Stanford [00:00:12]: Welcome back to the second season of the RSM River Mechanics podcast, a four episode mini season on reservoir sediment processes and management. At the end of last season, I asked for guest recommendations practitioners in these fields that you would like me to talk to, and todays guest was one of the most requested. Doctor George Annandale is one of the most recognizable names surrounding reservoir sedimentation and is something of an ambassador for the future of our global water supply, which is more connected than I realized before I got into his literature to reservoir sediment management. His book quenching the thirst includes a global assessment of water supply trajectories and some pretty shocking trends with unsettling implications about the future of fresh water availability and how reservoir sediment processes are already impacting that future. But Doctor Annandale doesn't just sound the warning about bleak trends. His work focuses on actionable reservoir sediment management practices that can mitigate reservoir storage loss. In addition to private work on reservoir sediment problems around the world, he's also worked for the World bank on several Reservoir sediment management studies at various scales and helped develop the Rescon software, a screening level tool that helps decision makers sift the broad menu of sediment management practices to figure out which ones are worth investigating at their reservoirs and which are nothing. Starters I thought this made a perfect intro to this series because George and I talked about a lot of important contextual ideas surrounding reservoir sedimentation that sets up the technical detail in the following episodes. We talked about the motivation and even the ethics of reservoir sediment management, and unpacked some of the economic models that got us where we are, and talk about how some countries and decision makers are revisiting those models to realign incentives. We talk a little bit about Rescon and just introduce the different methodologies available to manage sediment at a reservoir and extend the life of a dam, methods that will delve into in more detail in following episodes. So I'm thrilled to kick off this second reservoir centric mini season of the RSM River Mechanics podcast with a conversation with Doctor George Annadale. Stanford [00:02:11]: George Annadale, welcome to the podcast. George [00:02:13]: Well, thanks and I'm glad to be here. Stanford [00:02:16]: You grew up in South Africa and I suspected that that might affect the kinds of projects that you're interested in and how you approach water resources. Is there a biographical connection to where you came from and the kind of work you do? George [00:02:29]: Yeah, well, absolutely. You know, the thing is, and I grew up in South Africa, and it's common in South Africa to have droughts of five to seven years and then you have a wet period again and, you know, that obviously affects you. So since I can remember, there's always been water restrictions. As I grew up a little older then the issue that we losing storage became an issue with sedimentation. So that since a kid, that's always been in the back of my mind. Reservoir sedimentation. Currently I'm living here in Arizona and I feel really at home and very happy. And it's really sort of a south african climate. The parts of where my grandparents had their farm, which was sort of semi desert, and it was always problems with the rain, and that you have to have dams. And the other thing in South Africa is that groundwater resources is very scarce. So water supply really comes from surface water dams. Stanford [00:03:23]: In your book, quenching the thirst, you start out with the idea that sustainable development is a phrase that lots of people use to describe lots of different things. But maybe that phrase deserves some precision. So what is sustainable development? George [00:03:39]: What I try to do is to distinguish between the term sustainability, which is used generally, and it's got many meanings for many people. And then there is sustainable development. And that is really what we as engineers do as sustainable development. That's what we should aim at. And sustainable development is defined in the Brandland report in a very succinct manner. And it says that sustainable development seeks to meet the needs and aspirations of the present without compromising the ability to meet those in the future. Stanford [00:04:13]: And so I think I've been familiar with your work for about ten years. And the phrase that you use that I think has become the biggest problem of not only my lexicon, but my worldview, is this idea of intergenerational equity. What is that? And how does that have to do with sustainable development? George [00:04:29]: Intergenerational equity is really to make sure that in the way that we use our resources, that we don't harm future generations in the sense that we use them in a non renewable manner. Because you have renewable resources and you have non renewable resources. So that is really the issue that we make our decisions and design our projects in a way that future generations can still benefit. For example, I mean, a simple one that always