WT Interview with Junghyo Yoon, Ph.D. and Bruce Crawford, Research Laboratory of Electronics, MIT. The transciption below has been edited for clarity and length

WT: Welcome, and thanks for doing this. Can you tell us what’s different about your desalination technology from what exists on the market today?

Yoon: On the market, there are actually decent portable desalination units, based on reverse osmosis. This is a really good product that can produce some twenty-two liters per hour, and the specific energy consumption is really low. The problem with these reverse osmosis systems is that they need a high-pressure pump. The high-pressure pump is very heavy, so the weight of the unit is much larger than 20-30kg, really heavy to carry by one person.

The reverse osmosis product cannot be affordably initialized so twenty liters per hour is the minimum production rate in the current state of the art. This is too high sometimes, so even if they achieve a really low specific energy consumption, your power consumption will be large because it produces a lot of water, sometimes it is too much water for certain applications. Generally, it runs at 200-400 watts, it cannot be operated by off-grid systems, they do need a power grid.

The other thing about ours is, in order to change the production rate, we just change the number of the stack, this production rate is directly related to the weight of the device and the power consumption of the device. With this technology, we can achieve a weight of less than 10kg with less than 100-watt power consumption that can be operated by a really cheap solar panel, that you can buy from a website.

With reverse osmosis, if there is a molecule that is slightly larger than the water molecules, it will be stopped on the reverse osmosis filter membrane. Based on our customer interviews, they do need to change their filter every three months. As I commented at first, it is a high-pressure process, so you need to have a basic understanding of mechanical engineering. If you don’t appropriately change the filter, the water will be leaking, making a severe problem in the desalination process.

With our technology, we don’t need to change the filter, to remove the suspended solids. We have not operated our system for more than one year, but our process doesn’t require any filter change, although I am sure there will be some maintenance fees with our unit.

WT: My understanding of your technology is that if I run saltwater through a plastic conduit, there is a membrane on the top of the pipe, and a membrane underneath that pipe, and you charge one membrane positive and one membrane negative, and my understanding is that one membrane will attract salt and solids, and the other membrane takes out a stream of clean water?

You have used Artificial Intelligence (AI) in some way to maximize or optimize the charging of both membranes that you couldn’t do without AI? Is that close to accurate? How do the membranes work and how does AI come into this?

Yoon: Our (technology) uses the negatively charged membrane called the cation exchange membrane. These two membranes will be stacked with one water channel, under the electric field. So you will apply your saltwater from east to west and then you apply the electric field from the north to south, so all the alternative e charging, all the junk will be moved to the north direction; and then, the water will pass through the south direction, you can continuously supply your water from the east to the west. That is our basic technology.

Crawford: If I could modify the way you explained it, you are on the right track with the channel, membrane at the top, and membrane at the bottom. The salt and bacteria and everything is all sucked toward the top, leaving clean, pure water at the bottom. The drinking water stream is pulled out of the bottom and all the concentrated stuff at the top is sent out in a brine stream.

WT: How does AI come into this?

Yoon: That is our basic principle. It is a three-stage process, but in order to determine that the three-stage process was the really beneficial way, we needed to perform one-stage, two-stage, three-stage, four-stage, and five-stage tests in the lab-- This is a really time-consuming process, so we run the order, determine the stage process with different conditions, this gives us the basic database in the desalination process.

We put that database in the AI, and AI just learns which number of stages would show the best average consumption, and then they provide us with three stages with a certain number for the best energy efficiency.

Crawford: It’s not so much that AI is used in the device as it is operating, but it was used in the design to find the optimal process.

WT: So there is no AI in the system, that’s more for bench testing, is that correct?

Yoon: That is right, there is no AI in the system.

WT: When you separate, and now have clean water on the bottom, you are going to do this in a suitcase-type thing, which is a lot lighter, and it uses a lot less power and is that because of the way that these membranes are set up, on the top and the bottom? Or is that because they are charging at much lower electricity use rates, which saves power in the system, is that right?

Crawford: I’m going to give you the way a non-engineer thinks about it. Current devices in the market right now run on a technology called reverse osmosis – where basically you are jamming water through a membrane at upwards of 800 psi. To do that, you need really high-pressure pumps to get that 800 psi. So the pump is heavy, the plumbing is heavy, and the membranes often clog, as Jung said. The other problem with the pumps besides their sheer size, is they take a lot of electricity. If you have heard that desalination is energy-intensive, that’s why.

