International Colloquia on Thermal Innovations 20

Show video

I want to welcome everyone. Thank you for coming to InnoTherm. Today we have our 20th Colloquia with a talk by Professor Srinivas Garimella on Sorption Technologies for Space conditioning, Thermal storage and Carbon capture. I'm Professor Asegun Henry. I will moderate the discussion today.

We will start in just a minute or so and have Professor Garimella give a talk for about 45 minutes and then we'll follow that with about 15 minutes or so of questions from the audience. If you have questions, please put them in the chat or in the-- I'm sorry, in the Q&A box. We have some students taking a look at them and we will pull from the questions in the Q&A and I will ask some to Professor Garimella once he finishes his presentation. I wanted to remind you that we are recording and this will-- if your questions are here we'll get you the answers, will get recorded.

We'll post this on YouTube later, and we usually also post this on the website as well so have access to it later. The next webinar coming up is March 17. Generally been doing these bi-weekly. Next one is on Advances in Additive Manufacturing for Heat Transfer Applications. We'll have a series of panelists and moderation by Saniya LeBlanc. And with that, I will pass the mic to Srinivas and let me say a few words about Srinivas.

Maybe I'll go back to my slide on Srinivas first and say a few words. So Srinivas is the High Tower Chair in Engineering and Professor at the Woodruff school of Mechanical Engineering at Georgia Tech. He's there, he's the director of the Sustainable Thermal Systems Lab. And prior to joining Georgia Tech, he did his master's in PhD at the Ohio State University and bachelor's at the IIT. He was then a research scientist at Battelle Memorial Institute. Sr. Engineer at GM.

Then he was a research specialist in the Mechanical Engineering Department at Ohio State and then joined Western Michigan University, where he was on the faculty in the mechanical and aeronautical engineering department. Then he was a Associate Professor of Mechanical Engineering at Iowa State before then moving to Georgia Tech. He's published more than 250 papers, including a textbook on heat transfer and a textbook on condensation-- I'm sorry, textbook on heat transfer and fluid flow and mini and microchannels, in addition to a textbook on condensation heat transfer. He's ASME Fellow recipient of numerous awards, including NSF Career Award and the ASHRAE New Investigator Award. And his research involves applications such as residential and commercial space conditioning, automotive propulsion and climate control, chemical process industries, and other energy intensive applications.

His specific research focuses on fundamental investigations of phase change heat and mass transfer, supercritical fluid flow, heat transfer phenomenon and single and multi component fluids and micro and many scales towards the development of novel thermally activated absorption and vapor compression heat pumps, natural refrigerant, space conditioning systems, thermal management systems for high density lithium, ion batteries, and a number of other applications. So Srinivas has done a lot. We're very excited and glad to have him here. I'm happy to hear your talk and I'm really interested to hear about what you've been working on. So Srinivas if you want to go ahead and share your slides and go ahead and take it away. OK.

I've been getting to the slideshow. OK, can you see this OK? We can. We can see it fine.

Thank you. Thank you for the opportunity to present in this colloquium series. I've been attending some of them and it's been very, very educational and I really enjoyed listening to some of these talks, and I'm happy to talk about what we have been doing here.

So today my focus is going to be on sorption technologies for a variety of applications and we'll start getting into those as the presentation proceeds. So we'll kind of look at this today's talk as a journey through sorption. I'll first touch upon the basics, the thermodynamics of sorption. And then while thermodynamics describes the idea limits and so on, I'll then focus on the realistic constraints and then how to make those things happen through an understanding of absorption hydrodynamics heat and mass transfer. And then show you ways in which we've been miniaturizing them.

How we've been putting the system together and also focusing on how to fabricate them in an affordable manner. And then we'll switch gears and talk about other applications, such as carbon capture and gas separation using sorption. Also, water purification while doing space conditioning, and drying of textiles, and seasonal thermal storage and finally, the cold chain for developing countries. So sorption has a place in all of these applications and we give you a brief overview of those.

So first of all, what is sorption? Basically, it makes heat or species move in directions where it really doesn't want to do that. So sorption makes it possible, for example, to make low grade heat or low grade energy move uphill. So heat usually wants to go from high temperatures to low temperatures, but sorption makes it go the other way.

The basic principles that are used is there's a swing in the pressure or the temperature. If you change the pressure or the temperature of fluid or something like that, it has a different saturation temperature if you change the pressure and vise versa. And then you throw in concentration. We're usually talking about two species, so if you change the fraction of one species over the other, that gives you another degree of freedom instead of just the pressure and temperature that is available to compression systems. So you can look at sorption in the space conditioning applications as a thermal compressor. So some elementary thermodynamics.

Let's say you have 125 degree heat source and then you are at a 35 degrees ambient, which is fairly common in places like Atlanta. The ideal work output that you can get out of it has an efficiency of 0.226, but that's the absolute theoretical limit. It assumes that there are constant heat sources and sinks. There's isothermal heat transfer, infinite heat exchange and no reversibility. So it's just a dream.

