Field of the invention
The present invention relates to the non destructive evaluation of pulp wood traits from standing
tree using handheld NIRS (HHNIRS) by collecting spectra from (a) leaves and (b) stem below
the bark of the standing tree. More particularly, the present invention relates to evaluation of
pulp wood traits from standing tree using handheld NIRS by collecting spectra from stem below
the bark of the standing tree drilled at suitable height with a core sampler.
Background and prior art
NIRS was used for measuring pulp wood traits and mechanical properties of wood using
different sampling protocols. Downes et al. (2011) reported NIRS method for measurement of
cellulose in eucalypt wood meal (wood powder) prepared from increment core samples. Bailleres
et. al. (2002) applied NIRS for the measurement of wood characteristics (lignin, extractive
content, lignin composition, surface longitudinal growth strain and shrinkage) in wood meal.
Poke and Raymond (2006) measured lignin and cellulose content in Eucalyptus globulus by
NIRS using solid wood. Jones et.al (2005) used wood strips for measurement of density,
microfibrillar angle and stiffness by NIRS.
A global calibration model for the prediction of Kraft pulp yield (KPY) in eucalypt species
across a wide geographical range has shown that it is possible to predict pulp yield in a variety of
hardwood species from diverse geographies using a single hybrid model (Downes et al., 2009).
Blakey and Rooyen (2011) used handheld NIR for measuring moisture content of avocados
across a wide range of moisture content (51.0% -85.0%). The technique was also used for
measuring fruit quality in various fruits such as mangoes (Subedi et al., 2007), plums (Perez-
Marin et al., 2010) and nectarines (Perez-Marin et al., 2011). Rodgers et al. (2010) developed a
method for the routine measurement of cotton fiber quality parameters, micronaire, maturity and
fineness, using handheld NIRS.
Much work is not reported on application of handheld NIRS for measuring pulp wood traits.

Meder et al. (2009) compared different sampling methods such as core samples, chain saw frass
and bark free stem surface for developing calibration model for KPY measurement. Three
calibration models yielded R2 values 0.69, 0.62 and 0.54 respectively. They observed a good
correlation (R2 0.93) between bench top model and handheld NIR values for KPY on wood meal
samples. Downes (2011) reported a calibration model for cellulose and KPY with 47 samples
using handheld NIR. The spectrum collection was done directly with wood meal samples as well
as keeping the wood meal in ZIPloc bags and found that both methods showed similar
calibration performance.
Onsite application of handheld NIR for routine evaluation of KPY and other pulp wood traits
was hampered due to non availability of proper sampling method (representative of the whole
tree) for NIR spectra collection which resulted in high prediction errors.
Zeuhi patented (CN1936537A, 2007-03-28) a method for measuring density of Huoli wood
using near infrared spectrum. In this patent the author used core samples to predict the density.
He Wenming and Xue Chongyun patented (CN102192890A, 2011-09-21) application of near
infrared analytical technique to determine the timber chemical composition such as nitric acid-
ethanol cellulose, poly pentaglucose, acid-soluble lignin, cold-water extraction, hot water
extraction, dimethylbenzene alcohol extraction and 1% NaOH extraction. Rapid MRS analysis
methods (CN102288569A 2011-12-21) for lignin content, cellulosic content, etc of a fibrous
biomass have been patented. A method for predicting dry mechanical properties of wet wood
as well as from standing trees using NIRS has been described (CA2 401 301 (13) Al 2002-01-
03 ) by Meglen and Kelley. Here the authors used fibre optic probe to predict the mechanical
properties of the wood. Predicting the behavior of wood during processing for end use such as
the susceptibility of wood to form internal checks during drying, or predicting for one or more
characteristics of the wood after processing such as the dimensional stability of the wood during
or after drying or in service, or the elastic modulus of the wood after drying using NIR
(collecting spectra from surface) has been patented (WO 01/01113 Al, 3 2001-04-01). Kelley
patented (WO 03/004994 A2 2003-01-16 ) a method for determining the mechanical
properties of decayed wood that has been exposed to wood decay microorganisms, comprising:
a) illuminating a surface of decayed wood that has been exposed to wood decay microorganisms
with wavelengths from visible and near infrared (VIS-NIR) spectra; b) analyzing the surface of

