
M Mills		AWE, Aldermaston
A  Pereira		Reeves Technologies, Loughborough
T  Peyton		University of Lancaster
S  Shannon		John Caunt Scientific, Oxford
J   Silvie		Marconi Marine (Barrow)
N  Spyrou		University of Surrey
R  Stokes		DERA RPS, Gosport
Z  Tabatabain		University of Surrey
R  Tanner		National Radiological Protection Board, Chilton
J   Weaver		University of Lancaster
A  Williams		University of Surrey

and the following members of the Centre for Ionising Radiation Metrology, NPL:-

V E Lewis, D J Thomas, Section Heads, Neutron Measurement Group

M Burke, N Horwood, P Kolkowski, N Roberts, H Tagziria, G C Taylor, J R Winnington

Apologies had been received from:

J   Barnes		NE Technology, Reading
R  Benzing		Imperial College, Ascot
J   Caunt		John Caunt Scientific, Oxford
P  Coad		JET, Culham
M Chandler		AWE, Aldermaston
S  Croft		AEA Technology, Harwell
S  Franklin		Imperial College, Ascot
L  Howard		Centronic, Croydon
J   Harvey		Consultant
J   Roskell		Marconi Marine (Barrow)
R  Samworth		Reeves Technologies, Loughborough
T  Wilson		DERA Aquila, Bromley

 

1 WELCOME AND INTRODUCTION

On behalf of Julian Hunt (Head, Centre for Ionising Radiation Metrology) who could not attend, Vic Lewis welcomed the participants to NPL, in particular those who were visiting for the first time. The intention of these meetings was not only to disseminate information about research work undertaken at NPL but also for participants to talk about their facilities and programmes, and to discuss their problems and concerns. The success of these meetings depended upon an active participation by all.

The Neutron Users Club was run by NPL and funded under the National Measurement System (NMS). This is organised mainly in three-year programmes, one of which is the NMS Ionising Radiation Metrology Programme which began in October 1998. The formulation process for the next NMS Ionising Radiation Programme, due to start in 2001, would begin about April 2000 and would involve the consultation of anyone with an interest in the production, use and metrology of ionising radiation, including members of user groups such as the Neutron Users Club.

2 FACILITIES, STANDARDS and SERVICES of the NPL NEUTRON MEASUREMENT GROUP

Vic Lewis described the extensive accelerator facilities at NPL. All ISO-recommended neutron energies (from 27 keV to 19 MeV) can be obtained. The 3.5 MV Van de Graaff is used for the production of monoenergetic neutron fields with energies in the range 8 keV to 6 MeV using various reactions, and from 13 to 20 MeV using the d+T reaction. The neutron fluence rates produced using the ⁴⁵Sc(p,n) reaction are less than 8 cm^-2 s^-1 at 1 m (dose equivalent rate of 6 µSv hr^-1).  Fields produced using the d+D reaction with neutron energies higher than 5 MeV are not clean due to competing and contaminant reactions. Standards are restricted at lower energies to the range from 27 keV to 5.0 MeV, but the lowest energy for routine calibrations is 36 keV. NPL aims to establish better standards in the energy range between 8 and 36 keV, and then to extend to even lower energies. In the higher energy range there are problems due to neutrons produced by reactions of the deuteron beam with contaminant oxygen, carbon and deposited deuterium in the tritium targets. The Van de Graaff is not used routinely for d+T reaction due to the shortage of fresh tritium targets.

The 150 kV SAMES accelerator is used to provide more intense neutron fields in the energy range from 14.1 to 14.7 MeV. Emission rates of up to 2 x 10⁹ s^-1 are produced, yielding fluence rates of about 2.5 x 10⁴ cm^-2 s^-1 at 1 m (50 mSv h^-1).

Thermal neutron fields are produced by moderating fast neutrons produced by the interaction of 2.8 MeV deuteron beams on two beryllium targets in a graphite pile. A servo-mechanism controlling the beam steering ensures a medium term stability of better than 0.4% for fluence rate. Thermal neutron fields of up to 3 x 10⁷ cm^-2s^-1 can be produced in a small central cavity. Routine irradiations and calibrations of radiation protection monitors are performed using a thermal beam extracted from the pile where fluence rates of up to 3.5 x 10⁴ cm^-2 s^-1 (1.1 mSv h^-1) are available. The fields are standardised using gold foil activation. The spectrum of the non-thermal component of the thermal neutron beam has been measured using a set of Bonner spheres.

The UK standards are disseminated to customers through a range of measurement and irradiation services, all of which have been accredited by UKAS. Details of measurement and testing services provided by NPL and names of contacts are given in the NPL booklet Measurement Services. Further details of NPL and in particular the Centre for Ionising Radiation Metrology and its activities could be found on the Website at http://www.npl.co.uk/ionrad/.