comes up. If you have a dam in a reservoir and it completely fills your sediment over 30 years or 50 years, at the end, the generation that needs to take care of that decommissioning hasn't benefited from this at all. So the whole idea is to not let that happen. So if you, for example, have a dam in a reservoir that you cannot do anything about, but it's going to silt up, then, for example, you can create a sinking fund and you put money in the fund so there's enough money at the end of the life so that somebody can decommission that reservoir. And that is not really a new idea because that's used in mining. I mean, if you develop a mine, you must put money aside to decommission the mine right from the beginning so. Stanford [00:05:43]: That the generation that is benefiting is also paying the tail end cost. George [00:05:48]: Yes, that's exactly it. The problem is with the current way that we projects, and I know that is sort of standard in us government and World bank also use those techniques, is that we use interest rates that are so high that if you discount everything in the future, after about 30 years, whatever happens after that doesn't matter. So that doesn't create intergenerational equity. And there are techniques like the british government and the french government use it to really look at what's going to happen in the future. And those benefits and the cost that you also account for that in your assessment. I mean, economics is really a philosophy and it's got some math to make it a soccer. But mostly if you talk to the economists, they say, well, the next generation is going to be much wealthier and they can do this. No, it's not going to happen anymore. You have to look broader than that. And I think there is a path. I know there's economists in the World bank also that's starting to look at this and say, well, you know, we may have to look further into the future because a lot of the facilities, I mean, we currently, a lot of people live to 100 years. So, you know, it's not like it's far out in the future. You know, it's, and I think really if you look at it, you know, currently, if you look at the dams we've built in the past from say, 1900, and we look@the.net, storage after sedimentation has been taken off, we are really on a declining road. So we have not benefited the future generations, which are us now. Stanford [00:07:28]: I want to just pause on one of the things you just said because I think this is one of the most astounding findings in your book. It seems to me that especially outside of Europe and North America, we're building a lot of new dams. That means we have a lot more new reservoir storage. What are the actual trends in reservoir storage in the world? George [00:07:47]: Since I wrote the book, I got some more data from icold up to 2022, actually. And what happens is, you know, the, internationally, the trend of construction continues and if you look at the trend of adding storage space over time, it keeps on increasing. Okay? But then if you account for reservoir sedimentation, what we find is that now we are losing more storage space than what we add in. So there's really a negative trend in the amount of storage that's available. And then the second thing, if you look at storage per capita, it's even worse because it peaked in the 1980s and then it went down. So right now we are about where we were in 1950s and it's not moving forward. And in the United States, it's even worse because we're not building dams. And the decline in the net storage space is more significant than what it is internationally. And also the same thing with storage space per capita, the same thing, it goes down to 1950s, where we are now. It's not a good thing because as I try to explain in my book, is that we need to use renewable resources to be sustainable. And if you look at the most renewable freshwater source, it's rivers. If we want to use rivers, we have to build dams to make a reliable supply and we have to manage those storage spaces. Stanford [00:09:22]: This is one of the big arguments of your book. Let's pause on it just a little bit. There are other places we could get water. The other big one that you deal with a place where we get a lot of our water from, is groundwater. Why is surface water the future, and why does that mean more reservoir storage? George [00:09:40]: We need to step back and say, okay, we have renewable resources and non renewable resources. That's, those are the two. If you look at groundwater as such, and you look just at the fresh groundwater, for a resource to be renewable, the rate of replenishment must be greater than the rate of usage. You must be replenished more than what you use it. Stanford [00:10:04]: I'm not an economist, but I can do that math. George [00:10:09]: If you look at groundwater, you know, if you were able to, all the fresh groundwater on earth, you were able to remove it in an instant. It'll take a 1400 years for it to recover to where it is now. Stanford [00:10:21]: 1400 years? George [00:10:22]: 1400 years, okay. And on the other side, if you look at river water, if you remove all the river water on earth in an instant, it's back in about two weeks. So for river water, the replenishment