Real efficiency is going to be much, much lower because Carnot didn't leave us a whole bunch of money to actually make this happen. Unfortunately, thermodynamics dreams and decrees what it could be and heat transfer is burdened with the responsibility of actually delivering that potential and it ends up being far less than this. When you look at heat pumping, the measure is a coefficient of performance. And again, if you're going to have the temperature of 35, for example, for air conditioning and you want to cool the refrigerant to say five degrees so that you can cool the room, then the coefficient of performance can be quite high. So for every unit of energy coming in, you can produce 9 units of cooling.

However, that looks quite rosy, but that was assuming that there was electrical energy or work available. But if you want to do this from heat, then you have to multiply that previous efficiency with this coefficient of performance that then puts it down to about two. And this, also, is an ideal number.

The reality is much, much lower than that. So for one thing, one takeaway from this is if you happen to have a heat source, it's better to use cooling than work because work was only 20% efficient even ideally and then when you throw in the COP, you're able to do better. So that's one takeaway from this due to the heat pumping benefit. So if I want to look at thermally driven heat pumps, you can conceptualize it as a heat engine that is driving at the pump. So here's the heat engine with the current efficiency, and then here's the heat pumping part of it. And this produces work and then this needs work to make this happen.

It has the COP as shown here and then if you connect these if you use the work that the heat engine provides as the motive force for your heat pumping, then that's the thermally driven cooling that you are achieving. So the ideal performance can be defined by those thermodynamic formulas. This is for the engine heat pump or this is for the engine, and this is for the cooling part of it. But all of these are ideal considerations when you're trying to heat or cool something. Usually, you're trying to heat water or cool water, or worse yet you're heating air or cooling air.

So ultimately, it's not just the refrigerant that you're trying to provide at a certain temperature, but when you start coupling it to the liquid or the air, then liquid is usually much more beneficial because it has a much higher specific heat. It has a much higher thermal capacity. So when you heat or cool a liquid, it doesn't change temperature too much, so it sort of approximates those infinite reservoirs and so on a little bit better, whereas when you start heating or cooling air, its temperature changes so fast that then it runs out of potential to do anything more. Also, with liquids, you have the ability to have tight pinches on the temperatures that are available. So between two fluids, you can bring them much closer and extract more heat using liquid than with air.

And then when you use liquid versus air, you are able to use microscale heat transfer. You can use counterflow heat exchangers and so on. So whenever possible liquid is a better alternative.

However, that's not usually available to use. Quite often if you want to be comfortable in your room, you have to heat or cool the air. So keeping this in mind, let's consider a case where you have 120 degree heat source and you have a 35 degree ambient, for example. This is the same example I presented earlier. You might think you have an 85 degree delta T to play with. Well, not really, because some of it is consumed in that heating or cooling of the two fluids that are doing the job and then there is some delta T needed for the heat transfer.

So you're left with far less, you're left with only about 54 degrees of density and so you can see how reality has encroached on the ideal and you are getting less for that. Now if you have an even colder heat source of say 60 and then you're doing this indoors and not outdoors, maybe electronics cooling or something like that. Then, again, with the liquids the penalties for the delta T are smaller. And then with this 40 degree delta T, you're left with say 32.

So that's not as big a penalty. So these realistic considerations need to be kept in mind when you're designing actual systems that do this option. So one common example is we all get air conditioning or heat pumping using a mechanical/electrical compressor. It compresses the vapor to high temperature and then it condenses by using heat to the ambient, then it expands across the valve and then goes to the evaporator to provide you the cooling. But this requires electricity, and also many of these refrigerants that are used in these systems are being banned because of the global warming potential of these fluids. So what a sorption does is, it replaces that compressor with a bunch of heat exchangers, a generator, an absorber, and a recuperative heat exchanger.

In this way, you can use heat as the driving force instead of electricity. But then, you can see that it proliferates into more heat exchangers this is because, whereas the compression system that's in your house, you already have electricity provided to you. That electricity was done at great cost somewhere else with a big power plant.

Here, you're trying to actually put a power plant and an air conditioning system together in a small package, and so you need more heat exchangers and this sometimes reduces the viability of a system like this. So our efforts have been to try and make this still possible so that you can use a versatile sources of heat like solar, biomass, natural gas, and so on. So this is sort of the framework for the rest of the top. Now because there are all these realistic benefits people have tried to come up with more sort of internal recuperation and so on to reduce that external input and therefore make it more efficient. So this is the single with the basic cycle, then people have come up with things known as double effects cycles or dual cycles, and even generators absorber heat exchange cycles. Basically, what you're trying to do is once you generate the present in one place, you're trying to also use it again so that you keep extracting a little bit more and your efficiencies will increase.

But you can see that with this additional internal recuperation, you get higher cycle efficiencies, but then you increase the number of components. You make it more expensive. You make it more difficult to control and so on. So there's a fine line, there's a balance that has to be done between these criteria.

So when you are trying to develop sorption systems, it's a good idea to sort of do an overview comparison of what are my heat sources? What am I actually trying to achieve? And then fit the appropriate cycle to do that job, and I presented this sort of thermometer approach where these are my various heat sources, what can I do with them? Well, you can use it as a heat transformer to actually upgrade the heat to a higher temperature so that with medium temperature you can produce high temperatures. You can take things like exhaust air from trucks and so on, and do absorption cooling or organic ranking vapor compression cycles and so on to get either a work output or a cooling output and so on. So if you do this sort of survey of what you can do with your heat sources, then you can start building the systems. And so if you want to be able to do that, you have to understand the heat and mass transfer processes and the fluid flow that's happening inside. So I'm going to spend a few minutes talking about that.