3 RADIONUCLIDE NEUTRON SOURCES

Martin Burke (MB) described the two NPL manganese sulphate baths used for the absolute determination of the emission rate of radionuclide neutron sources. This involves the measurement of the activity of the manganese that has been activated by the thermalised neutrons. Corrections have to be made for the absorption of the neutrons by other constituents such as hydrogen, oxygen and sulphur. Corrections for the escape of neutrons from solution are determined using a long counter positioned close to the bath. A moderating detector system is also used as a secondary standard to measure emission rate by a comparative method. The NPL possesses a range of radionuclide sources which are mainly ²⁵²Cf, and ²⁴¹Am-Be sources and include also ²⁴¹Am-F, ²⁴¹Am-B, ²⁴¹Am-Li and ²³⁹Pu-Li.

Radionuclide sources are normally cylindrical in shape and, to ensure safe containment of the active components, doubly encapsulated in a material such as stainless steel. The resulting shape produces anisotropic emission of neutrons about the axis of the source. Typical examples of source encapsulation were shown. MB described the method used to measure anisotropy factors; this involved measuring the neutron emission variation with respect to the axis of the source using a long counter in the ‘low-scatter’ area at NPL. The use of the shadow cone enables subtraction of the room and air-scattered component of the counter response to be made.

4 MONTE CARLO MODELLING and SENSITIVITY ANALYSIS of the NPL LONG COUNTERS

Hamid Tagziria (HT) described the re-calibration of the NPL Long Counters (LCs) by modelling the response functions using the MCNP code.

At NPL, neutron fluences are standardised using de Pangher precision long counters (PLCs) which have a “near flat” response from 1 keV to 5 MeV, but require calibration because they are not 100% efficient. This involved:

• measuring the responses with well characterised radioactive neutron sources,
• simulating the response functions by Monte Carlo radiation transport techniques,
• normalising the calculated response functions using the source measurements

The counters were last calibrated more than twenty years ago using a similar approach to that above, although the response functions could only be empirically predicted using available hydrogen and carbon cross-section data. Regular source checks showed that the counters had been stable over the last 20 years. There were three main reasons for recalibration using MCNP:

• source spectra are now better known;
• cross-section data has improved and is more abundant;
• the Monte Carlo method in particle transport has become more sophisticated due to advances in computing power and improved cross-section libraries.

The de Pangher and McTaggart LCs used at NPL have been re-calibrated using five broad spectrum neutron sources (²³⁹Pu-Li, ²⁴¹Am-F, ²⁵²Cf, ²⁴¹Am-B and ²⁴¹Am-Be) and monoenergetic neutron fields, produced using the Van de Graaff accelerator, at energies of 144, 565, 1200, 2500 and 5000 keV. Measurements were carried out in the low scatter cell at distances ranging from 1.2 m to 5 m to determine the effective centre and efficiency of the long counters. The shadow cone technique was used to correct for inscatter.

Extensive Monte Carlo modelling of the response functions of both detectors was carried out using MCNP-4B, a general purpose Monte Carlo N-Particle transport code developed originally at Los Alamos in 1943. Details of the construction of the long counters were obtained from engineering drawings and from X-rays of the boron trifluoride tubes.

The calculations and source measurements agreed extremely well, especially for the de Pangher long counter whose design specifications are better known. A normalisation of only 3% was needed to adjust calculations to all source measurements.

New, more accurate response functions have been produced for both LCs. A novel approach using Monte Carlo methods has been developed, validated and used to model the response functions of the counters and determine more accurately their effective centres, which has always been difficult to establish. The sensitivity of the Monte Carlo calculations for the de Pangher LC to perturbations in the density and cross section of the polyethylene used in its construction has been investigated.

The most extensive modelling ever done for precision long counters has been carried out at NPL. This will result in improved standardisation and calibration services. The power of combining Monte Carlo radiation transport techniques with sound measurements has been demonstrated by this work.

5 The ASP ACCELERATOR at AWE, ALDERMASTON

Shaun Hughes (SH) described the accelerator facilities and services offered at AWE, Aldermaston. The ASP accelerator was installed and commissioned at AWE in 1964; the design was a joint venture between the High Voltage Corporation, USA, and the United Kingdom Atomic Energy Authority. The accelerator was designed to produce intense 14 MeV neutron fields using the deuteron bombardment of a tritium loaded disc  via the T(d,n)⁴He reaction. Since its installation it had been modified to enable operation in the pulsed mode. The accelerator is the most powerful of its type in the UK and can also produce 3 MeV neutrons using the D(d, n)³He reaction.