rate is much closer to what we use it on a daily basis. And if you look at groundwater, 1400 years is so far off, you cannot classify that as a renewable resource in that sense. McGill University and University of Woodraft, they've done a joint research project looking at groundwater usage, and globally, we are using three and a half times more groundwater than what's replenished annually here in the Ogallala aquifer, it's seven times more. The valley in California is just the same level. And then if you look at another place where we are getting a lot of our vegetables and produce from here in the United States is northwest Mexico. They're using 27 times more groundwater than it's replenished. So that's not a sustainable way to go. On the other side, if you have these river water, and obviously, the issue is that it doesn't flow constantly. Rivers doesn't flow constantly. So you need to build dams to store the water. You know, you catch water in times of floods and you store it for times when, when it's dry that you can release it. Stanford [00:11:51]: So your argument in the book is we are essentially mining groundwater, and into the future we're going to run out of that resource. We'll have to replace it. But then there are two amplifying factors. One, we have to replace it at a higher population, at a higher use rate, and there's the complicating factor of climate change and the covariance that comes with that. George [00:12:13]: If we look at using river water, obviously, first we need to look at reliability, as I just said. So we build dams to store water that we can supply it in a reliable manner. But the problem is that, as you know, is that reservoir sedimentation diminishes storage. And as it diminishes storage, it reduces the reliability that we can provided with. But the other problem that we are entering now is this whole issue of climate change. And with climate change, everybody is. And I think if I see the scientific community is in agreement that hydrologic variability is going to increase. So the issue with variability is that if variability increases, it means that the length of droughts is going to increase. Okay, so I said in South Africa, I grew up in an area with droughts of five to seven years. And most of Africa is like that. And the reason is that they don't have mountains of snow that can, you know, it's all rain fed. And in areas here, like where we have snow, you can, the flow is less variable. But as climate change comes in and snow starts to disappear and you still have this increased variability on rainfall, you can have increased variability on your river flow and you can have longer droughts. Stanford [00:13:37]: So we've mining groundwater in a way that's non sustainable. We've got more population. And more variability means potentially longer droughts. And a higher need to buffer that with storage? That's correct, but we're not actually adding storage faster than we're losing it. Which brings us to reservoir sedimentation. Can you just give us the simplest primer on why do reservoirs fill with sediment? George [00:14:05]: So, you know, even people that are not engineers or technicians, they would often look at the river and it's very dirty. So that sediment being transported by the river, and then as the river flows into a reservoir, the water slows down and it cannot carry that sediment anymore. So it settles out in the reservoir and the space that the sediment takes up replaces the space that you can use for water storage. So that is really the big issue. Stanford [00:14:34]: I feel like this is the one thing that I do that I can explain to people at a very boring cocktail party and they actually understand. But reservoir sedimentation, it didn't surprise anyone when we built these dams. We actually knew that this was going to happen. In fact, the engineers that designed the dams with very thin information and analysis methods often came very close to predicting how much sediment would fill these reservoirs. It didn't seem to be a concern. Why does it seem to be that people all of a sudden are caring about this? George [00:15:05]: Well, at this first thought, I think there's a few people caring about it because, you know, that's one of my big peeves. And I still work on a number of world bank projects on panels of experts, interact with design engineers. The thing is that there is a design paradigm that developed in the fifties, sixties, and people knew that reservoirs were filling with sediment, but they said, okay, that is, we can't do anything about it. Just we keep on building dams and they follow sediment and build another one. Okay. The problem is that that design paradigm still present today. Stanford [00:15:44]: Yeah. George [00:15:44]: It has not changed. There are a number of people that are concerned about what's happening to our water resources in terms of storage, but it hasn't sunk into the design engineers. Stanford [00:15:55]: What is this design paradigm that you're talking about? George [00:15:58]: Well, the design paradigm is really, people use this concept of a design life. And, I mean, it's a good concept to use for roads and other stuff. I mean, a road, just think of it. You can design for