Personal Context: About ten years ago, I was part of a private equity team that managed and exited a water & wastewater treatment company. I spent a lot of time with the management, and lot more time on the company’s strategy and financials, creating the pitch for sale, and prospectus for an IPO. All the work with and on the company helped refine my understanding of global water availability, the infrastructure and its pain points. The aforementioned company was an engineering company, providing a suite of treatment options for treating fresh water, sea water and waste water. Desalination being one of them. It is also the most capital intensive one. Exploring it in some depth here.
Water, Water everywhere
2.5% of the world’s water is freshwater (rivers, groundwater aquifers, lakes). Rest 97.5% is sea water, full of salts and minerals. (Yes, Water, Water everywhere/ nor any drop to drink). In ideal situations, there should be enough fresh water freely available for human needs. But we do not live in an ideal world. The world has a lot of people. And we use a lot of water. And at times, water may not be easily available in the areas it is most needed in.
Where does it all go?
Most of the water usage is for agriculture (around 70%). Followed by industries (20%) , mainly for generating power and then some for making stuff (that pair of Jeans).
And then come our daily needs as human beings which are supplied by municipal pipelines (around 10%). As an aside, it is an interesting triangle worth exploring – water, food and power.
Journey to the tap
Water is treated before sending it through the urban pipelines as potable water. Since safe, even if not free drinking water is a fundamental requirement for any civilization, in most countries, it is government responsibility (local or state).
Depending on the input source feed water (fresh, sea, brackish or waste), and the output usage needed (potable, industry), the treatment can be designed. This is what most water engineering companies do. And then hand out the different parts of the project to different construction and equipment companies. Of course, each treatment option comes at a different level of cost.
One of the ways is desalination
Desalination is expensive. If one (here, it means the government) had the option of either providing water from freshwater with minimum needed treatment, or providing water extracted from salt water in a complex process of time, energy and capital to remove the salts to get potable water, one would always choose freshwater.
But we are not talking about choice any more. Over the last few decades, we have moved beyond it. The freshwater supplies in the world are limited. And some of them are under stress. Also, water is a local resource and a very local problem requiring custom solutions. Surface water gets replenished regularly (if there were no such things as decade long droughts). But ground aquifers not as quickly. Hence, desalination comes in picture.
Desalination through heat
Desalination is the process of removing salts from water to make it potable. There are many ways of achieving this. The simplest method uses thermal energy to segregate the water vapour from the salts. The idea is to heat up the water and to collect the vapour. This is the idea. But to do it efficiently for large quantities of water, several ways have evolved over time. One of them is Multi Stage Flash (where steam is created and collected in flashes, sort of). Another is Multiple Effect Distillation. Both these processes use some of the energy they generate by heating the water, and work best along with tandem power plants where the desalination can use the spare heat of power plants and the water helping to cool the plant. Hence making the economics symbiotic and efficient for both.
These are the old ways of doing this. And most of the older desal plants use this method. It is interesting to note that the largest and oldest desalination plants are located in the Middle East, where fresh water has been scarce and where fossil fuels have been easily and cheaply available. Countries like Saudi Arabia rich in oil but not so in fresh water have been using the modern Thermal Desalination methods since the 1960s. And they continue to build more to secure the water supply to its citizens. Earlier this year, the country announced 9 more desalination plants along the Red Sea.
Desalination through Membranes
The other desalination methods, Reverse Osmosis and Electrical Dialysis are membrane-based methods. Salty water is sent through membranes where either reverse osmosis helps the water through membranes leaving the salts behind or electrical dialysis helps in ionising and pushing the salts out, leaving pure water behind.
To the extent possible, I’ll avoid borrowing external images/text, but in this particular case of osmosis, the following image captures the point very well. The following image is from Hong Kong Water Supply Department.
I’ll not get into the science which I have a limited understanding of. However, osmosis is a natural process, taking place in all living cells, where water molecules pass from low concentrate solution to high concentrate solution through the semi-permeable cell membranes. The Reverse Osmosis method using this natural property, puts pressure (which needs energy) to overcome the osmotic pressure of water, and helps in reversing the flow through the semipermeable membranes. Membranes are a selective barrier, not a filter, but a sort of partition. Membranes are expensive, and they get used over time, need replacement and hence these processes tend to be expensive.
