Fonterra spends a lot of energy and money in New Zealand to concentrate milk through ultra-filtration

We recently received an inquiry from Glen McGillivray who has been working on building a business case around the use of forward osmosis to replace expensive ultra-filtration processes in New Zealand’s dairy industry. Glen is missing some key FO membrane performance parameters to complete the business case and has therefore asked ForwardOsmosisTech’s community for help.

Have a look through Glen’s initial thoughts below and feel free to share your feedback in the comment fields.

Reducing logistics and milk concentration costs are important innovation drivers for Fonterra – New Zealand’s largest co-op

There is an incentive to Fonterra: who have to haul raw milk from the farm to the plant for processing, to try to reduce the volume of milk being hauled allowing for less trucks to perform the same task, leading to a variety of competitors trying to sell ultra-filtration to farmers, which is generally looked down upon, having high operating costs in a market driven by trying to keep costs down, as Fonterra suffers from New Zealand being a long distance from any market they can sell to, as the nation is too small to consume it’s own product.

Milk processing companies like Fonterra have to deal with extreme levels of variability in milk volumes with the main spring flush can be double the later autumn production in volume: low running cost high material costs are not attractive as such facilities will end up nonfunctional outside of peak flow.

At the other end: assets that will be used every day on the farm are usually considered attractive, assuming there is a financial incentive to help offset the cost of setup/operation.

Given Fonterra already pays to halve the volume of milk with Ultrafiltration prior to Pasteurization, I can’t see why they wouldn’t pass any savings from prior concentration on to their farmer/owners. Which would probably be all we would need to encourage forward osmosis installations.

The forward osmosis value proposition in a nutshell

Utilizing the principles of osmosis it is possible to create an osmosis gradient in the score of multiple atmospheres of pressure using materials as commonly found on farm as Magnesium Chloride.

With that sort of math behind it: and the required materials to build a concentration system being a few water tanks and a boiler system on farm: the entire idea would hint at offering lower operating costs AND potentially lower setup costs.

According to theory it all looks good as a application. Milk gets stored on farm for pickup usually once a day to once every two: offering plenty of time for a relatively small setup to extract water which really only adds bulk to the product, which could be recovered from our salt feed-stock by any boiling method you choose: it would remain more energy efficient than ultra-filtration using large vacuum pump. My personal idea is to use Solar evaporators with s more standard boiler as your cool weather/cloudy day backup.

Concentrating the milk on farm reduces volume meaning less trucks on the road thus less fuel and less maintenance, less ultra-filtration (the only means fast enough to extract water from millions of liters a day without massive setups), Less bulk for drying the final product reducing energy costs yet again, and the means to avoid the whole Permeate discussion entirely (returning the permeate from the ultra-filtration step is re-added to standardize milk, my personal opinion is local regulations make milk taste like it is from spring all year round: a bit watery.

Farmers have a lot of assets and relatively low rate of return, any process fitted can cost a bit up front, but operating costs need to be kept small, otherwise they effect their rate of return, and may just need to be turned off during periods of economic uncertainty. The potential of forward osmosis to offer a lower energy thus lower cost alternative at the cost of a capital outlay is more attractive than more inexpensive ultra-filtration setups that consume power ravenously trying to drag water  out of the milk

What are the potential challenges of using FO for milk concentration?

Given what I COULD find, I’ve determined that A: there are milk safe products, and B: Flow rates should enable reasonable outputs. I could not find much information regarding Cleaning/Maintenance nor lifespan. The usual cleaning products are strong acids and bases (admittedly well diluted) and so long as the material could withstand this standard practice and avoid needing swapping out too often all would be well.

Potential issues include the recovery of this water to afford all those savings (which again, would need to be passed back to farmers to make use of this potential) happens on site, and there are only small volume charges to producers at present, more may need to be done to encourage on farm concentration. Plus the removal of water mean that to reach proper standardized milk for consumer consumption would necessitate re-adding water to restore the proper character. Would consumers understand/appreciate milk processors adding water back in? Is there a market for double concentration milk (just add your own water)?

Completing the business case requires key FO membrane parameters, which are not readily available

Unfortunately two limiting features of the idea need to be addressed: first, just what are the values of those membranes which are needed for flow rate calculations. And secondly: anything coming into contact with milk for food production MUST be tolerant of cleaning products/processes for ensuring food safety, is there even a membrane technology that might be compatible with those needs?

I find it frustrating that basic functions like permeability which helps determine flow rates are concealed: particularly when sizing of potential uses can determine the viability of the application. All my basic math was needing were ballpark numbers, how many 10^x do I need to use a number that makes sense?

I know the volume of liquid to remove, just how much removal media to I need using the concentration of salts I am considering? This lack of information is not very helpful.

It is easy to flaunt values like bulk flow rates from their testing, but the number that would give guidance is left out. Makes running ballpark figures remarkably difficult..

Closing remarks

Please let me know if any companies show interest in trying to deal with the numbers, I can suggest various scales to consider, it should be noted that farms vary from small to VERY large so assumptions really should be spread across a number of scales to check how reasonable they might be in practice, a setup fine for 1000 cattle may not be viable for a farm with less than 200.

Thank you for your time.

Glen McGillivray


  1. Milk concentration is one of the more difficult tasks for membrane based systems to accomplish RO tried for many years in the early 1990’s to do this but suffered from severe fouling by the milk fat products. Several FO companies have tried this, including us, Trevi Systems. Our goal was to try and concentrate whole milk to about 55% milkfat, no matter the seasonal variation of the initial milkfat. Draw solution selection compatible with milk is straightforward and the regeneration step trivial. The big problem we encounter is the severe fouling of the membrane and therefore the cleaning regimen that must be developed to prevent this. Currently we are experimenting with different cleaners and run time intervals, and expect to have that sorted out by Q2, 2019. We are then ready to provide 500m3/day FO systems to dairy farmers. Unfortunately, the cost of such systems given the low volume of membrane production, makes it difficult to compete with UF based solutions.

  2. Thanks for sharing John. In your view, why should dairy farmers then opt for FO systems if the price (both OPEX & CAPEX?) is higher?

  3. It depends on the product that they want. A product that can be made with UF or RO will be cheaper with those established technologies which have already traveled down the cost reduction curve. However, a fresh milk concentrate is worth the current cost because it can be transported at reduced costs to overseas markets and water can be added later to produce a product that tastes like and has the nutritional value of fresh milk. This is not possible with evaporators which create a concentrate like condensed milk.


Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.