Membrane housings in a Reverse Osmosis facility.  From Baker’s Membrane Technology and Applications, 2nd Edition

On Friday, March 14, Korean industrial concern LG Chem announced they were purchasing NanoH2O, an RO membrane maker that took its first external funding at the peak of the CleanTech bubble in 2005 for $200 MM.  Industrial technology has proven to have a much longer adoption cycle than these initial investors expected – looking at the dynamics of the current NanoH2O business can help us understand how the initial investment assumptions in 2005 led to the current state of affairs in 20104. As with any business, we must start with the company’s product to understand how their customers work with them (from the company website as of 3/16/2014):

NanoH2O, Inc. develops, manufactures and markets (1) reverse osmosis (RO) membranes that lower the cost of desalination. Based on breakthrough (2) nanostructured materials and industry-proven polymer technology, NanoH2O’s thin-film nanocomposite (TFN) QuantumFlux membranes improve desalination energy efficiency and productivity.

0.  The customer buys a membrane cartridge

The customer purchases a membrane cartridge from NanoH2O that they load into a housing in a multi-million (and often several billion) dollar reverse osmosis facility, from which they produce usable (either drinkable or industrial) water from saline water.  The customer pays per cartridge, but there may be service components, discounts based on large volume purchases, and long term supply commitments.  The customer’s goal is to get a cubic meter of water as cheaply as possible, and NanoH2O’s solution, in the form of that membrane cartridge, is part of their calculus in driving that cost as low as possible.

Membrane cartridge like those made by NanoH2O, and what would be loaded into the housings shown above.

1.  Reverse osmosis (“RO”) and desalination (aka “Desal”)

There are three methods used in liquid filtration – RO is a special case of nano-filtration and is in general the most sophisticated and specialized.  The simplest is particle filtration – think of a metal sieve where we want to pull out pieces of gravel and then let water pass through.  Microfiltration is the same process, but at smaller scale.  The US FDA has clear guidance on how microfiltration can be used for the sterilization of medicines and therapeutics. In most micro-filtration activities, all of the liquid passes through the filter, and the filter after it is clogged is then scrapped/recycled.  This is very different than the process used in RO, where only some of the fluid passes through the membrane, and the fluid that does not pass through becomes more concentrated – in desalination, the more concentrated fluid has even more salt after than it did before, while a portion of the water flows through the membrane and has its salt removed.

Types of Liquid Filtration

Ultrafiltration is more sophisticated than microfiltration – the particulate we are looking to remove is much smaller, and in many circumstances may be considered ‘dissolved’ into the liquid.  Here the particles we are removing are 0.1 um and smaller.  In most filtration activities we look for pressure to improve the system’s performance – higher pressure across the filter media / membrane allows the process to occur more quickly, but it does so at the cost of more energy.  As we get into ultrafiltration, many processes require higher pressure in order to function at all. With those higher pressures, we encounter higher capex costs in our system.  The mechanical components of an ultrafiltration system start to increase – this is important in the case of NanoH2O, as it means the relative value of an improved membrane start to decrease as the CapEx of the system in which the membrane will function start to increase. Nano-filtration, of which RO is a special case, is the removal of small molecules from a fluid stream – some processes are separating nitrogen from oxygen – there are many, many flavors of commercially interesting nano-filtration processes and a bewildering array of technical approaches to achieve the needed economics.  RO is unique in that it is widely known by the public and the market for such materials has been growing since its initial invention.  For RO to work, water is put under pressure and pushed against a membrane – a small portion of that water passes through the membrane and is thus ‘purified’, while the remaining water, and all of the salt, is kept on the other side of the membrane and becomes concentrated. A water molecule is only 3 angstroms in size – and the salts we are pulling out are even smaller.  To do this best requires high pressure – again making the entire system cost much higher than the cost of the membranes across which the ions are separated.

2.  Nanostructured materials and industry-proven polymer technology

NanoH2O makes a membrane cartridge – the heart of their technology is the membrane that goes into that cartridge.  Industry-proven polymer technology simply states that the company isn’t attempting to introduce radically different chemistry into the membrane.  This makes sense.  If you’re spending $500 MM on a new facility, your desire to take a risk with a never-before-proven polymer is low.  The customer wants to improve performance, but they don’t want to make radical leaps to achieve that performance. When NanoH2O says, “Nanostructured materials,” they are doping the polymers used in their membranes with some type of structure that improves its performance once it is in the RO production facility.  This can be seen in their patent literature (example here is WO 2009129354 A2), which often calls out specific formulations of different nanoparticulate or additives to enhance performance.  This gets to be a blurry realm, when combining ‘nanotechnology’ (herein meaning nanoparticulate) and polymer chemistry – when the two components are often close to the same size.

3.  Thin-film nanocomposite

This is an industry standard term for the type of membranes that are used in RO.  The membrane itself is a very thin, precisely made material.  It is often made on top of a conventional nonwoven or even on top of a lower grade microporous membrane.  While this term may sound advanced, it is commonly used in the industry.

All RO membranes are thin film composites – many of the newer water or desal start-up technologies try to get around the use of thin film membranes using other methods.  This may itself turn out to be the big debate about new water systems – are they able to move beyond thin film, or should they continue to invest in 10%, 20% or higher improvements.  This is similar to challenges in the semiconductor industry where challenges at both sides of the size continuum – fine scale lithography and larger 450 mm wafers, force the industry to make big, long term decisions on how they deploy capital.

Closing

The product is a membrane cartridge that an end customer buys in order to get clean water at a cheaper price per cubic meter than their other options.  NanoH2O doesn’t sell the water.  They don’t sell the RO system.  They sell a component that feeds into that industrial supply chain in order to help the customer get what they really want.  The performance delivered by NanoH2O helps deliver either more water, lower total capex, or cheaper water, but it is essential to understand that in an industrial supply chain NanoH2O doesn’t hold its fate in its hands – demand for its product must be fostered at multiple points in the supply chain.

All of these factors together make for a complex adoption process that takes time – time that initial investors in this space didn’t fully appreciate.