So we're changing gears. And one of the most common geometries for absorption systems is a bank of tubes. You see three tubes here and fluid flows over them. In early days, people understood this as just a form of liquid flowing over a bank or tubes, but didn't consider the intricacies of how that happened. So you can see that there is a liquid droplet formed. It elongates.

It breaks up due to instabilities. And then what it impacts the next tube, there's sort of a splash-like thing and the boundary layer gets mixed up. And then when it falls on to the surface of the tube, there are waves that are spreading sideways. And then some of these droplets, they don't merger it because of surface tension effects. So you have to model all of these fluid phenomena to be able to understand how the heat and mass transfer happens in these systems.

So for that example, for example, we looked at how the droplet actually propagates and then we did some analysis of the progression of these droplets to find out how much surface area there is for heat and mass transfer. And if you take that same fluid and break it up into multiple droplets and so on, you get more surface area for you in a volume. So if you track that, then the surface area for volume increases substantially and this is a benefit, and this within the same surface area helps you do more.

We then did computational analysis of these. This is on the left side. You see the experiments on the right side. You see our computational work without taking any information from the experiments and just providing the initial conditions. And so you can see that we've been able to replicate that quite nicely all the phenomena of droplet formation, extension into a filament, and then it breaks up into these satellite droplets and so on and also the propagation on the surface of the tube has been captured.

So this provides us now some design tools for these kinds of systems. We can also do this for evaporation, so those were around tubes. And if you want to use microchannel tubes, for example, these are flat rectangular tubes with microchannels in them. And we studied how the fluid flows over them to achieve more compactness.

So you can see four different conditions how the fluid wet the surface and flows, and the way the mass that is created. And then other practical considerations that people sometimes ignore are how do you distribute this fluid over these tubes because if all the fluid just falls in one location, you're not doing any heat or mass transfer. So we had to design some good distributors, a wide variety of configurations to get to some ideal geometries. And then we did heat and mass transfer analysis, also.

There's coolant flowing through this tube, and then the lithium bromide water solution that flows all over that-- and then as long as it is in contact with the tubes, it cools down the blue corresponds to it cooling down. But then when it starts departing from the tube, then it starts heating up because it's still picking up vapor, which heats it up, but it has a means to cool down so then it falls on the next one. So this is a full 3D transient analysis of that process to understand absorption in these geometries.

With this understanding, you can design better geometries. This is a terrible design, and we get these insights because this isn't providing those nice droplets. This isn't a thin film so there's a lot of heat and mass transfer resistance that is generated if you use a geometry like this. So our insights guide us into what is a good design for these things.

So we use these insights several years ago to develop a very compact residential absorption heat pump, and we're just about 12 centimeters by 12 centimeters by about 50 centimeters big heat exchanger, with these microchannels crisscrossing and providing a lot of surface area per unit volume, the coolant flowing through those microchannels. And then the liquid flowing down on them vapor coming up, you do the absorption, you take away the heat. This provides a lot of compactness because of the high heat and mass transfer coefficient on the outside.

The high heat transfer coefficient inside the tubes and the large surface area to volume ratio. While this shows an absorber, you could also use this exact same geometry, which helps with manufacturing to do this option. Instead of sending cooling fluid through these tubes, you can send heating fluid through them and then it boils off your refrigerant from that solution to provide you the cooling. So with just a simple change in the drippers that you use on the top, you can do this. So after that, we had achieved quite a bit of compactness, but we took on the challenge of making it extremely small.

So one of my PhD students thesis was to develop a full absorption system, not just one heat exchanger, but a full absorption system using microchannels that are smaller than the textbook that we use for the heat transfer class. And in fact, he built it in this functioned and provided us 300 watts of cooling using just waste heat. The way this is done is very thin sheets-- 1/2 millimeter sheets with etched passages in them carries the fluid. And so this is one representative plate where all functions of the heat pump are captured in this and the fluid just kind of goes around this plate. It does this option, which is the vapor production, then it goes up gets purified and then rectifier condenses.

There's some recuperation going on again. And then it evaporates to provide the cooling and then gets absorbed into the solution. So the entire heat pump function is done in one single plate. Of course, that's not enough heat so you then stack a bunch of them together and put a cover plate and then it becomes a very modular system.

You change the features of these passages and the size of these plates to satisfy whatever capacity you need and you also increase the number of plates to meet the duty that you need. But when you're doing things like this, you have to remember that when you're trying to flow fluid through microchannels quite often maldistribution of the flow is an issue. You can see here that if you bring in the fluid at the top in the header, most of the liquid just flows down near the center and nothing goes to the side. So all the circuits area that you provided was of no use.

And so we did a lot of detailed misanalysis and came up with how this wetting is happening. And you can see that the flow is really pretty badly distributed. And so we came up with some simple ways of fixing that where we provided little breaks in the passages, and then the flow gets a chance to mix again at these breaks. And the breaks are also offset so that there's a lateral more movement of the fluid. So with these very passive and simple techniques, we were able to improve the maldistribution if you see on the left side as you go from the top to the bottom.