The deuteron ion source of the accelerator is a Duoplasmatron type. The deuteron beam is repelled from the source by the application of a 50kV, negative anode voltage and passes through an analysing magnet before being accelerated. Generally, molecular deuteron beams with an energy of 300 keV are produced. The tritium-loaded target is cooled either by compressed air or chilled water depending on the type used. Neutron fluence rates of up to 2.0 x 10¹⁰ cm^-2 s^-1 at 10 mm and 0^° from the target face can be produced, with a mean energy of about 14.7 MeV. Two ²³⁸U fission counters are used for routine monitoring of the neutron output. Associated alpha particle monitoring is used for absolute calibration of the neutron field.

Pulse rates of up to 200, 000 s^-1 can be achieved, in the pulse mode, and with a pulse width of 1.5 µs with typical fluence rates of 1 x 10³ cm^-2 s^-1 at 10 mm and at 0° from the target face.

The accelerator is currently used for non-destructive testing employing activation analysis. A pneumatic system transports the irradiated sample in less than six seconds from the irradiation position to the room in which the induced gamma ray activity is analysed. This is essential for short lived isotopes. The technique can detect light elements in concentrations as low as fifty parts per million. Other work being carried out at ASP includes linearity tests on portable neutron monitors, radiation hardness testing and calibration of diode dosimeters.

Future work is expected to include alpha particle irradiations using the T(d, n)⁴He reaction and also the development of ASP as a neutron radiography system.

6 TRITIUM TARGET PRODUCTION at AWE, ALDERMASTON

Michael Mills (MM) stated that AWE was now managed and operated by Hunting-Brae on behalf of the Ministry of Defence and that they were now encouraged to diversify and produce tritium targets for customers other than the MoD.

The AWE had the facilities and experience for handling large quantities of tritium (50 up to 70 Ci) and the production of good quality tritium targets. MM showed some comparisons of other tritium target manufacturers with targets produced by AWE.

Manufacturer	Identifier	Neutron Flux n cm-2 (x 109)	Tritium* Loading Ci	Average Flux n cm-2 (x 109)
Supplier 1	a b c	4.7 5.9 6.8	60 64 55	5.8
Supplier 2		5.0	50
AWE	a b c d	12.0 8.5 9.9 9.2	52 52 52 52	9.9

*    Tritium content as estimated by the manufacturer.

The AWE targets showed 40 - 50% increase in neutron output and were more durable than those produced by other manufactures.

At the present time, MM could not quote any prices for targets since it depended on the charge that MoD levied for the tritium. Also, the manufacturing of the targets was a batch process and the unit cost per target depended on the number in the batch. He thought the cost would be very approximately between  £300 to £600 if a complete batch of ten targets are produced. Anyone interested in obtaining tritium targets should contact MM for an accurate quotation.

7 The VIPER REACTOR at AWE, ALDERMASTON

Michael Cooper (MC) described the Versatile Intermediate Pulsed Experimental Reactor (VIPER), which had been operating since May 1967 and was the first fast pulsed reactor in Western Europe, and has a unique variable core composition. VIPER provides isolated pulses of intense neutron and gamma radiation for a wide range of radiation effect studies.

VIPER differs from conventional reactors in that two classes of neutrons are produced in the fission process: ‘prompt’ neutrons, emitted virtually instantaneously; and ‘delayed’ neutrons, emitted during the radioactive decay of the fission products, over a much longer timescale - the various groups of such emitters having half-lives ranging from around 0.15 s to over 50 s. The delayed neutrons constitute, in the case of U235 fission, only 0.75% of the total neutron production and are of great importance in reactor control, because they effectively lengthen the timescale on which the reactor changes take place in normal operation of conventional systems.

MC described the concepts of criticality, and reactivity. A reactor operating at any power level which is constant with time (assuming there are no extraneous sources) is said to be ‘critical’, or, more precisely, ‘delayed critical’, in recognition of the fact that the 0.75% of delayed neutrons are required to keep the neutron population constant. While conventional systems are not allowed to reach reactivities much over delayed critical, pulsed reactors are designed to operate into the region of reactivity above prompt critical in order to achieve a short burst of very high power.

Having obtained a rapidly rising power by suddenly placing the reactor in a super-prompt-critical condition, however, the next requirement is to terminate the rise by some guaranteed mechanism which will operate in the time available, and bring the power level first to a peak and then back to a low level. In VIPER, as in other fast pulsed reactors, this mechanism is inherent in the system physics, viz., a negative temperature coefficient of reactivity. As the reactor power rises the fuel is heated, the reactivity reducing as the temperature rises, due to fuel expansion and to the nuclear Doppler effect. It can be shown that the reduction in reactivity during the pulse is such as to leave the system finally as far below prompt critical as it was initially above. The reactor can then be shut down below delayed critical by mechanical means. At first, a step reactivity input is made to above prompt critical, whereupon the power begins to rise sharply; at 2.5 ms, the temperature rise starts and reactivity correspondingly reduces. At 3.0 ms the power reaches a  maximum, the rates of temperature rise and reactivity fall, being at their highest at this point. After this reactivity is below prompt critical, and the power falls rapidly to stabilise at about 4 ms, when the temperature rise is comparatively slow and reactivity is almost constant. This situation continues until the system is shut down. The maximum fuel temperature is 450° C.