I am unable to get hold of cumulative data, but over the last 15 years, most new, big desalination plants across the world are RO based plant – these being all those in Australia, Israel, Spain, Saudi Arabia, rest of Middle East, the largest plant in US in San Diego, new plants in India.
Reverse Osmosis, as a process, is less than 50 years in the making. It is a combination of improved engineering, the materials and the pressing need to secure more water supply that has made them viable over the last two decades. Similar to computers, the tech has improved over time. Not as starkly though. There are better membranes, new filtration technologies, Nano materials. But once a plant is commissioned, it is a big commitment to the method it is made for.
The Reverse Osmosis technology is where most of the recent Desalination money has been put to work. And where most of the future money is headed.
So, what is the scale of what we are talking about?
Given the general issues with climate change, droughts, and increasing urban populations, over the last few decades, the number of desalination plants around the world has increased. There are some 18,000 desalination plants in the world now. (a count of plants with capacities of over 20 Mega Litre Per Day). To better understand the metrics, consider the following:
The 1 cubic metre daily for an average household is a rounded up assumption.
So, we are talking about plants with capacity of 20 Mega Litres Per Day (MLPD). Or shall we say, with capacity to meet daily requirements of around 20,000 households (?). So, there are around 18,000 such desalination plants in the world and growing.
A lot of these plants are new (commissioned over the last 5-10 years). Especially the big ones. And they are mainly located in Middle East, US, Europe, and now Australia.
Around 44% of the global desalination capacity is located in Middle East and North Africa. Saudi Arabia has over 10,000 MLPD Desalination capacity (around 14% of global desalination installed capacity). US has around 9000 MLPD (or 11%), followed by UAE, Israel, Spain. Relative to their size, the other gulf countries are where most desalination plants located. Understandably given the aridity of the region and the limited fresh water.
Let’s consider for a moment what it means for these countries. For Saudi Arabia, with a population of around 32 million people, around 50% of the water needs are met through desalinated water. Contrast this fact to that only 1% of the global water supply is desalinated water. Israel, with a population of 8.5 million people, has desalinated water as 60% of its total water supply. And with plans to build more desal plants. Similarly, Singapore, with a population of around 5.6 million people, meets 25% of its water needs through desalination, and more plants to come up by 2020.
These are rich countries with very low fresh water supplies and can afford desalination of large scale and as main source for their water. Other countries are currently using these as back up. But if we look to the future, more and more urban centres will need this.
With the evolution of RO economics over time, many governments around the world are making desalination an urgent infrastructure agenda item to be discussed to secure water supply. It is interesting to note that over 50% of the world’s urban population is surrounded by sea water. In the arid regions of Gulf, US, Australia etc, this proportion is around 75%. And as water stress rises, and desalination becomes a viable option, more governments begin to consider it.
Let’s consider Australia.
From what I gather from several internet sources, Australia currently seems to have 11 desalination plants which have more than 100 MLPD capacity (many small ones). And 2 more big ones planned. As newness goes in the water infrastructure world, these are very new plants. Sydney Desalination plant (ready to operate in 2010) has 250 MLPD capacity. Melbourne (was ready in 2012) has 410 MLPD capacity. The one in Adelaide has 270 MLPD capacity (ready in 2012). Sydney’s plant can meet 15% of Sydney’s water demand, and the one in Victoria, around one-third of Melbourne’s needs.
Australia is an interesting case. Most of the desalination plants were built 2009 onwards, as a reaction to the Millennium Drought. For around 10 years the country suffered drought, with low water levels in dams. Hence different state governments got these Desalination plants commissioned. Most of these are now in hibernation or stand-by mode. Or as the Sydney desal plant website defines itself, “Insurance”. There is enough fresh water in dams.
Although not being used, they do come at a huge cost, and the operation and maintenance can be expensive (like repairs after a tornado). Understandably it draws a lot of ire from the water bill paying public.
Given the recent stories from drought in South Africa, one would think such capacities are needed. These plants were built to fight off droughts. They are water security of sorts. It just so happens that there have been not much of a drought once they have been tested and operated. It is a long-term call taken at a time of distress which will be paid over years. But by already investing and installing the capacity, what is lost is the option of investing in it in future in better, cheaper technology. And that is the question – are they too big? Are they too soon? Were they really needed? It stays the moot point.
However, while trying to understand all this, the most interesting bit is that to be recommissioned, or to wake the Sydney desalination plant up, from the time it is decided to do so, it will take around 8 months for it to begin operating!