And then here you will see a poorly distributed fluid with that V shaped valley of just a lot of liquid and nothing else starts getting better as you go through the remaining passages. So these are some ways to address some of these realistic issues and design components that would match your predictions. So the other thing is you have to integrate. It's not just building a small microchannels component, but you get real compactness by eliminating a lot of plumbing and things like that. So the idea is to put many components together to function as a unit.

So this component you see on the screen, actually does multiple functions. It's a full vapor generation unit. It boils the refrigerant, it purifies it, and then it strips off the water of fraction from the ammonia water vapor that is being generated.

So if you open that plate, then these are the geometries that are there so you pour some liquid over here. And then it fills up these trees. It's kind of like a miniaturized distillation column.

And then as the vapor rises, it goes up and then you cool it at the top so that you condense off the stuff that you don't want and only pure refrigerant goes to the other side. So on the other side of these plates, you provide these heating channels here so that the refrigerant is boiled. And then once it is boiled, you provide cooling channels on the other side so that it is purified.

And then you apply some more sort of ingenuity and provide a lot of contact due to mass transfer contact between the vapor and the liquid so that you're accomplishing that in a compact package where you can see that the liquid is flowing in this serpentine fashion, so that its residence time will also increased and then you send the vapor through these small passages so that there's ultimate mixing between them and this sort of flow pattern is set up. And then we did some high-speed videography and saw what kind of interfacial areas are being provided. And with these interfacial areas, we can model the heat and mass transfer better and design the components better. So these are the figure on the right is what we use to do our modeling based on these experiments. One other way instead of just playing around with the geometry is to use surfactants and they induce new sort of convection mechanisms. And you can see that whereas without surfactant, it was kind of a blob that wasn't doing a whole lot.

You are increasing the interfacial area when you put in some surfactant. So these are some ways to take one design and make it for a function even better, so we have been doing this. And you can see that the thermal resistance of that side of the absorption side goes from this tall blue bar to a much lower value and you get 37% reduction in the resistance, or you can get much more compactness with this.

Not only that, it doesn't affect the system at all and you are able to lower the pressure of the absorber and the evaporator, so you are able to do more cooling at lower temperatures when you throw in some surfactant into the strenght of this cycle. OK, so it's one thing to design these, but somebody's got to build it and somebody has got to build it in a way that someone actually wants to buy it. If you make it very, very expensive, then nobody will buy it. So for our lab experiments initially, we use photochemical etching to make these channels. All of these were photochemical etching surfaces, and those provided these microchannels passages. But this is an expensive process, it's a batch process.

Then we are working with a company came up with high volume stamping techniques so that you can just really stamp these plates out and you get these microscale passages. Not only that, these sheets and the original designs were diffusion bonded. And diffusion bonded need a slow and very, very expensive batch process, so then we have developed to take these templates. We have developed brazing processes at these micro scales so that you can actually do continuous atmosphere, controlled atmosphere brazing and this is how you are able to, not only make good designs but make them with high volume and low production cost techniques.

You also have to control these systems. Quite often researchers will just sort of come up with some nice designs and then it would function at one condition, and then one day if it's a warmer day or your waste heat is not at the same temperature, then the system breaks down. So we did a lot of study of the controlled parameters. We predicted the trends in performance. And then we developed control algorithms so that even if the outdoors is just going haywire, you are still getting comfort. And so we first looked at it with high resolution phase change models, which using this finite volume technique to track multiphase, multi component fluids is a very, very expensive computationally expensive technique.

But then we developed a moving boundary technique so that we get basically the same fidelity. But with almost a 60 times improvement in computation so that you don't need a supercomputer to actually run your absorption system. You need something much less that can be packaged into a controlled system that goes into this. So for example, we are able to track and control a variety of transients that you expect or weather changes, or suddenly your heating is not what you expected.

And it will go through a small transient, but then without control scheme, it'll rise back up so that you are able to track and still provide comfort at the same COP so your performance metric didn't change. So these innovative control systems are able to help you run the system without compromising occupant comfort. So these are some examples of what we have done over the years. This was a one time or 3 kilowatt cooling system, an entire cooling system packaged into this using this sort of microchannel heat exchanger concept.

This not only does the absorption heat pumping, but it also has a combustion module in it, so you supply natural gas to it and then you get out cooling liquid just to cool your house. This was an old system where we did this for the defense applications. And then we scaled this one up to a 2 kilowatt system.

And again, the increase in volume was not that much. The one on the right that you see-- maybe I'll use my laser pointer. This unit is a unique one. We have not seen anything like this anywhere at all. It is for forward operating bases and it provides cooling at 52 Celsius ambient.

Think about that 52 Celsius ambient. We're using just waste heat from diesel generators on the forward operating bases. And so no electricity, you take the waste heat from those generators, the diesel engine generators and provide cooling to the soldiers in their tents and so on, even at those very high ambient temperatures. Then on the left side, you see a water heating system. We all have water heaters in the basement. This is the storage tank and then there's some sort of either gas is burned or electricity is provided.

There are electric heat pump water heaters, but we built-- this was another PhD thesis. We built a full absorption system that fits right on top of the water heater. And for every unit of heat that you put in, it gives you almost 1.6 times the heat that you put in

because it does the heat pumping and gives you much more bang for the buck there by reducing your water heating costs significantly. This is another system we built recently. This is a residential heat pump system, but it is tailored to be functioning at ambients as high as 44 Celsius.