The versatility of the VIPER system lies in the ability to vary the time width of the pulse, between approximately 400 microsec and 6 msec, by variation of the core composition. The energy spectrum of neutron flux in VIPER is intermediate between the very fast (almost fission spectrum) typical of solid metal systems, and the thermal spectra of, for instance, AGR and Boiling Water power reactors, and TRIGA thermal pulsed reactors.

The VIPER core is made up of cylindrical fuel elements stacked in vertical channels in a matrix of horizontal layers of moderating and diluting materials. There are three sections of core, a fixed central block and two movable ‘safety blocks’ on either side of it. The latter  are hydraulically driven and can be dropped out rapidly when required, to shut down the reactor. Fuel element/matrix arrangements are similar in all three blocks. Reactor control is effected by four boron-loaded absorber rods positioned at the edge of the core, and to provide the sudden reactivity jump to above critical for pulsing, a ‘pulse rod’ consisting of a slug of fuel material can be fired pneumatically into the central hole. A copper reflector surrounds the core on all sides.

The matrix layers consist of plates of copper as diluent  and aluminium-filled epoxy resin moderator. Variation of the proportion of these two materials is used to obtain the desired versatility in pulse width. Rods of moderator or diluent can be substituted for some of the fuel if further composition change is required.

The fuel rods are 1.0 cm diameter and 29.0 cm long, of uranium alloyed with molybdenum (1.4% weight), the U235 enrichment being 37.5%. The design ensures that the rapid expansion of the fuel during a  pulse is symmetric about its point of attachment and does not produce impulse forces on the reactor structure. Over six hundred standard fuel elements are used in the present VIPER core arrangement. The pulse rod is a 4.1 cm diameter cylinder of the same fuel material  as the standard elements, and is also enclosed in stainless steel.

Two cavities have been provided to accommodate irradiation samples of various sizes. The smaller of these is a 4.8 cm diameter hole entering the core region to within some 10 cm of the centre, and the larger is in the reflector next to the edge of the core. In a full-sized pulse of 3.5 x 10¹⁷ fissions, with peak power of approximately 2 x 10⁴ MW, the neutron fluence and gamma dose in the central cavity will be 10¹⁵ cm^-2 and 300 Gy, respectively. Levels of about one third to a half of these are produced in the reflector cavity.

VIPER is the only UK facility capable of testing defence systems for radiation hardness. Items tested include Trident, battlefield, communications and aircraft equipment. Additional work being pursued includes drug treatment effectiveness, seeds for prostate cancer treatment, BNCT cancer treatment, and neutron radiography. Support is given to university physics and engineering education

Any members interested in visiting VIPER should contact MM, who would make the necessary arrangements.

8 RELOCATION of the DEPARTMENT of NUCLEAR SCIENCE and TECHNOLOGY of the ROYAL NAVAL COLLEGE to HMS SULTAN and the STATUS of NEUTRON METROLOGY RESEARCH

Philip Beeley (PB) reported that DNST had been re-located from Greenwich to HMS Sultan, Gosport in October 1998. The JASON reactor at Greenwich was almost de-commissioned. HMS Sultan is the naval school of air and marine engineering, providing support for both civilian and naval requirements in the nuclear propulsion programme. The department organises courses of differing complexity on nuclear reactor technology.

Present work being undertaken by the Neutron Metrology Group at DNST includes -

• development of fixed active area dosimeters for use in naval vessels; information from crew duty rosters is used to determine doses, eliminating the need for many personal dosimeters;
• improvements and modifications to the transportable neutron spectrometer (TNS) to improve the sensitivity, portability and reliability;
• neutron activation analysis in collaboration with Imperial College.

PB describe the irradiation facilities at the new location which include a PANTAK X ray set and a ¹³⁷Cs/⁶⁰Co gamma-ray irradiation facility.

It was proposed to purchase a 14 MeV neutron generator. This neutron generator was likely to be a French machine which can produce 10⁸ neutrons s^-1 in the dc mode and up to 10¹¹ neutrons s^-1 in the pulse mode. The initial design stage is underway for the biological shield and experimental areas. The machine will be used for both teaching and research and proposed work to be carried out will include activation analysis, fast and thermal neutron measurements, prompt gamma experiments and time of flight measurements.

