Energy Consumption of an Automatic Milking System

Automatic milking systems are appearing all over the UK and researchers at Teagasc have been looking into the energy requirement issues that these new systems may bring.
calendar icon 9 April 2013
clock icon 4 minute read

J. Upton1, G. O’Riordan 2, B. O’Brien1 and S. Fitzgerald1
1 Teagasc, Animal & Grassland Research and Innovation Centre, Moorepark, Fermoy, Co. Cork., 2Dept. of Mechanical Engineering, Cork Institute of Technology, Cork


Recent data shows that there are approximately 10,000 commercial farms worldwide using automatic milking systems (AMS) to milk their cows. This figure is expected to grow rapidly in the coming years (De Koning, 2011). Therefore, the energy consumption of AMSs will become increasingly important. Studies by Bijl et al, (2007), and Artmann & Bohlsen, (2000), showed that electricity costs were greater with AMSs compared to conventional milking systems. However these studies did not give detailed component breakdown information. The aim of this study was to document the electricity consumption of a recently installed AMS at Moorepark for the period 1st May to 31st October, 2011.

Materials and Methods

A herd of 63 spring calving cows were milked from pasture using a Merlin AMS*. Data presented here pertains to the period from 1st May to 31st of October. Cows were milked on average 1.8 times per day. Average milk yield over the period was 17 litres per cow per day.

The AMS was washed with hot water 3 times per day during weeks 1-6, and twice daily, thereafter. Water heating was provided by electrical water heaters. Milk was pre-cooled using a Packo tubular cooler (model TT2) supplied with well water and subsequently, by a 5,000 litre ice bank tank.

The vacuum pump was a vane pump with a 3kW motor. Vacuum level was controlled by a standard regulator from weeks 1-6, while a variable speed drive (VSD) controlled pump with 1.1kW motor was used thereafter. Compressed air was supplied by a 2.2kW compressor.

The energy consumption of the AMS was monitored using Sinergy Escot energy monitoring equipment and software. The Escot data-logger can measure power consumption of multiple electrical circuits using clip-on AC current transducers. The logging software records cumulative kilowatt-hour (kWh) readings every 15 minutes.

Measurement equipment was calibrated and accurate to ±1 per cent of reading. This equipment allowed for measurement of the following individual components in the dairy: milking robot, vacuum pump, air compressor, milk cooling and water heating.

Results and Discussion

Water heating accounted for 37.1Wh/l (Watt hour per litre) of milk produced. The milk cooling system, air compressor, vacuum pump and robot consumed 14.8Wh/l, 14.7Wh/l, 11Wh/l and 2.9Wh/l, respectively. Miscellaneous items such as lighting and an office consumed 16.4Wh/l. The average total electricity consumed per litre of milk produced between 1st May and 31st October was 97.4Wh/l.

When the relevant tariffs are applied the average cost of electricity was 1.25 Euro cent per litre of milk (c/l). The cost of electricity varied from 0.82 to 1.87 c/l as the volume of milk harvested by the AMS varied over the season from the spring-calved herd (Figure 1). Water heating was the largest consumer of electricity in this study at 38 per cent of the total.

This requirement is a consistent fixed cost irrespective of the volume of milk produced because wash cycle scheduling is time based. These average costs of 1.25 c/l were high compared to an audit of conventional milking systems (0.43c/l) on 21 commercial dairy farms (May-October 2010) (Upton et al,2011).

This may be due to (a) reduced milk output from the AMS due to it being the start-up year, when milk yield is expected to be reduced by 10-15 per cent (Wade et al, 2004) and/ or (b) under utilization of the AMS. 63 cows were milked in this grazing based study, whereas the possibility to extend capacity to 112 is considered achievable (Jago et al, 2006). The change to a VSD vacuum pump in week 6 reduced vacuum pump running costs by 55 per cent (from 0.28 to 0.13 c/l).

Figure 1. Seasonal variation in electricity costs per litre of milk harvested (c/l) with an AMS system


Average electricity costs of the AMS tested with 63 cows was 1.25 c/l. This is likely to be reduced as cows become familiar with the AMS and optimum herd size/milk output for the AMS is reached. Suitability of a heat recovery system and cold detergent wash cycles will be investigated in 2012 to moderate running costs further.


Artmann, R. & Bohlsen, E., 2000. Robotic Milking, Proc. Int. symposium Lelystad, The Netherlands, 17- 19 August 2000, Wageningen Pres, p.221-231.

Bijl, R. and Kooistra, S.R., 2007. J. Dairy Sci., 90:239-248.

De Koning, K., (2011). Encyclopedia of Dairy Sciences. W. F. John. San Diego, Academic Press: 952-958 Jago J.G., Davis K.L., Newman M., & Woolford M.W., (2006). Proc. New Zealand Society of Animal Production, 66: 263.

Upton, J., Murphy, M. & French, P.(2011) Proc. Teagasc National Dairy Conference: 101-106.

Wade, K.M., van Asseldonk, M.A.P.M., Berensten, P.B.M., Ouweltjes, W. & Hogeveen, H. (2004).pp 62-67. The Netherlands, Wageningen Academic Publishers.

Further Reading

You can view the full report by clicking here.

April 2013

© 2000 - 2023 - Global Ag Media. All Rights Reserved | No part of this site may be reproduced without permission.