Trials on nir perten 2014

NIR Analyses in fishmeal An Icelandic Network

2014-03-27

Stefan Lundgren Perten Instruments AB

Perten Instruments Group

AGENDA

• Briefly about Perten, really briefly...

• NIR Analyzers from Perten Instruments

• Parameters analyzed in Fishmeal by NIR

• Results from trial at Matis 2013

• NIR Networks

Our History Harald Perten, a cereal chemist, founded the company in 1962.

Analytical solutions for Agri Food Industries

Grain

Flour

Food

Feed

A long history in NIR

In 1980 Perten Instruments developed its first NIR instrument.

Global – Local

Icepack Ehf

Reykjavik

DA 7250 NIR Analyzer

• 6-second analysis

• Grains, slurries, powders and more

• No sample preparation

• SEE IT OUTSIDE!

DA 7300 Online NIR Analyzer

• Inline analysis

• Grains, pastes, powders and more

• Built-in camera

DA 7300 Online NIR Analyzer

• Over-belt analysis

• Grains, meals, powders and more

Plus Platform overview

NetPlus

server

NetPlus - Reports

Reports on e.g. production

Production manager

NetPlus - Remote

Instrument performance

and configuration

Instrument manager

NetPlus - Application

Organize calibrations,

find data, export/import

Application specialist

Results Plus

Sample analysis, with GUI

Process Plus

Process analysis,

No GUI

Results Plus?

Future additions

Instrument software Browser software

NIR in Fishmeal – Analysis parameters

Parameter Unit Reference method

Moisture % by weight Drying Oven

Protein % by weight, as is

Kjeldahl

Fat % by weight, as is Soxhlet

Ash % by weight, as is Ashing Oven

Salt % by weight, as is Titration

TVN

Cadaverine

mg/100g

g/kg

Steam distillation

HPLC

NIR in Fishmeal – Trial at Matis

• Some 200 samples were provided by SVN and Matis

• All samples were scanned on DA7250 • Analyzed for Moisture, Protein, Fat and Ash, using Perten’s global

calibration models

• Results were good but not optimal

• Matis provided samples with Salt and TVN values

• New calibration models were developed by Perten

• Feasibility study carried out on Cadaverine in Fishmeal

Moisture

Parameter SEC SEP R2 Range C-set V-set

Moisture 0.43 0.35 0.93 2.1-9.3 184 38

Protein

Parameter SEC SEP R2 Range C-set V-set

Protein 0.87 0.83 0.87 65.9-74.3 184 38

Fat

Parameter SEC SEP R2 Range C-set V-set

Fat 0.52 0.51 0.95 6.3-15.4 182 38

Ash

Parameter SEC SEP R2 Range C-set V-set

Ash 0.39 0.40 0.97 9.5-17.8 184 29

Salt

Parameter SEC SECV R2 Range C-set

Salt 0.31 0.55 0.97 2.2-8.4 53

TVN

Parameter SEC SEP R2 Range C-set V-set

TVN 1.9 1.9 0.92 0.7-30.0 123 24

Cadaverine

Parameter SEC SEP R2 Range C-set V-set

Cadaverine 109 218 0.96 100-1800 35 8

NIR Networks

•Basic Ideas

•Advantages

•An example

•Organisation

•Calibration Monitoring

• Standardisation

• Instrument Monitoring

•Costs

NIR Networks – Basic ideas

• Ensuring a uniform, fair and cost effective form of quality assurance and basis for payment

• A common accredited reference laboratory

• Rapid analyses based on international standard methods

• Easy-to-use analytical instruments with identical specifications

• Minimized operator influence on results

• ’Approved’ calibration models • All instruments are compared to a

’Master Instrument’

NIR Networks - Advantages

• Centrally controlled calibration models and reference analyses

• No local influence on results gives a uniform and fair classification of products

• Simplified quality assurance for the users • Calibration monitoring through the network using

web based software

• Daily check sample

• Powerful tool for implementing an Icelandic standard

• Harmonized quality control on produced fishmeal

NIR Networks - Advantages

• Building a common sample database • Better and more robust calibration models gives

reliable analysis results

• Few reference samples per network member -> Lower analysis costs!

• Yearly updates

• Reduced development costs for new analysis parameters and products, eg

• Cadaverine

• Raw materials

• Same calibration models used for both process control and out-loading

NIR Network - Organisation

Calibration Center

Perten Applications

Group

Administration Center

NetPlus Network

Manager

Networked

Local instruments

Reference Laboratory

Master instrument

NIR Network - Organisation

• Steering Committee • Representatives from users, reference laboratory and instrument

supplier

• Reference Laboratory • Matis Laboratories • Accreditations: ISO 17025, SWEDAC • Certification in acc. With ISO 9001 och ISO 14001 • Controls:

• Ring tests in Iceland and Europe • Third party revisons by customers

• Calibration Center • Perten Applications Group – Stockholm / Matis Reykjavik

• Administration Center • Matis Neskaupstadur

• Technical Support • Perten Technical Group – Stockholm & Icepack - Reykjavik

NIR Network – Calibration Monitoring

• The Master Instrument is compared to reference analyses at the accredited laboratory

• Samples are sent in by network members on a regular basis

• Some 100 control samples collected from all participating factories each year

• Results are available on the internet

• NetPlus Reports – Web based reports software

• Updated regularly with new samples

NIR Network - Standardisation

• The Master is validated and adjusted against reference analyses

• The Reference Laboratory sends out reference samples once a year for instrument control

• Typically 10 samples per product type

• Scanned on every instrument in the network

• Scans collected via internet by the Admin. Center

• Predicted results are compared to the Master

• Individual instruments are adjusted via internet if needed

NIR Network - Standardisation

• The Master is validated and adjusted against reference analyses

9.8

9.9

10

10.1

10.2

10.3

10.4

10.5

Reference lab DA7200 Master

NIR Network - Standardisation

• Individual instruments are adjusted against the Master via internet if needed

8,5

10,0

11,5

maste

r

slave

1

slave

2

slave

3

slave

4

slave

5

slave

6

slave

7

slave

8

slave

9

slave

10

NIR Network– Instrument monitoring

Continuous monitoring of instrument stability

• Instrument Self-Diagnostics

• Daily Check Samples

• Relevant sample material

• Synthetic sample (Polystyrene)

• Analyzed daily to confirm stability

• Statistics in Results Plus instrument software

NIR Network – Outlier handling

• ’Outliers’ during analysis

• Samples that deviate from samples in the calibration sample database

• Spectral difference or difference in content (range)

• Sent to Reference Lab for chemical analysis

• Added to calibration sample database for updating and improving calibration models

NetPlus Reports

• Information on all reference samples is available for all participants in the network via internet

NIR Network – Costs

Network price for the instrument

• DA7250 NIR Analyzer: € 48 000

Annual fee

• Xx xxx ISK per instrument

Includes

• Yearly standardisation against Master instrument

• Specified number of reference analyses per year

• Updated calibration models

• Annual maintenance service of instruments

• Remote support via internet

NIR Network – Summary

Participating in an NIR Network cooperation

• Simplified and cost-effective quality assurance

• Increased credibility for NIR analyses • Quality certified by accredited laboratory

• Reduced time spent and costs for upkeep of NIR analyses for product controls

• Lower costs for developing new NIR analyses

Questions?

Your partner in Agri Food Quality Control