9 NEUTRON SOURCES USED IN BOREHOLE LOGGING

Alex Pereira (AP) described the techniques used by Reeves Technologies in borehole logging and other evaluation services for the oil industry. It is essential to know the type of rocks that oil companies will be boring into, and to do this various instruments electrical and radioactive sources are lowered into the borehole. To determine if it would be cost effective to extract oil, the companies need to know if they have bored into the correct reservoir and also how porous and permeable the rocks are, what fluids are contained within and if the pores are connected.

Both radioactive neutron sources and d+T neutron generators, steady and pulsed, are used. The logging is carried out at depths of down to 3 miles and in temperatures ranging from -40°C up to 150°C and pressures of up to 125,000 psi. The logging is carried out at a very rapid rates with the probes moving at 10 m per minute and up to 30 m per minute for special applications.

Porosity is measured by the moderation of neutrons from a source using two detectors, at some distance from the source, that measure the epithermal neutron flux. However, not all the hydrogen is combined in the rocks and many correction factors are applied to obtain a true porosity value. Other more sophisticated methods use gamma detectors, in conjunction with the neutron detectors, to determine the individual element concentrations and computer modelling is used to characterise the tools used for borehole logging.

Very large sodium iodide detectors are used to measure the natural gamma background due to potassium, thorium and uranium in the rocks. Other tools are used to make electrical measurements to determine conductivity, salinity and magnetic properties, as well as ultrasound to measure the density of rock formations. AP described how the combination of the results from different tools could indicate the presence of hydrocarbons. These tools are large and therefore cannot be used in highly deviating or curved wells. Current technology being developed is leading to slimmer, smaller tools which can pass along sharp curvatures and along horizontal boreholes.

10 IMPLICATIONS of the NEW ICRP 60 QUANTITIES for NEUTRON DOSIMETRY

David Thomas (DJT) talked about the implications of the changes to radiation protection quantities recommended by the International Commission on Radiological Protection (ICRP) in ICRP Publication 60. As these will be implemented in the new UK Ionising Radiations Regulations due to come into force around the turn of the year, this was an appropriate time to try to clarify them.

To explain the changes, it is necessary to understand the system in place prior to the recommendations of ICRP 60. For this reason, DJT gave an outline of  the basic quantities, absorbed dose and dose equivalent, explaining why these on their own are not adequate for describing the detrimental effects of ionising radiation, and why effective dose equivalent was introduced in an attempt to define a quantity which is a reasonable measure of the risk to an individual. Pre-ICRP 60, effective dose equivalent was the ‘protection quantity’ in terms of which radiation limits were set. However, effective dose equivalent was deemed to be unmeasurable, and therefore the so-called operational quantities, ambient dose equivalent, directional dose equivalent, and personal dose equivalent were introduced by the ICRU as the quantities in terms of which instruments and dosemeters should be calibrated. The properties of these quantities and their relationship to the protection quantity were described by presenting the various neutron fluence to dosimetric quantity conversion coefficients.

The changes proposed in ICRP 60, i.e., the introduction of the radiation weighting factors and a change of the quality factor dependence on linear energy transfer (LET) were then described. As the use of operational quantities as a measure of the protection quantity is still recommended, the implications of these changes on the relationships between protection and operational quantities were outlined. Although the operational quantities are no longer an overestimation of the protection quantity at all energies, they can still be considered acceptable for the types of neutron fields in which nuclear workers are exposed.

Finally, the implications for calibration of area survey instruments and personal dosemeters were illustrated by showing the changes in the spectrum-averaged fluence to operational quantity conversion coefficients for commonly-used calibration sources (Am-Be, ²⁵²Cf) of using the new quality factor - LET relationship.

11 The TEPC as a RADIATION PROTECTION INSTRUMENT

Graeme Taylor (GT) described the construction of a tissue equivalent proportional counter (TEPC). It is a simple gas-filled detector constructed from materials with a composition similar to that of living tissue (muscle). The main compromise in both the wall material and the gas filling is that the high oxygen content of living tissue (from its high water content) has to be substituted by carbon for reasons of stability. The hydrogen and nitrogen contents of the materials are matched more or less exactly. By applying a suitable voltage to the central anode, the output pulse from the detector is proportional to the amount of energy deposited in the gas by the passage of charged particles.

The TEPC differs from other dosemeters in that it mimics a microscopic volume of living tissue that is smaller than a typical human cell. It is able to do this because a charged particle loses the same amount of energy crossing a small, high density thickness of material as it does crossing a larger thickness of a proportionately lower density material of the similar composition. Thus a TEPC about 12 cm in diameter can simulate a microscopic volume of tissue if filled with tissue equivalent gas at the appropriate pressure (of the order of 0.1 atmospheres).