So this is a full household air conditioning system using just heat, and it provides cooling at very, very high temperatures. OK, so even in an absorption heat pump, you still need a solution pump, a liquid pump. Instead of a compressor, you need a liquid pump.

But in some places don't have very reliable electricity in distant places in developing countries and so on. And if that expensive compressor breaks down, there's nothing you can do. Even the pump if it breaks down, there's nothing you can do. So we wanted to eliminate even that pump. And so we took inspiration from bubble heat pump where the actual heat makes the bubbles and raises the column of liquid, and that is what is circulating the fluid around the cycle.

So there is no need for a pump, the heat does everything. So then we did some high speed photography, and then we tracked it dynamically with this image analysis. And then we also came up with computational fluid dynamics simulations of that, and then we built this passive system to be able to take waste heat at just about 100 to 120 degrees Celsius and provide cooling for things like vaccines. When we develop this, we had no clue that there was going to the COVID, but here's a system that is providing refrigeration using just waste heat. So then you might think that these microscale systems are only good for very small things, but if you actually build it are very, very high surface areas, then you can actually take, for example, a shell and tube heat exchanger and you can make the microchannels inside. So this is a cascade system where exhaust heat on a Naval ship.

There's a lot of heat that's being wasted in the motive force for the system. You take that and you provide cooling at a medium temperature, and then you use that cold fluid to provide heat sink for a CO2 vapor compression system, so that we were able to provide cooling using to 200 megawatts of waste heat. We were able to provide cooling of 90 megawatts at 5 Celsius, and then also minus 40 Celsius, we provided 50 megawatts. So this is good for bursts of cooling that you need in some of those applications. And we were able to achieve that by using CO2 condensation on the inside and falling for water vapor on the outside.

OK, switching gears. We have looked at sorption, now this is adsorption for CO2 capture. So the way this works is that you have flue gas from the power plant and you send the flue gas over these sorbent coated tubes that are about a 3 micron particles and about 300 micron passages.

These are all sorbent particles and they will squeeze out the CO2 from that flue gas. And then when that flue gas is absorbed that is heat provided and you send coolant through it and that will capture the carbon from the flue gas stream. So if flue gas is coming in and you send the coolant through the center of the tube, and once you have absorb it then in the next phase, you can then heat it up and then capture it in whatever tank you want. And this is the principle that we use. Now it's all well and good, but if you use a lot of energy to do this, then you're producing more CO2.

So we came up with this thing called the thermal wave, whereas when you are doing the absorption it is picking up heat and so you travel along the tube and pick up that heat stored it for the next phase and then use that same heat in the next phase of the cycle to do your desorption. So a lot of internal recuperation so that the parasitic losses for carbon capture are much lower. We then apply this to power plant systems and we did a detailed analysis of carbon capture integrated into the power plants. And we compared it with a variety of other techniques that are being used in our study. The parasitic load is much lower than most other techniques that people have looked at. So another avenue for absorption is gas separation.

When methane comes out of the ground when you're fracking or something like that, it comes with a lot of impurities. And in the previous system, you saw that there was cooling fluid going through the tubes and then the absorption of the gas that is being absorbed going through the outside of the tubes. We thought, well, we can get almost half of the size by sending both fluids through the same channel. So the liquid flows through and then chases the gas down, so the coolant as well as the gas are flowing through the same channel in a batch wise mode. We track the species concentration. We check the temperature profiles, and we did energy analysis on that.

And basically, the separation capacity is very, very high for our system, and also the energy required to do the separation is much, much lower than most other computing techniques. So this is where our system sits. A lot of separation capacity, but doesn't need a whole lot of energy. So that's another avenue for sorption, and here you see the breakthrough curves as the gas is being absorbed. The different temperature peaks and you can see the species also going through. And this is sort of how that happens, the liquid chases the gas and then we go backwards.

So there's the adsorption and desorption phases in here. And we were able to get some extremely good agreement between our predictions and our models. OK, a couple more things and I should be getting close to the end here. Everyone knows about the energy-water nexus. There's a need for pure water potable water and the world is in dire need of water and it causes the lack of access to clean water. It causes a lot of sickness and death and so on.

So we wanted to see if we can do something about this. And we recently came up with this idea that, well, there are purification techniques available in the context of desalination and so on. But how about we combine space conditioning with water purification? So we all take showers and then we have other dirty water, clean water. We can perhaps purify that while also achieving space conditions.

So we broke a typical absorption cycle. We broke it up and made it an open cycle. And instead of sending the refrigerant back to [INAUDIBLE],, which is water back to the cycle. We open it up so that this is your pure water collection and the dirty water, the gray water comes in here goes through one distillation through evaporation, goes through this absorption cycle goes through another distillation. So the double distillation to purify your water, while also providing cooling. So this system does both of those functions and this is sort of the energy ratio what are you getting for this input.

You're getting cooling COP and you're getting a performance ratio, which is how much water have you purified. And so this sort of double dipping for the same input, you're getting almost double the bang for the buck. So this is one way to look at combining the energy and the water needs into one system and not increase the energy input requirements. So my student just recently conducted these experiments. And you can see in the background, the absorber or the desorber and then this is the condenser here in one evacuated chamber.