the decayed wood using a spectrometric method, the method generating a first spectral data of
wavelengths in VIS-NIR spectra region; and c) using a multivariate analysis to predict
mechanical properties of decayed wood by comparing the first spectral data with a calibration
model. The calibration model comprising a second spectrometric method of spectral data of
wavelengths in VIS-NIR spectra obtained from a reference decay wood, the second spectral data
being correlated with a known mechanical property analytical result obtained from the reference
decayed wood.
The methods reported for pulp wood traits require collecting core sample, powdering it, and then
measuring using bench top NIRS. Portable / handheld NIRS was used for powdered samples,
chain saw frass and wood surface. The prediction errors were high when using wood surface
from standing tree. Hence it could not be used for routine evaluation of pulp wood traits.
US6525319 B2 (2003-02-25) relates to the spectral analysis of wood, and in particular to a
method for predicting dry mechanical strength properties from the visible region of near infrared
(MR) spectra of green wood using a multivariate calibration model. It is an object of '319 to
provide a rapid, accurate method for predicting the mechanical properties of standing trees,
which is useful in assessing the value of a stand of timber, by quantitatively measuring the
quality of the timber via VIS-NIR spectra. The invention of '319 provides a method for
predicting the dry mechanical strength for a green wood comprising: illuminating a surface of the
wood to be predicted, the wood having a green moisture content; analyzing the wood surface
using VIS-NIR spectrometric method from a reduced range of wavelengths in the range of from
about 400 to about 1150 nm, the method generating a first spectral data; and using a multivariate
analysis to predict the dry mechanical strength by comparing the first spectral data with a
calibration model, the calibration model comprising a second spectrometric method of spectral
data of a reduced range of wavelengths in the VIS-NIR spectra obtained from a reference wood
having a green moisture content, the second spectral data correlated with a known mechanical
strength analytical result obtained from a reference wood when dried and having a dry moisture
content.
The present invention relates to illuminating the stem surface at a depth preferably 25% to 35%
of the stem diameter and drilling width is 6.0 to 8.0 mm with light of wavelengths from near
infrared region (NIK) having spectral range of 939 to 1797 nm using handheld NIRS and

collecting the NIR spectrum without the use of fiber optic probe. The patent of'319 may not be
user friendly for standing trees and for trees having smaller diameter. It further requires a fibre
optic probe for spectral collection.
US6031233 (2000-2-29) relates to devices for analyzing a material according to its optical
reflectance or transmission spectrum. The invention discloses a handheld device for infrared
reflectance measurements of samples to identify the sample material and comprising a self-
contained portable unit built into a handheld housing. The housing includes a window and optics
on a bench adjacent to the window so that the optics is aligned with the sample when the device
is placed directly against the sample. The invention thus relates to a portable, handheld AOTF
spectrometer which is lightweight and small enough to be carried to the filed for analyzing
samples.
The measurement unit used in the present invention is with improved technology and
measurement processing like MEMS technology, digital light projection, programmable micro
diffraction grating and the user friendly software.
Objects of the invention
It is an object of the present invention to overcome the drawbacks of the prior arts.
It is another object of the present invention to provide a method for evaluation of pulp wood
traits from standing tree using handheld NIRS.
It is yet another object of the present invention to provide an improved method for evaluation of
pulp wood traits from standing trees using handheld NIRS by collecting spectra from (a) leaves,
and (b) stem below the bark of the standing tree drilled at suitable height with a core sampler.