The advantage of using this approach is that the detector is a much closer simulation of the system being investigated (i.e. radiation interactions with living tissue) than the traditional moderating/thermal capture area survey meters. By using tissue equivalent materials, the number, type and energies of the secondary charged particles generated by neutron interactions in the detector will be similar to those created in living tissue exposed to the same irradiation conditions. Consequently, studying the patterns of energy deposition in the detector gives a useful insight into the energy deposition occurring in living tissue.

Simulating a region of tissue on the micrometer scale not only has biological relevance in terms of the cell's substructures (organelles); it also means that the energy deposited by crossing charged particles contains information about the particles' LET. In fact the spectrum obtained from a TEPC is often regarded as being an LET spectrum (although this is something of a simplification). Nevertheless, the results can be combined with the ICRP Q(L) factors to generate a mean quality factor for the energy deposited, enabling the calculation of a value for the dose equivalent. Hence the TEPC can provide a measurement of dose, mean quality factor and dose equivalent for a radiation field in a single measurement. Furthermore, the LET information contained in the spectrum can be used to separate the photon and neutron components.

Although this makes the TEPC a very powerful tool, it does have certain disadvantages: it has a lower counting efficiency than moderating detectors, it is more complicated to use and has a poor ambient dose equivalent response to intermediate energy neutrons (~1 keV - ~100 keV). This last point is not surprising considering that the lower energy charged particles produced by the neutron interactions deposit energy in a way that degrades the information on the particles' LET. It should also be remembered that ambient dose equivalent is defined 10 mm inside the surface of a 300 mm diameter sphere, which allows plenty of backscatter when compared to the thin-walled TEPC, which has less than 1% of the sphere's mass.

Recent work at NPL has sought to correct for the poor ambient dose equivalent response to intermediate energy neutrons by means of a more sophisticated analysis regime. The shape of the spectrum gives a strong indication of the mean energy of the neutron field involved and this can be used to determine a correction factor for the measured dose equivalent.

One area of work which has seen increasing use of the TEPC is in the dosimetry of the cosmic radiation, particularly the exposure of air crew. The cosmic radiation fields extend to incredibly high energies (hundreds of GeV and above), although the bulk of the ambient dose equivalent from neutrons occurs between 1 MeV and 200 MeV. The TEPC has been shown to have a good ambient dose equivalent response over this energy range and thus makes an excellent cosmic ray dosemeter.

12 NEUTRON COINCIDENCE COUNTING of PLUTONIUM

Patrick Chard (PC) described how passive neutron coincidence counting (PNCC) for the non-destructive assay of plutonium-bearing material was based on the detection of time-correlated neutrons emitted by the spontaneously fissioning isotopes ²³⁸Pu, ²⁴⁰Pu and ²⁴²Pu. For usual types of nuclear material, the ²⁴⁰Pu fraction accounts for between 65% and 90% of the spontaneous fission neutrons; therefore the concept of a ²⁴⁰Pu effective mass is frequently used. The quantity is expressed as a weighted sum of the masses of ²³⁸Pu, ²⁴⁰Pu and ²⁴²Pu. Determination of the weighting factors can be made using nuclear data evaluations of the spontaneous neutron fission neutron half lives and multiplicity distributions for the three species. However, these calculated values were subject to uncertainties of about 2% to 3% which was not acceptable for safeguards in plutonium assay systems for high burnup materials.

To perform a new evaluation, Harwell was commissioned to produce several high purity, isotopically enriched plutonium oxide samples. Three ²³⁸PuO₂ samples, two ²⁴⁰PuO₂ and two ²⁴²Pu O₂ samples were measured in two independent PNCC counting chambers which are used for safeguards measurements. These included a Harwell Instruments N95 high efficiency neutron chamber (Euratom OSL - counter), as utilised in the Euratom on-site laboratories.

Measurements were made using neutron multiplicity electronics. Statistical uncertainties of better than ± 0.2% (at one standard deviation) were achieved on the coincidence count rates. An accurate determination of the samples’ plutonium masses and isotopic compositions was obtained by destructive analysis using mass spectroscopy of the samples of the same materials. This permitted the derivation of the ²⁴⁰Pu_eff coefficients for ²³⁸Pu and ²⁴²Pu by calculating  the experimental ²³⁸Pu and ²⁴²Pu count rates per unit mass relative to that for ²⁴⁰Pu. Correction factors for multiplication effects taking place within the samples were estimated within the spirit of Bohnel’s neutron coincidence counting model, using leakage self-multiplication factors for each sample that were generated using the MCNP code.

The measurement results for the N95 chamber indicate that the ²⁴⁰Pu effective mass for plutonium bearing material is given by:

m_eff   =   2.72 m₂₃₈   +   m₂₄₀   +   1.66 m₂₄₂

The coefficients obtained using the OSL counter gave very similar results, within the uncertainties, and suggest insensitivity to details of the respective counter assemblies. To good approximation, the determined specific ²⁴⁰Pu-effective coefficients for ²³⁸Pu and ²⁴²Pu could therefore be regarded as universal constants in the determination of the ²⁴⁰Pu effective mass.

