Technical Field The present invention relates to a method of selecting bulk wood units for mechanical pulping. As used herein, the term"bulk wood units"refers to logs or log segments or large planks of wood; the method of the present invention is designed predominantly for use in selecting or classifying unsawn logs, but it could also be used for selecting or classifying log segments or large sawn planks.

Background Art The mechanical pulping of logs to produce pulp for papermaking may be carried out by a number of different known techniques; broadly, these techniques involve grinding or disc refining the wood against an abrasive surface to produce a mixture of long fibres, shorter fibres and fines. Mechanical pulping processes provide a high fibre recovery rate (typically 85-95 percent of total fibre is recovered) but consume a great deal of electrical energy :- electrical energy consumption typically ranges between 1250 and 3000 kW hours per tonne of fibre, and can account for up to one-third of the total processing costs. It is well-established in the industry that some tree species pulp more readily than others, and hence require less energy to pulp. However, there is a considerable amount of intrinsic variation in wood properties even when the trees concerned all come from the same forest stand or forest compartment, i. e. a block of trees in the forest which are of similar age and which have received similar management throughout their life.

The traditional method of sorting trees at the point of harvest for the log are to categorize logs in grade according to their diameter, length, straightness, diameter eccentricity and visual defects; the logs are placed in categories which reflect log diameter, logs size and log grade. The basic assumption is that logs in each category are substantially identical. However, so far as energy consumption for mechanical pulping is concerned, logs sorted in the above described manner often prove to be far from identical, and may vary widely in their energy requirements. Attempts have been made to predict likely energy consumption of logs during mechanical pulping from a wood quality parameters such as wood density and wood age, but these attempts

have not yielded reliable results. One expert in mechanical pulping (Corson"Course Notes for University of Auckland Diploma in Pulp and Paper Technology"1997) stated "Pulp energy consumption is influenced by factors that effect the related parameters of freeness and sheet density development. However, a knowledge of basic wood density, tracheid length or subsequent mean pulp fibre length is insufficient to predict refining energy requirements." The ability to be able to grade a batch of logs according to their likely energy consumption during mechanical pulping could result in substantial savings in electrical energy, and hence in pulping costs.

It is known to use acoustic speed measurements to predict some qualities of logs :- U. S. Patent No. 6,026,689 discloses a system for predicting the modulus of elasticity of a log by generating a stress wave along the length of the log by striking in the log (e. g. with pneumatic hammer), picking up vibrational signals from a standing stress wave in the log, and using this information to calculate the speed of the stress wave in the log, and hence the predicted modulus of elasticity for that log.

PCT/NZ99/00137 discloses a method for predicting one or more characteristics of wood fibre or wood pulp by first determining the velocity of sound through the solid wood and then comparing the determined velocity to known experimental information on fibre characteristics for that wood.

However, there is no known relationship between the modulus of elasticity of a log and that log's energy consumption during mechanical pulping. Similarly, although it might perhaps appear likely that the fibre characteristics of a log would govern that log's energy consumption during mechanical pulping, in fact no useful relationship between these two factors has been established. Thus, neither of the above described methods can be used to predict a log's energy consumption during mechanical pulping.

Scope of the Invention An object of the present invention is the provision of a method whereby a batch of logs may be reliably and accurately graded according to their likely energy consumption

during mechanical pulping, without significantly increasing the overall processing costs.

The present invention provides a method of sorting a batch of bulk wood units comprising the steps of: 1) Establishing a reference scale for the timber group to be sorted by: a) selecting at random a plurality of sample units of bulk wood from the timber group; b) measuring the acoustic velocity through each of said sample units using a predetermined measuring technique; c) recording said acoustic velocities and grouping said velocities into two or more velocity bands; d) processing all or part of each of said sample units to pulp using a predetermined mechanical pulping process; e) measuring the energy used to pulp each said sample unit; f) measuring the freeness of the pulp from each said sample unit; g) for each of said velocity bands, graphing pulping energy measurements against the corresponding pulp freeness measurements to produce a reference scale ; 2) Measuring the acoustic velocity through each of said bulk wood units in turn, using said predetermined measuring technique; 3) Comparing said acoustic velocity measurements against the reference scale and said velocity bands to predict the mechanical pulping energy required to pulp each tested unit; and 4) Dividing the tested units into subgroups according to the predicted mechanical pulping energy.

Preferably, before said acoustic velocity bands are selected, the acoustic velocities from all of said sample units are graphed to show the distribution of acoustic velocity in

the total sample, to enable the velocity bands to be selected such that a predetermined proportion of bulk wood units fall within each of the selected velocity bands.

Preferably, each of the batch of bulk wood units would be of the same, or a similar, species and would have a broadly similar history, i. e. each of the bulk wood units would be of about the same age, and have been grown and managed under similar conditions.

Brief Description of the Drawings Fig. 1 is a flow chart illustrating the method of the present invention; Fig. 2 is a diagram showing in the method of taking an acoustical measurement from a log ; Fig. 3 is a distribution curve of the acoustic velocities in the sample units; Fig. 4 is a graph of freeness/energy consumption; and Fig. 5 is a graph of fines content/freeness.