And we were actually able to take shower water, which has this conductivity and reduce it to almost pure water while doing cooling. So we were able to demonstrate the two functions through this technique. We also need clothes and we also need to clean our clothes and dry them.

And as you know, when the clothes come out of the washer, they need to be dried and it's a very, very energy intensive process. It actually triggers pretty significantly international energy consumption figures. And as it turns out, the energy that's being put into these dryers, a lot of it is wasted, 58% is wasted. And the typical temperature profile in a dryer through the drying cycle offers us some opportunities. Initially, the clothes have to be heated up and then the evaporation starts and then most evaporation has happened and then the clothes have been heated further so that the moisture stuck in their internal polls can also be removed. You can see that as a heat transfer engineer, I see an opportunity.

Here this a hot stuff and there's cool stuff. So we try to sort of match them together and use some thermal storage using sorption. And so we did the analysis at the board level at the fabric level at the dry air system level, and we were able to predict the clothes temperature, the upper air temperature and the evaporation rate. And then these are the experiments in my lab.

This is an actual commercial dryer. And we were able to come up with the techniques of changing the parameters, the rotational speed, and the recuperation doing some absorption of the moisture into silica gel, for example, and removing that. And therefore, providing hot dry air to be reused in the next cycle.

With that, we have been able to reduce the energy consumption significantly by almost 20% to 23%. Some energy saving techniques have been there in the industry, but they usually increase the time required for the drying. And when you're impatient to get dry clothes, you don't want that. So the techniques we have employed, not only reduce the energy consumption, but also decrease the drying time. So this is a significant accomplishment in the drying techniques.

Thermal storage is another big area. And there is a challenge when you're trying to do seasonal storage because heat tends to dissipate and you need to provide lots of insulation and that doesn't usually work. So again, using absorption background, we use this absorption cycle, but then we didn't close it and we took water and then we took the concentrated solution stored them. And then whenever we need the heat, we can mix them to produce that heat again.

So this is the basic concept for thermal storage using sorption. And then if you want to get good thermal storage for short-term, like, 5 or 6 hours then throw in some PCM Phase Change Materials, so a combination of phase change and absorption gets you both long-term storage, as well as short-term storage at reasonable efficiencies. Finally, I covered clothing, I covered comfort, I covered the air. We can't do without food, right? In developing countries, one of the biggest challenges is that farmers produce a lot of produce vegetables, fruit, milk and so on.

And by the time it gets to the consumer because of the hot temperatures and the lack of access to electricity, it spoils. And so that's a lot of wasted effort, and the post harvest losses can be almost $13 billion annually just in India. And so we are teaming up with a very innovative company called New Leaf and they have been building these systems. I'll show you what they are, to be able to extend the shelf life of producing milk and so on using sorption concepts.

And so we are working with them to help them refine their designs. Basically, the way this works is there's a lot of wood chips, coconut hulls and shells and so on, bagasse. A variety of biomass is available. You use that to produce a gas, and then that hot water drives this sorption heat pump like I have been describing in this presentation. And that way, you then tie it to that room that shared like thing that you see in the back, which is where you store your bell peppers or whatever you are producing fruit, apples, even flowers and milk and so on.

So of course, for this, you have to do a detailed component and system level development all the way from inter-particle and intra-particle to inter-particle to then optimizing the heat or mass exchanger in the absorption bed, as well as the desorption bed. And then we've done some optimization of those designs, and this is my last slide. We scientists think a little bit beyond publishing in nature and science to actual impact, then our work can actually benefit by using agricultural waste and putting it to something good so that there is also an economic benefit.

There is no carbon footprint at all here because it's all biomass. And even the refrigerant that is being used as zero global warming potential. There's a lot of price sensitivity and variability month to month in these kinds of products and so you can smooth that out by doing this. And of course, you're combating food spoilage. And in addition to that, you may have heard that countries like India have a lot of air pollution problems, and this will address that, too. So usually, you end talks with the concluding remarks what better concluding remarks than to say that we can sorption can address your basic needs.

It can address air, water, food, and clothing and comfort. So I'll stop with that, and these are my students next to a somewhat famous guy from the past on the Georgia Tech campus that helped me make these kinds of presentations. So again, it's all yours. I think I finished in just a little less than 45 minutes. Yeah. Thank you so much, Srinivas.

It's a fantastic talk. Very, very, very interesting stuff. A number of questions in the chat. One that comes to mind you talked earlier about adding 1% surfactant.

Yeah. I was curious. What's the trade off, like, why not go a higher percentage? What changes as you go to higher percentage and why stop at 1%? Yeah, it's all the convection that is induced, we actually did this study, the exact study you're talking about.

First of all, we looked at different surfactants and we looked at different fractions. What happens is there's a plateau that happens in the surface tension as you increase the concentration. And because these are surface tension driven convection currents, once you reach that plateau anything more doesn't do anything for you.

And then adding more would just only dilute the actual material you actually care about, right? Yeah, I mean, you are nowhere near, but anyway but, yes. It doesn't serve any useful purpose. Got it. OK.