The uncertainty on the values of these coefficients of better than 0.4% is almost an order of magnitude lower than the previously assumed values based on nuclear data. The improvement in the accuracy of the new coefficients reported in this work is important for accurate analytical PNCC measurements. The neutron coincidence response from a sample depends on numerous sample parameters, and the evaluation of sample-specific correction factors, as obtained in the present work, should become an integral part of the routine data analysis. Using this method, calibration of analytical PNCC measurements could be reduced to a normalisation measurement against a single reference sample, provided that there is an unbiased link between the ²⁴⁰Pu effective mass values of the reference sample. The ²⁴⁰Pu-effective coefficients for ²³⁸Pu and ²⁴²Pu determined in this work provide the means of meeting requirements.

13 RECENT DEVELOPMENTS with the TRANSPORTABLE NEUTRON SPECTROMETER (TNS)

John Weaver (JW) briefly gave the background to the TNS and described recent improvements and future development work to be carried out on the instrument. He described the construction of the TNS which is made up of a probe unit and a power supply unit. The probe unit contains six sensors - two BF₃ counters, three hydrogen gas counters and an NE213 liquid scintillator. The device is intended to cover the energy range from 50 keV up to 10 MeV.

Recent development work had reduced the weight of the power supplies by more than 50% and circuitry has been modernised. The present system uses analogue signal processing which is to be replaced with a digital technique using field programmable gate arrays that will improve the data handling and speed the production of results.

The TNS had had reliability problems and the digital techniques being developed will have ‘self test’ capability and the silicon components used are specifically chosen for reliability as well as test procedures being formalised to ensure correct operation of the instrument.

The TNS is used in radiation hazard areas such as submarine shielding and is also used for environmental correction factors for other neutron sensitive instruments. One of the advantages of TNS is that it produces broad spectrum, accurate data on neutron fields and it is an essential instrument for VSEL in the submarine programme.  The possibility of using the instrument for both gamma ray and neutron spectroscopy will be investigated.

14 COSMIC RADIATION EFFECTS on MICROELECTRONICS from INTERPLANETARY SPACE to SEA LEVEL

Clive Dyer (CD) reported that, with the increasing use of microelectronics with ever diminishing component size, systems are becoming increasingly susceptible to single event effects (SEEs) arising from highly ionising interactions of cosmic rays and solar particles. These SEEs include soft errors, involving both single and multiple bits, and hard errors due to latch-up or burn-out. For space systems an increasing body of evidence has accumulated over the last twenty years.

Cosmic radiation is composed primarily of energetic charged particles. These comprise 85% protons, 14% alpha particles, and 1% heavier particles covering the full range of elements stripped of all electrons. Most theories of their origin favour supernovae or the resulting pulsar. The cosmic rays have very high energies. The Earth’s magnetic field reduces the cosmic radiation striking the earth but this effect is much reduced at the poles. In addition to geomagnetic shielding, the Earth is greatly shielded by its atmosphere in which primary cosmic rays interact with air nuclei, generating a cascade of secondary particles comprising protons, neutrons and nuclear fragments. A build up of secondary neutrons occurs, reaching a maximum at about 60,000 feet, reducing by a factor of only three at 30,000 feet and slowly decreasing to sea level.

The penetration of these galactic cosmic rays into the atmosphere of the Earth is influenced by conditions on the sun which emits a continuous wind of ionised gas, or plasma to form a bubble of gas which extends beyond the solar system. This carries magnetic field lines from the sun; the strength of the solar wind and the geometry of the magnetic field influence the cosmic ray intensity.

The primary cosmic particles can deposit enough charge in a  small volume of silicon to change the state of the memory cell so corrupting the memory and leading to possible erroneous commands. These are soft errors and are referred to as single event upsets (SEU). When a particle upsets more than one bit this is known as multiple bit upsets (MBU).  Certain devices can be triggered into a state of high current drain, leading to burn-out and hardware failure; such effects are termed single event latch-up or single event burn-out. The interactions of individual particles are referred to as SEE to distinguish them from the cumulative effects of radiation or lattice displacements.

As well as directly ionising interactions with electrons, particles may interact by imparting recoil energy and generating secondary particles. These events when occurring in or adjacent to the depletion region of a semiconductor device may cause a SEE. Collisions with nuclei are less probable than those with orbital electrons but in the presence of intense fluxes of protons or neutrons this mechanism can dominate and this situation occurs in the earth’s inner radiation belt. This situation can also occur in the atmosphere where there is a build up of secondary neutrons and is thought to be the dominant SEE hazard for current and near-future avionics.