Summary of the invention
It is an aspect of the present invention to provide a process for predicting wood traits of standing
trees (from Eucalyptus species) comprising the steps of:
(i) Preparation of a sampling point by debarking a portion of the tree and drilling a hole
wherein the hole is made at a suitable place in the debarked window;
(ii) Illuminating the sampling point with near infra-red radiation'and spectra collection;
(iii) Determining the pulp wood trait of the standing tree by comparing the spectral data with
a calibration model of the relevant pulp wood trait.
Brief description of the accompanying figures
Figures 1 and 2 illustrate the debarked window with the hole.
Figure 3 illustrates spectral collection using Phazir handheld NIR from Polychromix in the field.
Figure 4 illustrates spectra collected from stem drilled using handheld NIRS.
Figure 5 illustrates processed NIR spectra (1st derivative mode) for KPY (drilled stem).
Figure 6 illustrates calibration model for Kraft pulp yield measurement using handheld NIR.
Figure 7 illustrates calibration model for acid insoluble lignin using handheld NIR.
Figure8 illustrates calibration model for S/G ratio using handheld NIR.
Figure 9 illustrates calibration model for KPY in E. urophylla using handheld NIR.
Figure 10 illustrates prediction error (%) to the measured value for KPY measurement in E.
camaldulensis (stem drilled).
Description of the invention
According to the invention it has been shown that it is possible to directly and continuously
predict wood traits of standing trees like Kraft pulp yield (KPY), lignin, S/G ratio (Syringyl /

Gauiacyl ratio) and cellulose content by collecting spectra by illuminating the stem below the
bark of the standing tree drilled at suitable height with a core sampler, wherein the drilling depth
is preferably 25% to 35% of the stem diameter and drilling width is 6.0 to 8.0 mm, with IR light
in near infrared region (NIR) having the spectral range of 939 to 1797 nm using handheld NIR.
The invention especially relates to the determination of wood traits such as Kraft pulp yield,
lignin and S/G ratio in Eucalyptus species, which comprises Eucalyptus camaldulensis (4 years
old trees) and E. urophylla (2-3 years old).
Onsite screening process for predicting wood traits of standing trees comprises of the following
process:
1. Tree on which measurement to be carried out is debarked in order to produce a debarked
window having breath of about 3.5 to 4.5 cm and length of about 5.5-6.5cm ( Fig 1).
Sampling point is made by drilling a hole wherein the hole is made at a suitable place in
the debarked window such that the distance between the drilled hole and upper end of
debarked window is preferably 1.5 -2.0 cm and distance between the drilled hole and
lower end of debarked window is preferably 4.5-5.5 cm ( Fig 2). Further, hole depth is
preferably 25% to 35% of the stem diameter and hole width is 6.0-8.0 mm (Fig: 2).
Moreover the hole prepared should be free from any core sample so as to maintain the
uniformity of the hole. Drilling is done with a core sampler. The drilled hole is marked on
four directions in order to ensure that NIR source is kept properly on to the drilled hole
and measurement consistency is maintained (Fig:2a).
2. Infra red spectra of the stem surface at 25-35% depth are collected by illuminating (Fig:
3) the IR radiation having wavelength 939 -1797 nm into the sampling point prepared as
mentioned above.
3. Developing a calibration model involves collecting the spectral data in the NIR region
(Fig: 4) (939 to 1797 nm) using handheld NIRS wherein the data is obtained from
reference set of standing trees using handheld NIRS. The collected spectra are correlated
with known analytical results from the said reference set of standing trees. The

coefficient of estimation (R2) of the spectral data with known values is checked and it is
improved by selecting suitable wavelength region (Fig: 5) and spectral processing
parameters (S.Golay, normalization etc). The calibration model accuracy and precision
are determined by conducting cross validation / external validation and checking
correlation coefficient, root mean square error of cross validation (RMECV ) and
external validation (Table: 2, Fig : 10).
Predicting the wood trait of the standing tree at the site by comparing the spectral data
using the calibration model of relevant trait with suitable accuracy and precision.

Table: 2 Calibration and validation model characteristics for KPY, lignin and S/G ratio in
Eucalyptus camaldulensis and E. urophylla

Methods used to generate data for calibration models:
KPY was estimated using bomb digester with 300gms of dried wood chips (wood to white liquor
ratio: 1:2.8). The digestion was carried out by increasing the temperature of the bomb from
ambient to 98°C in one hour, 98°C -165 °C in 50 min and maintaining at 165°C for 90 min. After
completing the digestion the pulp was washed, refined, dried, weighed and KPY content was
calculated at Kappa number range 19-22.