Now, to understand the dollar numbers
It is worth considering a few desalination plants around the world to get a hang of the numbers and expenses:
The largest desalination plant in the world is Ras Al Khair, Saudi Arabia. With capacity of 1,037 MLPD, it is a hybrid plant (that is, it used both MSF/thermal and RO methods). It began operating in 2015. It cost USD 7.2 billion. It is operated by the Saline Water Conversion Corporation of Saudi Arabia. (Founded in 1974, this body operates all the Saudi desal plants, making it the world’s largest desalination provider). A point to note here is that thermal plants may be more expensive to build than membrane based. But thermal ones can go on for much longer without major refurbishment.
The largest RO plant in the world is Sorek, Israel. It has a capacity of 627 MLPD. It cost $500 million to be built. It began operating in 2013. The Sorek plant was designed and built by IDE + Hutchison water (IDE eventually bought out Hutchison’s stake in the plant).
If you look at the costs of some of the other desalination plants in the world, the Sorek number seems quite low. This is very heartening. By comparison, Victoria, Australia desal plant was eight times more expensive and has two-thirds of the Sorek plant capacity. Or on a per MLPD basis, the Sorek plant is one-tenth of the cost of the Victorian plant. Normally the total costs go down as capacity increases. On the costs and how they are not directly comparable plant to plant, just in a moment.
The largest operating RO desalination plant in Europe is located in Spain, Barcelona. It has a capacity of 200 MLPD. It started operations in 2009. There are larger desalination plants in Spain, but this one is the still operating one. I am unable to find proper cost totals for this one. One of the numbers is Euro 230 million or USD 300 million.
The largest RO based desalination plant in the United States recently started operating at Carlsbad in Califorina (2015). It has a capacity of 190 MLPD. Cost to build USD 1 billion.
One of the largest desalination plants in India built by IDE with the company I refer to above in my personal context is located at Nemmeli, Chennai. It has capacity of 100 MLPD (full capacity in 2013). It was built at around USD 100 million cost.
As to the Australian desalination plants, the Sydney plant was made by Veolia and John Holland, cost of around ~A$1.8 billion for 245MLPD capacity. It is managed by Veolia. Victoria (Degremont, Thiess/Suez, Macquarie) cost ~A$3.5 billion for 410 MLPD capacity. Sydney, Queensland and Victoria plants are quite expensive as well. Around $7-9 million CAPEX per MLPD capacity.
The true cost of desalination
If you think of the true cost of desalination, there are two key components to it. One is the capital cost of building the plant. And second, is the operation cost over the life of the plant.
As the above listed plants show, the setting up costs are not directly comparable for several reasons. There are several variables that go in the total cost – the kind of water to be treated (salinity levels) which inform the design of the plant, the source of water (from how far the water needs to be brought in), the cost of the coastal land where the plant needs to located, the design and engineering costs (which are primarily salary costs of the engineering team and will vary significantly from which country they are located in), the method of desalination, the amount of pre-treatment required (most RO based plants require considerable pre-treatment of water before putting it in through the RO plant), the construction cost, the mechanism of discharge of the brine thus created, and finally, the cost of remineralisation, storing and feeding the water thus generated into the municipal pipelines. Not to forget, the kind of capital used and the cost of that capital (Ownership and capital is a separate big game by itself).
Each of these variables can be quite different for each desalination plant. There are economies of scale that come in with higher capacity builds, and the technology has evolved over time. Which might imply that the overall life costs of plants built later should be lower than older plants.
The second component is the cost of operating those plants, and the maintenance needed over time. For example, for the Sydney plant there is a $1.8 million per day fee payable. The maintenance contracts may be for 20-30 years.
The operating cost is mainly power, consumables (like membranes and chemicals), cleaning downtime, personnel & admin cost and maintenance of the plant. Over that, there is the life of the plant itself. Some of the operating cost is related to the water itself. The source/feed water could be sea water or brackish water (saltier than freshwater with 0.5 -3% salinity. This is a very big range. Sea water has around 3.5% salinity). Cost wise, it is cheaper to treat brackish water. However, there is limited brackish water available. Mostly in dry regions, and it is already being used. If a plant uses brackish water, it will perhaps use less pre-treatment chemicals and time than sea water. Then the big cost of energy to put water through those high-pressure processes.