I was curious about also the question came up around the heat exchangers when you talked about them being stamped. I was curious about the geometries because I am a bit familiar with these heat exchangers that end up getting effectively the fusion bonded or a bunch of plates that are combined together. Have you thought about other materials that could be used there that might actually enable, I guess cheaper or more complex geometries with the different processes? It's photochemical etching itself, I guess that's the expensive step. So I guess I was curious what other options that might exist for that? Yeah, very good question and I have a bunch of answers for that.

Yes, photochemical etching is extremely expensive. And when you do the stamping, you do have to sacrifice a little bit of the tolerances and a little bit of [INAUDIBLE].. So they have to be designed to account for that. And so you're giving up a little bit of that microchannel advantage to be able to build it cheaply and so on.

So that's one thing. You're asked about the different materials. One of the issues with the fluids we use, for example, if you're using ammonia and water, it is not compatible with say aluminum. Ammonia is compatible with aluminum, but not ammonia water. And similarly, it's not compatible with copper and so on.

If you are looking at very, very low temperature operation, maybe some high temperature plastics could be brought to bear and then that leads me to the other option. Until now sort of 3D printing and additive manufacturing and all have been kind of coming along at a slow pace, they have done some wonders. However, when we ask about high pressure tolerances at larger scales than just small pieces. So if you have a long channel that's 2 or 3 feet long, to maintain that tolerance has been a little bit of a problem and also some distortion in the channels and so on. And then their speed.

As the additive manufacturing process gets more and more refined, I think this is where this all the technology is headed because then you can actually tailor these geometries, not only as I showed you in the slides, but maybe you're going to have expanding and contracting channels and so on to match the velocity profiles. To match the density of the fluid as it boils or condenses and so on. So that's sort of the outlook for this technology. Interesting.

With the spec on this channel point, though, can you do non-straight channels? And I guess, are there, like, I've seen what it's like airfoil geometries of one not inside the channels, does any of that help you? Yes, actually, it may not have come through very clearly. But when I talked about the maldistribution when you distribute the fluid among these parallel channels, then there is maldistribution and so we break those. Instead of full microchannels, we make them sort of elongated things shall we say, I mean, they're not really thins. But there are sections about an inch or so and the front and leading edges can be-- so we already have a [INAUDIBLE] edge. The other side can also be tailored.

Yes, the airfoil chip can be easily incorporated, especially if we go to these additive manufacturing type techniques. It doesn't matter, which shape you want. So yes, those are some possibilities. That's great.

Someone also asked about in transportation applications. What effect does the acceleration of the vehicle, and I guess vibrations and whatnot in the vehicle affect the ability for these systems to operate properly? It seems like to some extent some part of it still relies on gravity to actually pull some things down. So how does that work or how does that impact the performance? Yeah. Again, an excellent question.

If you were using lithium bromide water systems, then definitely the falling [INAUDIBLE] configuration would suffer a little bit if you are in a mobile transportation system. However, if you use something like ammonia water, it doesn't really depend on gravity. It is more force convective flow, and these are called bubble absorbers. And so that's very tolerant to changes in intonation and so on and so forth. And then furthermore, if you look at adsorption systems, they are actually quite resistance because it's not a liquid. It's not adsorption in a liquid, it's an absorption in a solid.

So different levels of tolerance for mobile applications. Another mobile application that is actually quite a good candidate for this is marine applications. Because if you have a big ship or a fishing vessel or something like that, a fishermen is out there in the sea for a long time and there's a lot of fish. And before they come back, they need to be preserved. So sorption can be a wonderful solution for that because also heat sink is right there. It's a beautiful heat sink.

The water is nice and cool, so your system efficiencies can be very high. So yes, those applications are possible and you can do them without dependence on the gravity or the effects of gravity. Interesting.

OK. Another question that came up is around the CO2 capture system. Does it matter if the flue gas is wet so to speak, if it's got steam in it? There are competitive adsorption processes that happen and you may, in some cases, change the adsorption crystals. And in some cases, you may have to do some [INAUDIBLE] before you get there. And but in one of the techniques that I've described for the gas separation, we are actually using the heating and cooling to get rid of the residual moisture in there, too.

So it can serve the dual function. But yes, the water are presence, the water is an important issue that needs to be addressed when designing such a system. Can you just basically condense it out when you said pre-treating-- is that what you mean? You condense it out first then put the remaining gas through? Is that the idea? One other way we have done is we have done this in stages where the first stage is to get rid of that. And then there's also some innovative sort of recuperative type of systems that we have come up with one set of fibers, but then you do it in stages. You go from one level of purity to the other and then you use some purge gases. Purge gases becomes very important.

So between the adsorption and desorption phases, you have a full cycle to get rid of that. You can purge it with nitrogen or you can purge it with that same gas. So there are some ways to address that. Now, out of curiosity, have you developed, I guess a fully-- you mentioned basically getting rid of the electrical input and replacing with a thermal pump, have you managed a thermal compressor, I guess? Have you managed to make a fully solar-driven airconditioning system that's basically, I guess solar-- maybe you probably need to get some concentration right to get to a high enough temperature to drive, to have a nice hot thermal input.