S/G ratio was estimated by pyrolysis GC/MS. Pyrolysis-GC/MS was performed using CDS
Pyroprobe (5150) and Agilent GC /MS (GC 7890A MS 5975C) instruments. Wood powder of
0.1-0.2 mg of 200 mesh sizes was placed in a quartz tube (25mm * 2mm) and pyrolyzed at 550°C
for 10 seconds. Pyrolysis chamber, transfer line and interface were maintained at 280°C
temperature. Pyrolysis products formed were transferred to the GC column (DB 1701, 60 m *
0.25 mm ID, 0.25 mm film thickness) by purging with helium gas. The following temperature
program was set for GC/MS: - injector at 280°C, column at 45°C for 4 min to 280°C at the rate of
4°C/min and held at 280°C for 15 min and interface at 280°C. NIST (National Institute of
Standards and Technology) and Wiley libraries were used for identifying the mass spectra of the
separated compounds. GC/MS analysis was done in selective ion mode (SIM) by choosing
characteristic mass fragments relevant to the compound. Klason lignin was estimated as per
Tappi test method: T 222
Calibration plots for KPY, lignin and S/G ratio in E. camaldulensis and KPY in E. urophylla are
presented in Figs 6-9. Statistical parameters for calibration model were calculated using
Polychromix method generator software. Peak processing parameters and calibration
characteristics of each calibration model are presented in Tables 1 &2. It is found that surface
sampling has lesser R2 value (0.87 / 0.83) (which gives a measure of accuracy) for KPY
compared to 0.91 for the proposed method. The R2 value of the developed model (0.91) is
higher than the value (0.62) reported by Meder et al (2009). In this paper the authors have
compared handheld NIR with bench top and portable models using three different sampling
techniques for the estimation of KPY in E. globulus and E. nitens and found that bench top
model shows good prediction compared to other two methods, where as current sampling method
showed satisfactory correlation between bench top and handheld NIR models. The maximum
prediction error +/- 2% was found between bench top and handheld NIR models. Out of 20
samples tested sixty percent of the samples showed error <1.0%, twenty percent <1.5% and
remaining 20% < 2.0% error. The data is presented in Table 3.

The method developed has been applied for measuring pulp yield in the field. Figure 10
demonstrates the prediction error encountered with the current calibration model. It takes just 5
minutes to carry out one measurement with HHNIRS including making bark window and
sampling point. Moreover the instrument is portable and can be carried easily anywhere. To
use bench top model we need to take a core sample (2-3 gm), dry and powder it prior to
measurement. The whole process takes minimum 3-4 hrs apart from other infrastructure like
oven, special powdering machine (to powder small quantities), vials and additional man power.
For HHNIRS a bark cutting tool (knife) and core sampler/ auger of suitable size are sufficient
for preparing the sampling point. The method helps in grading the trees in the field itself.
Conventional methods of pulp yield measurement requires 0.3 to 1kg sample ( depending on the
model) as well as infrastructure to carry out pulping and the whole process is time consuming
( about 8 hrs).
The protocol currently developed for onsite screening of pulpwood traits in eucalyptus species
covers E.camaldulensis and E. urophylla. It can also be extended to other species for pulp wood
traits/ mechanical properties with suitable calibration model.

WE CLAIM:
1. A non destructive process for predicting pulp wood traits of standing Eucalyptus species
trees comprising the steps of:
(i) Preparation of a sampling point by debarking a portion of the tree and drilling a
hole wherein the hole is made at a suitable place in the debarked window;
(ii) Illuminating the sampling point with infra-red radiation and collecting the
spectra;
(iii) Determining the wood trait of the standing tree by comparing the spectral data
with a calibration model.
2. The process as claimed in claim 1, wherein the drilling depth is 25% to 35%.
3. The process as claimed in claim 1, wherein the drilling width is 6.0 to 8.0mm.
4. The process as claimed in claim 1, wherein infra-red Spectra are of wavelength 939 -
1797 nm.

ABSTRACT

The present invention provides a process for predicting wood traits of standing Eucalyptus
species trees comprising the steps of:
(iv)Preparation of a sample point by debarking a portion of the tree and drilling a hole
wherein the hole is made at a suitable place in the debarked window;
(v) Illuminating the sample point with infra-red radiation;
(vi) Determining the wood trait of the standing tree by comparing the spectral data with a
calibration model.