Detailed Description of the Preferred Embodiment Fig. 1 illustrates the sequence of steps needed to utilise the present invention.

In step 1, each of the batch of bulk of wood units is of the same or similar species, and preferably also has a similar history, as defined above. Whilst it may be possible to treat all bulk wood units of the same tree species as forming part of a single batch, i. e. being sufficiently similar to be classified using only a single set of reference tests, it is probable that before bulk wood units can be treated as forming part of a single batch, they must have a similar history. Thus, it is envisaged that separate reference tests will be required for bulk wood units of the same or similar tree species but with a different history.

The extent to which separate reference tests are required will become apparent in the course of industrial use of the method of the present invention; as data is accumulated from large-scale use, it will become apparent to users of the method whether more or

fewer reference tests are required to meet particular conditions.

In step 2, sample bulk wood units are selected from the batch, to carry out the detailed testing need needed to establish a reference scale. Typically, 100-300 samples would be taken from a batch, assuming that the characteristics of the batch were completely unknown.

In step 3, the acoustic velocity of each sample unit is measured, using the standard method represented diagrammatically in Fig. 2. The equipment for, and techniques for measurement of, acoustic velocity through a bulk wood units are known, and therefore are not described in detail. One typical system is shown in Fig. 2, in which a bulk wood unit 20 is supported and is struck on one end 21 by a hammer 22. The acoustic wave generated in the bulk wood unit by the impact of the hammer 22 travels down the length of the bulk wood unit, is reflected from the far end 23, and travels back to the end 21 where it is detected by an acoustic sensor 24. The detected signal is analysed by signal analysis apparatus 25, which also computes the velocity of the sound. The velocity is calculated from the time taken for the sound wave to travel along the length of a log and back divided by a distance equal to twice the length of a log. The apparatus is controlled by controller 26.

There are a number of known types of apparatus available for measuring acoustic velocity, and the above described equipment may be varied in a number of ways :- for example, the velocity may be determined from a single reading taken at the opposite end of the bulk wood unit 21 to the hammer 22. Further, the hammer 22 may be replaced by any device capable of generating an acoustic wave in the bulk wood unit, e. g. a piezoelectric device or a wave from a sound generator.

In step 4, the readings of acoustic velocity obtained in step 3 are graphed to give a distribution curve of the type shown in Fig. 3. The distribution curve shows the acoustic velocity range in which any specified percentage of the test samples fall.

Thus, a study of the distribution curve enables the operator to select acoustic velocity bands which will include or exclude a specified percentage of the bulk wood units.

The actual figures selected for the acoustic velocity bands will depend upon the operator's requirements :- if the operator wishes to select for mechanical pulping only those bulk wood units which are optimum for this purpose, then only a single acoustic

velocity may be selected, as discussed hereinafter.

However, more typically the operator would wish to select acoustic velocity bands such that a majority of the bulk wood units would be selected for mechanical pulping, and only those units which were clearly unsuitable would be rejected. In this case, the limits of the acoustic velocity band into which bulk wood units to be selected would fall, would be set to include a large portion of the units, e. g. : 3.0 km per second in the example shown in Fig. 3.

In step 6, all or a predetermined proportion of each sample bulk wood unit is pulp, using any of the standard mechanical pulping techniques. The amount of energy required to pulp each sample is recorded.

In steps 6a and 7, the fines content and the freeness are measured for each pulp sample, using established techniques.

In step 8, the measurements of freeness and pulping energy for each of the samples are arranged in groups corresponding to the acoustic velocity band into which that sample falls. A reference graph is then prepared of freeness/pulping energy for each of the selected acoustic velocity bands; this result in a graph of the type shown in Fig.

4. In the typical results shown in Fig. 4, four velocity bands were selected :- Speed 1: velocity S 2.52 km/second.

Speed 2: velocity 2. 7 km per second but zu 2.8 km per second.

Speed 3: velocity 3. 1 km per second but < 3.2 km per second.

Speed 4: velocity 3. 4 km per second.

The graph enables the operator to select the velocity band in which the desired level of freeness can be achieved for minimum energy input.

As Fig. 4 shows, the four velocity bands selected experimentally in practice group reasonably well into two separate bands, since the results for speed 1 and speed 2 lie

close together, and the results for speed 3 and speed 4 also lie close together. It follows that for practical purposes, the results could be grouped into two acoustic bands:- The first having a velocity < 3 km per second; The second having a velocity > 3 km per second.

Of these first and second bands, the second gives the lowest energy input for optimum freeness, and it follows that in step 9, where the acoustic velocity through each unit of the batch of bulk wood units is measured, the operator should select for mechanical pulping only those bulk wood units having measured acoustic velocity falling into the second acoustic band, since these are the units which will pulp most economically by mechanical pulping means.

For some pulp applications, the fines content of the pulp may be important and in these cases a graph of the type shown in Fig. 5 may be prepared, showing fines content/freeness for each of the four initial acoustic velocity bands.

To prepare this graph, the fines content is measured for each of the sample bulk wood units, using standard measuring techniques. These measurements are then combined with the corresponding freeness measurements and the acoustic velocity measurements to prepare a graph of the type shown in Fig. 5 :- a plot of this type enables the optimum velocity band to be selected for the desired fines and freeness properties.