But I guess in principle, as far as I'm just curious, have you guys done that yet or is there any particular challenge with doing something like that? Can you make a miniature solar-thermal airconditioning system? So we have not actually coupled it to the sun due to an actual solar collector, but the heat source we have provided to that massive system that I described for vaccines and so on was, in fact, run using water, hot water that was only at just over 100 degrees Celsius, which is fairly easy to do. And that was a fully passive system. Basically, you set it out in the sun and it just does its cooling. There's nothing more you need to do. So we have built that, except for the fact that we did couple it to a solar collector, while we couple it to fluid that would be representative of what's possible in the solar collector.

I guess if you one takes a step back and thinks about how the cost of PV has dropped, how does this approach stack up against just using cheap PV renewable electricity from, I guess PV or wind and just running vapor compression directly? I mean, if you definitely don't get the advantage of the effectively zero GWP approach, but in terms of I guess just the economics of delivering cooling, how do they compare? Does it make sense to do, like, a solar thermal version? Is that actually work out better in the end in terms of economics or are they comparable, or do you have any thoughts on-- This is a layered and evolving issue. When you go through the PV route to get that electricity, you're still at a fairly low efficiency. There is no fluid available or there will be no acceptable fluid available in the near long-term to run those kinds of systems.

That's one issue. The other issue is let's say a village in India or some remote village in India, they are more likely to have twigs and cow dung and stuff like that. Variety of biomass is far more readily available and it can be done on a community scale or-- it doesn't have to be an individual house scale, and so on. So this kind of thermal route is far more versatile because when the villages are full of farmers and farmers have things like corn stover, they have rice husk stubble and so on. There's a lot of biomass available, which then you won't be able to use if it's a photovoltaic system. Now photovoltaics definitely has made leaps and bounds of progress and it has also become cheap.

So I guess the best answer for that is on a case by case basis depending on the availability of the energy source, and what that in this application is both are viable options. Now, this gets to an interesting question about-- is there any way, I mean, one of the things that I think is quite amazing with what you've been doing is the compactness. Is there a way to possibly integrate this in the context of electric vehicles? Can you maybe use the waste heat from batteries from the actual battery to help run some of the space cooling, or is that it's not hot enough? So that's an interesting question that goes well with my history. When I finished my PhD and I went to work for GM, I was the lead thermal engineer for the very first electric vehicle concept.

It wasn't even the EV1, which is my car [INAUDIBLE] came the impact. They called it the impact. And this was a big issue because electric vehicles, you want the battery energy for range rather than for comfort.

And every jewel that you are expending on comfort, you're losing range. So I actually developed a system and I have pattern to collect at least in those days, the battery waste heat, as well as the inverter mode of waste heat and then throw it all together. You're right, it isn't very hot, but it serves as a much better heat source than the ambient air. So if you're driving in minus 15 degree weather and you, at least, have say 10 degree heat source due to collecting all of this waste heat and that serves as your heat source to drive your system.

You gain a reasonable advantage in that. So yes keep pumping with this is a good option except that heat could be used either in a vapor compression system or in an absorption system. Again, matter of trade-offs there.

Very interesting. Maybe one last question, I guess, is thinking about-- is there a very large market opportunity where you can imagine maybe some standardized version of this technology gets mass-produced and really has a tremendous impact across the globe or in some major sector? Is there a particular market that's jumped out at you that is like the one you would go after? Excellent question. In 10 years ago when I started working on these very compact systems, that was the sort of goals that we had, which is to make them compact and modular. And we achieved the compact and the modular. The market over the past few years has been going through a lot of volatility in terms of a carbon signal, in terms of gas prices, in terms of the move towards electrification that you see.

So much as I would like to tout this as a one size fits all for the entire world. I think there are some steps to go before you will get there. I think the mobile market looks attractive, but I don't think it is the car that is the first target. It is more refrigerated transport, for example, for your semi-trucks and things like that or the marine application that I mentioned. So that you can learn and refine these technologies at that larger scale where it is not an individual's price sensitivity, but it is a market segments price and their energy costs.

For a semi-truck that is delivering all of these goods, their primary energy cost isn't their primary cost is, in fact, energy. And so if you can help them that really-- so the point is that the essence of the technology remains the same. It's either you're building a small system or a big system. There is no more scaling issues the fundamentals.

Whether you build a 1 kilowatt system or a 1 megawatt system, the Reynolds number at which you will be functioning will be exactly the same. And so the scientific challenges by enlarge being addressed. It's a question of finding the right application and sort of the market variability, which is well beyond my capabilities to make any sound predictions. I can predict the past very well. [LAUGHS] All right. Well, thank you so much, Srinivas.

This has been a fascinating talk. It's been very, very educational to learn about your work and get it all put in context. It's very, very intriguing, very interesting, very clever design, very clever choices and innovations.

So thank you so much. We'll go ahead in the session here. Let me just remind everyone that our next webinar will be March 17, Advances in Additive Manufacturing for Heat Transfer Applications. Maybe there'll be some discussion about possibly making some of the geometries needed in these absorption systems, and that'll be moderated by Saniya LeBlanc from George Washington University. So thank you, everyone. And thank you all.

I enjoyed doing this. Thank you. Thank you. Thank you so much for doing this, Srinivas. It's, like, I said very fascinating work.

2021-03-09

Show video