Cosmic ray effects are now a normal part of system specification. During the past ten years there has been increasing evidence of single event effects on aircraft electronics, and the problem is expected to increase as more low-power, small-component size avionics are used. New legislation was now in place to deal with the allied problem of the effects of these neutrons on aircrew and frequent flyers. The next maximum for solar activity, at the turn of the millennium, is likely to produce large solar particle events which penetrate to aircraft altitudes. The Cosmic Radiation Effects Working Group has been established to pool information on this problem.

Much evidence from the space industry shows errors, computer crashes and hardware failure resulting from radiation. These effects were predicted in 1962, but computer technology did not become sensitive until 1975. Numerous papers on these effects have been published and many examples of space craft malfunctioning have been recorded and related to cosmic ray effects.

The extent of the problem has led to software modelling and DERA have participated in electronic upset experiments and radiation monitoring onboard many flights of Concorde, a Boeing 767 and the Space Shuttle to gain information on the radiation type and intensity at different altitudes so that computer modelling techniques can be validated. The in-flight measurements were shown to be correlated with solar particle activity and cosmic ray activity. The comparison of results from computer modelling, when all the particles and interaction processes are taken into account, agree reasonably well with measurements and show the possibility of SEE in certain environments.

The importance of soft error production by radioactive contaminants, uranium and thorium, in semiconductor chip packaging has been recognised but despite efforts to eliminate this source, soft errors still occur at a rate of about 10^-12 per bit hour in modern RAMs at sea level. Data obtained from major computer installations and from biomedical devices had shown the soft error rates to be consistent with the intensity of secondary neutrons at sea level.

Future systems are likely to rely on faster and larger solid state memories using smaller devices operated at lower voltages. This trend is likely to be accompanied by the use of higher flight altitudes making SEE increasingly significant so that the influence of cosmic rays must be considered when assessing reliability and cost effective mitigating strategies adopted.

A collaborative research programme is under way to address these problems, the key elements being:

• development of models of the atmospheric radiation environment and its interaction with devices to enable accurate prediction of SEE rates;
• development of cost-effective ground-testing and screening techniques using irradiation facilities and laboratory techniques such as lasers;
• validation and improvement of models by comparison with flight data from environment monitors and SEE experiments at a wide range of latitudes and altitudes;
• investigation of cost-effective monitoring equipment for  SEE and air crew dosimetry;
• creation of a database of observed anomalies with follow-up analyses pursued wherever possible to ascertain if radiation is the cause.

15 The IONISING RADIATION METROLOGY FORUM (IRMF)

Vic Lewis briefly described the activities of the IRMF, which meets bi-annually and covers all fields of ionising radiation metrology. Membership is free and open to any representatives of organisations involved in the metrology of ionising radiation and radioactivity. Details, including the IRMF Catalogue of Calibration and Measurement Services available in the UK for Radiation Protection Intrumentation, could be found on the Website at http://www.npl.co.uk/irmf/. VEL, as Secretary of IRMF, could also supply details. The next meeting would be at NPL on 3 November.

One important activity is the organisation of comparisons of monitoring of ionising radiation and radioactivity. There was not sufficient time to describe that relating to neutron survey monitoring, but members were invited to the wash-up meeting for the Second IRMF Inter-comparison of Neutron Area Survey Monitors that was to be held directly after the present meeting.

16 CLOSE of MEETING

Vic Lewis said that it had been the best attended meeting yet. A wide range of aspects of neutron metrology had been presented and a number of diverse facilities had been described.

Members were asked to complete the questionnaire in their information pack, and return it to NPL. The members’ views were greatly valued and would help in planning the future work programme and the format of future Neutron Users' Club Meetings. Proceedings of the meeting would be distributed; these would include a list of all present along with telephone numbers and email addresses.

In closing the meeting, VEL thanked everyone on behalf of the Neutron Measurement Group for contributing. He hoped their visit to NPL had been both enjoyable and informative, and looked forward to seeing everyone again both as friends and customers. The next meeting is expected to be in approximately eighteen months time.

17 LABORATORY VISITS

Ten attendees visited the accelerator and irradiation facilities in the Chadwick Building (Building 47) after the meeting.

18 WASH-UP MEETING for IRMF COMPARISON OF NEUTRON MONITORING

Several members, including those from the participating laboratories (AEA Technology, Marconi Marine, NPL, NRPB) attended. DRaStaC had also participated but were not able to attend. A report of this meeting was drawn up by VEL and distributed to all participants in the exercise. The comparison had been successful and had shown excellent consistency in all calibrations.

A report will be given at the next IRMF meeting on 3 November. It is intended to publish an NPL report in the CIRM series by the end of the year.

P Kolkowski

29 October 1999