With this technology, you’ve got our conduit and the electrical field sucking out the bad stuff. We are not shoving water through a membrane at any point. There are membranes, but water is not passing through them. We can do this all at low pressure, so no high-pressure pump, no high-pressure plumbing.

That means, now we have eliminated the pain of the clogging membrane, we have eliminated the heavyweight of the high-pressure components, and we have eliminated the need for all that power because we can do it at low pressure. So, what you get is a device that can be much smaller, much lighter. It’s more reliable because you don’t have membrane clogging. It’s easier to use because you don’t have to deal with balancing the pressure in the pump, so our setup time is way fast.

The other benefit is you don’t need much electricity at all, we can run on a small solar panel, and you don’t need a generator.

WT: People in the Canadian water space, are always looking for a better, stronger, cheaper solution, like any investor. Are you two gentlemen close to looking for investors to get it off the bench and into the market? Where are you at in the commercialization space?

Crawford: We are close. The prototype is basically functioning how we want it to. It needs to be scaled up just a little bit to be useful for a family, or a group of ten people. We think we will be there by the end of the year.

In terms of funding, we are fortunate to be at MIT, where there is a lot of attention and grants given to scientific breakthroughs. We are not looking for dilutive venture funding at the moment. Right now we are thinking we can get from (readiness level) zero to one with grants from the National Science Foundation, with MIT start-up support programs. We are just participating in one next week, to hopefully win $100,000.

We think we can get to the point where we are selling our small portable units without taking a lot of investment. But then to go from (level) one to two, and take this technology, and say we are going to scale up and disrupt big desalination, or we are going to apply the separation breakthrough to some other segment, like hydrogen production, or drug manufacturing or oil/water separation, something totally different, that is where we will be looking for investment to take that next step.

WT: So by the end of this year, people will be able to buy a simple unit that will desalinate water in a sane presentation, I mean without the huge pumps and not a membrane crusher? How much do you expect the cost will be?

Crawford: By the end of this year, some select beta testers will be given the device, by the end of the next year, we hope we will be through a couple of iterations of beta testing and will be selling. In terms of cost, for portable desal units right now, the most portable, and most affordable right now is Rainman, and there are several others similar. You are going to pay at least four thousand dollars to get that portable set-up or five thousand dollars for the one with the portable generator. We are hesitant to release pricing at this point, but we are hoping we can go to market at a third of that.

WT: You are talking a global market, into Saudi, UAE, all this sort of thing? If I have a bar on the beach in South America, I can buy one of your systems and my clients can have a clean glass of water, is that how you see it?

Crawford: The thing with a technology like this is there are so many places we can go, and we want to go everywhere, including both those places you mentioned. First, we are thinking that we’ll start with sailors because they are the primary users of portable desalination right now. They will be really good to validate our prototype because they use it all the time.

Then we want to move to emergency preparedness, starting close to home, here in the United States where willingness to pay is a little higher, and in Canada. There is a growing need to be prepared for natural disasters, if you live near the coast, or in dry areas. People don’t realize it, but in tropical storms, hurricanes or where there is flooding, people lose access to clean water.

Those are our first two markets because they are reachable, and we think they are strategic because people can afford the device and they will use it a lot. Once we can get some validation that the market likes the product, that the product works, and some cost economies, then we will try to go more places, go global.

WT: When and how did this idea occur to you, charging two membranes? As a water filtration expert, you have sand filters, you have...

Yoon: Actually, for my Ph.D., I used a protein concentration for a biomedical application. If we can concentrate something, we can purify some water. Fortunately, after I got my Ph.D. my group got a postdoc to solve the brine problem in the states, I joined his group for brine treatment at first. While I am doing this project, my PI always had an interest in portable desalination. At first, I did not agree with his idea because did not think it was feasible.

Usually, when we make these units smaller, energy consumption always increases. So, based on my first experience with desalination, energy efficiency is not feasible for practical uses. but I ran my experiment over and over again, and suddenly the energy number was reaching really reasonable rates.

WT: I know that Canada is buying a bunch of frigates, and we might even receive them soon. Would you consider shipping our north coastguard guys one to test?

Crawford: Absolutely! We are looking for beta testers in extreme environments. If it can work for your Coast Guard in the far north, then that’s it.