On the operating cost, there can be a big difference in thermal and RO based plants. The thermal/ old tech plants are more robust. They require less or no pre-treatment of water before sending through the process. And they don’t need high level of maintenance (cleaning, membrane replacing every 3-5 years). And the plants seem to have a much longer life. What’s more, they are normally co-located with power plants which brings down some of the power cost (which thermal plants need a lot of – almost five times than in RO processes).
The overall calculation depends upon the demand and supply of water in a region. Then the ease of power being available to carry out the processes. And then, who owns it, how is the project funded adds artificial costs to the physical process of actually desalting the water. If the projects are owned by government, then great. Else, the equity cost also gets built in the cost of each glass of desal water coming out of a plant.
Here’s all this on an image. The cost of desalination is the combination of all the following costs.
So, the same glass of desalinated water may be five times as expensive if it’s made in Australia vs if it were made in Israel.
The good thing is that there are new breakthroughs in this space, and the overall trend points towards reducing costs. The RO process is becoming more efficient. The energy consumption is going down (almost one-fourth of 1980s). The recovery is improving. And the life and quality of membranes are improving. The question to be considered is how much of the new developments can be applied to the existing plants?
The ‘so what?’ and a few other concluding notes
In a water stressed world, desalination is here to stay. It is expected that the world’s capacity will be doubled by 2030. There are so many numbers. At one place, I read $10 billion of investment is earmarked over the next five years for 5700 MLPD desalination capacity. The problem is these numbers can be very different depending on what parts of costs they are including.
Before I started writing this note, I thought that may be desalination is the place where new tech (as in new way of desalination) can help bring down the cost significantly. But there are so many problems with that premise. The volume of water that passes through means investment in infrastructure is as big a number as investment in the process (which is what tech would improve). Hence, by committing to a tech and building a plant around it (like in Sydney), might mean that you give up the option of building a new desal plant 10 years down the line with a tech which does not exist today unless the economics is compelling. Water is not a precious commodity in that sense. People wouldn’t change the infrastructure just because a marginally better technology comes on the market.
Of course, new ways of desalinating will be welcome. It is still only a few developed countries where most of these plants are located. It is still only 1% of the world’s supply. New cheaper method might mean more investment by the developing and water stressed countries.
And in terms of existing desalination plants, there are so many other fields that contribute to ‘efficient’. Breakthroughs or improvements in the membranes (making them cheaper or making them last longer), the pre-treatment methods and chemicals, power efficiency, may be power generation in a cheaper way, can all help in bringing down the cost of desalinated water.
However, if one is looking at water security, then perhaps one should look towards agriculture. If water saved is water created, then given that the maximum consumption is in agriculture (70%), shouldn’t the new efforts be focused on making it more efficient? Any improvement there would free up water equivalent to several expensive desalination plants.
Listing some of the questions yet to be explored:
- How much costlier is desal water compared to fresh water? I know this would be region specific, but one memo on Atmospheric Water (?) I read, showed it as twice (desal to freshwater treatment). I thought it would be much more.
- Would it make sense to consider cost of recycled water (waste water treatment) with desalinated water?
- Would waste water be a feed for desalinated plants? Seems like no as of the moment. It’ll need so much more pre-treatment.
- The companies in this space.
- The ownership structure and implication on costs. Who owns the infrastructure? The plants? The tech? And hence who benefits as the market grows?
- What is China doing for its water needs?
- The triangle of food, water and energy.
- Graphene as the new breakthrough in membranes? And graphene itself as an explore-worthy topic. Similarly. more clarity on Nano Materials. Is graphene a nano material?
- Case study of two actual plants. To get a sense of overall plant cost proportions and maintenance. Say, Sorek and one of the Australian or US plants?
- What is the life of a desalination plant? Seems like RO based are 15-20 years and thermal ones can go on for 30 years or more without major refurbishment
- To get a better sense of the trend of at-home RO. It is interesting to note that in countries like India, and the US as well, people who can afford to, will not trust the tap water. And families are installing these RO-based Desal units at their own homes. Rich countries do it at source. Rich people in poor countries do it at their home.
Closing Note: The objective of the note is to get a perspective on desalination and its economics in global context. I have referred several web-links and data points for this note. The data referred is quite spread out, and I have tried to consider and compare several sources to arrive at the above. Still, the numbers may not be exact, but they are close enough or approximated to get the broad picture. Happy to be corrected if you spot any glaring errors of fact or thinking. And keen to hear more on the open questions.