This invention relates to the detection of impurities in metal agglomerates such as scrap metals and in particular to methods of screening scrap metals for unwanted impurities.

The recycling of metals has always been an important industry, and it is currently becoming even more important. Metals such as aluminium, copper, zinc, brass and steel are collected as scrap and are then reprocessed for re-use. One example is the electric arc steel making process which uses scrap almost exclusively for its feedstock.

Because of the wide variety of sources from which scrap may be collected its quality can vary widely and the scrap may be mixed with a number of non-metallic impurities or contaminants. Commonly found contaminants are ceramic materials. Such impurities contain naturally radioactive constituents to a greater extent than metals. Alternatively, there may be radioactive contaminant materials themselves in the scrap. To determine quickly the amount of such contamination in a lorry load of scrap when it is presented at a scrap yard is not an easy matter. One particular problem is the shielding effect of scrap loads themselves. There is, as is well known, a naturally

occurring background level of gamma radiation and therefore any detection . system must be set to ignore this. Unfortunately it is found that a load of scrap lowers the measured background level of gamma radiation, the effect being greater for greater loads. It will be realised that this reduction in the background level measured can allow some large loads to pass as uncontaminated when they are in fact contaminated, if the detection system is set to pass uncontaminated small loads. Any adjustment of the detection system to compensate, runs the risk of rejecting uncontaminated small loads. It is therefore an object of the present invention to provide a method of analysing the quality of scrap metal with respect to contamination quickly and easily and reliably.

According to the invention there is provided a method of detecting impurities in a metal agglomeration comprising measuring the gamma ray emission from the agglomeration, analysing such emission and from such analysis determining the radioactive contamination in the agglomeration. The radioactive contamination may be due to traces of natural radioactivity in the non-metallic content of the scrap, or may be due to radioactive contamination. The analysis may be by differentiating the gamma ray radiation by energy bands. The gamma radiation energy bands may be those of potassium 40 and the decay products of uranium and thorium.

The invention will now be described by way of example only with reference to the accompanying drawings in which:

Figures 1 and 2 are diagrammatic front and side views

of a lorry being analysed;

Figure 3 is a possible arrangement for the analysis of scrap on a conveyor belt; and

Figure 4 is diagram of gamma ray energy bands measured in the detection and analysis system.

As has been explained the analysis of scrap for non- metallic content is not easy and there is a requirement to be able to do this reliably, simply and quickly in scrap yard environments. The method adopted is to measure the gamma ray radiation from a particular load of scrap on a lorry or railway truck or from a particular position on a conveyor belt or indeed in any other situation in which scrap is being handled.

The method relies in part on the fact that in general the non-metallic contaminants do have a level of radioactive activity, albeit quite low in absolute terms. Typical levels are 0.3-1.0 Becquerels/g. The impurities commonly found are glass, concrete, dirt or slag and it is known that these contain traces of uranium, thorium and their decay products as well as potassium. Gamma radiation of the isotopes and energies shown in Table 1 will be found. These are all relatively high energy gamma rays. In general, metals contain extremely small amounts of radioactive isotopes and therefore have a significantly lower gamma ray activity than non-metallic materials of typically ten times lower. Any significant detection of radioactivity therefore indicates contamination of the metal. It is also known that metals in the dimensions and

quantities of interest with respect to the invention do not wholly absorb gamma radiation. Therefore either the radioactive contamination of the non-metallic impurities are those generating the preponderance of such gamma radiation in a load of contaminated scrap and the radiation being emitted, or the radioactivity is due to a specific radioactive contamination of the scrap, rather than the natural radioactivity of non-metals.

Turning now to Figures 1 and 2 which show respectively a side and front view of a road vehicle containing scrap diagrammatically positioned between a typical arrangement of radioactivity detectors. The vehicle 1 contains scrap 2 which may contain non-metallic impurities. The six detectors 10 to 15 are arranged in a matrix around the lorry and the output from these are analysed into energy bands and the totals in each energy band are summed. The detectors may for example be caesium iodide scintillation counters or semiconductor counters such as high purity germanium.

In Figure 3 the arrangement of detectors about a conveyor belt carrying scrap is shown. The belt 20 is carrying a supply of scrap, for example in bales 21 and the detectors 22 and 25 surround the belt. The detectors are connected in a similar way to those of Figures 1 and 2.

The information provided by the detectors in either case can be used in two particular ways. First, where there is a considerable amount of non-metallic contaminant or a heavy radioactive contamination in the scrap itself

the overall reading will be relatively high, above a predetermined threshold, and although a detailed analysis could be made by the method outlined later in all probability the high level of contaminants shown would mean the rejection of the scrap load or its down grading.

Where the level from non-metallic contaminants is less than the predetermined threshold the level of ionising radiation picked up will be significantly affected by the naturally occurring background radiation and therefore it is necessary to resort to some simple signal processing in order to estimate the level of contamination. As has already been mentioned, to add to analysis problems the actual measured background level will vary with the size of the scrap load. What we do is analyse the radiation received by energy bands. The energy bands in which radiation is expected from typical contaminants is shown in Table 1 and by computing the amounts of these isotopes on the basis of the radiation found at these specific energies a level of contaminants present can be calculated.

Turning to Figure 4, this shows in (a) the normal background spectrum that might be found for gamma ray radiation between 20KeV and 1.5MeV, the intensity for each band being shown by the height of the respective bar shown at (1).

In (b) a scrap load has been introduced. The levels of background radiation measured have fallen from the levels (as shown in (a)) illustrated by the dotted lines to new levels shown solid. However, the whole spectrum falls

in the same proportion. If radioactive contamination is present it will occur at specific energy levels and therefore will increase the intensity of some, but not all bands. This is shown in Figure 4(b) where the energy band 5 was originally at intensity 1 without the scrap load present, and would have fallen to the intensity 2 when the scrap load was introduced without any contamination. But this load is contaminated with a gamma ray emitting contaminant, whose characteristics energy falls in the band 5. The band 5 intensity rises to the level 3, which can be detected as being a level out of ratio with the other energy bands, the ratios being of course determined by the relative intensities with no scrap load present.

These measurements will in general give the amount of non-metallic material or specifically radioactive contaminant present in a load of scrap on a truck or at a particular point on a conveyor line, for example. In order to estimate the percentage of contamination the details of the bulk density, total weight and cross-sectional area of the scrap load itself may be needed. This might be conveniently obtained for example from weighbridge date and type of truck, and, for example, it would be possible to arrange for an operator to enter the make of a truck, for the weight automatically to be supplied from the weighbridge and for the measurements of the radioactivity to be made simultaneously. The density, volume and cross- sectional area of the scrap could all thus be determined automatically and an estimate of the percentage

contamination of non-metallic material in the scrap could be obtained.

TABLE 1

Element Isotope Energy, Mev

Potassium K-40 1.46

Actinium Ac-228 0.91

Bismuth Bi-214 1.12 and 1.76

Lead Pb-211 0.83

Thallium Tl-208 2.61