17 AUGUST
NEW ZEALAND GAZETTE
2275

95% of milk yield measurements to be within ±7.5%
99% of milk yield measurements to be within ±10%

The first requirement has been interpreted as ±7.5% where the true volume is 4.0 litres or more, and ±300 grams where the true volume is less than 4.0 litres.

When herd test determinations are carried out at every milking
95% of milk yield measurements to be within ±15%
99% of milk yield measurements to be within ±20%

The first requirement has been interpreted as ±15% where the true volume is 4.0 litres or more, and ±600 grams where the true volume is less than 4.0 litres.

The target accuracy shall be maintained under the following conditions:

(i) flow rates between 0.5 litres/min and 9.0 litres/min, that exist for at least two minutes.

(ii) air pipeline vacuum levels between 36 and 56 kPa.

(iii) air admission rates between 4 and 12 litres/min, with a milk flow rate of 3 litres/min and air pipeline vacuum level of 50 kPa.

(iv) the meter should perform to requirement on all types of installations when fitted in accordance with the manufacturer’s instructions.

Yield Range
The measuring range for total yield shall not be less than 0-20 litres.

Hygiene Requirements
Surfaces of the meter which come in to contact with milk and wash water/chemicals shall be of materials which conform with the requirements of MAFQAL hygiene specifications.

The materials used for the meter shall be capable of repeatedly sustaining a continuous temperature of 95 degrees C for a period of 3 minutes, without affecting accuracy during subsequent milkings or causing visible deterioration of materials.

2. Component Sampling Accuracy
   Target accuracy for subsample measurement shall be:

(i) within ±0.1% for milkfat and/or protein units of the reference sample for 95% of all measurements.

(ii) within ±0.2% milkfat and/or protein units of the reference sample for 99% of all measurements.

(iii) subsampler bias less than ±0.1% for milkfat and/or protein units compared with the reference sample.

(iv) standard deviation for subsampler errors less than 0.05% units.

3. Component Analysis Accuracy
   When herd test determinations are carried out less frequently than at every milking:

C_v* = Coefficient of variations (standard deviation divided by average value of samples tested).
Reference methods: Rose-Gottlieb. Kjeldahl calibrated analysis.

Somatic Cells

Measuring range ... up to 10,000,000 cells/ml

Repeatability (relative SD) ... 4% at 500,000 cells/ml
... 5% at 300,000 cells/ml
... 6% at 100,000 cells/ml

Accuracy (rel. mean diff. from direct microscope count MC) ... Typically better than 10% in the range 100,000 to 1,500,000 cells/ml

Correlation to DMC ... 0.96 - 0.99

When herd test determinations are carried out at every milking.

The average of all of the determinations carried out shall not be less accurate than those specified for the previous category.

LIST OF APPROVED EQUIPMENT

The following is a list of equivalent approved by the database manager (refer 2.5) as at 1 August 2000.

Milk Meters

• Tru Test

  • Models – *F.V.
  • *H.I.
  • *F.C. 5.6/ML/L
  • *F.C. 7.7/ML/L

• Carter Holt Harvey

  • Model – Waikato Mark 3

Sampling Devices

• Pipette
• Southey Probe
• MMB Dipper

Analysis Equipment

• Fat, protein, lactose and total solids

  • Foss Milkoscan
    • Models – *203
    • *300
    • *605
    • *133
    • *303
    • *4000

• Bentley

  • Models – 150
  • 200A
  • 200M

Somatic Cells

• Coulter Counter

• Fossomatic

  • Models – *450
  • *180
  • *215
  • *360
  • *90
  • *300
  • *5000

• Bentley

  • Models – Somacount 150
  • Somacount 300A
  • Somacount 300M
  • Somacount 500