a design life of 15 years. After 15 years, you repave it and essentially you have a sustainable resource. You developing it sustainably. Stanford [00:16:18]: Yep. You come back every 15 years and you put up the cones and you slow down traffic, but you haven't lost. George [00:16:23]: The resources and it can be used for many generations. I mean, you say highway system in us is a typical example of that. But the thing is, with big projects like dams, if your reservoir is filled with sediment, you can't do anything about it. I mean, the cost of removing that sediment once it's filled is hundreds of times more than what the dam cost to build original. So this whole concept of a design life, because I'm working on Rogan Dam at the moment in Tajikistan, and we're struggling with that right now because at first when I got there, the World bank said, oh, this is 100 year design life. Now this is going to be the biggest dam in the world, the highest dam in the world, 100 years. It's not good enough. Stanford [00:17:05]: No. George [00:17:05]: So you have to get past that concept and you have to get to a life cycle kind of concept where you can continuously work on this facility and you make it last forever. That's the ideal. I mean, it's not always possible, but you want to last much longer. Now, just to give you an example, Rogan, I can say it's going to be the highest dam in the world once it's done, and it's got a huge reservoir. And the only way that you can manage sediment in that to extend the life of the facility is to vent density currents, because there are density currents, we know that. To vent density currents, we know that's there. And I just, you know, and I did a quick assessment, and you can increase this life another 50 or 60 years by just doing that. Then obviously in that case, you have to sometime it's going to sort of. But if you can extend it long enough, multiple generations can use it. Stanford [00:18:01]: What did you do for the World bank and how did that affect the way that you thought about projects? George [00:18:05]: When we moved here from South Africa, the United States, in 91, I always tried to get into the World bank kind of projects. And then eventually in 2001, I made a presentation on reservoir sedimentation and reservoir sedimentation management to the World bank. And it so happened that Alessandro Palmieri had just joined the bank at that time as the dam safety engineer, and he was obviously also looking for something to do, I guess. And he is interested in sediment management. And then there was a professor Shah, University of Connecticut, that was just doing a sabbatical there. And then he went back and we decided to start off with this. So what happened is we are working and developed this reservoir con program, reservoir conservation software. In that time, in 2003, when we finished it, it was sort of just a spreadsheet, and that's been around for a long time. I think that sort of set off my path at the World bank and I've done a number of other projects for them. You know, for example, in Nepal and India, did a three week tour there to look at sedimentation of those reservoirs. But then recently, about three years ago, the austrian government funded a project, a much larger project. They funded about $2 million, where we were looking at ways to advance reservoir sedimentation management and to enhance the, the health of rivers. And so that amongst other things, it included studies on the Niger, the Danube and the Mekong river. And then what was also became a book out of it with Gregory Morris and myself and Praveen Karki. And then I think a number thing that was good that was developed was the advanced version of the Rescon program and it's been released recently. And what that allows you to do is to do a pre feasibility level assessment of the potential to make, do reservoir sedimentation management on a new dam or existing dams or for policy on a whole group of dams for a country or whatever, how you want to prioritize things. Stanford [00:20:30]: It's not a detailed hydrodynamic model. It's a screening level tool that lets you look at a lot of options fast. George [00:20:35]: Oh yeah, that's the whole idea. It's really, I mean, we use back of the envelope techniques to do things. And so what you do is you would, you know, first is putting the data for the reservoir and it will screen all the reservoir sedimentation management techniques at a high level and say, well, these are potentially feasible, potentially not. And then we do an economic analysis on that, which is very important. In this case, we introduced new methods for economic analysis, which is not normally used. And then it also does a very preliminary climate change assessment and helps you to identify the optimal solution based on the uncertainties associated with climate change. But like I say, it's a very high level and it's just the first cut. Stanford [00:21:26]: Okay? So it's hard to get your head around what is the comparable action to repaving a road, even if you wanted to, even if you set it up from the beginning. But that's a lot of kind of what you're famous for and what you cover in your book. There actually are ways that you can build or retrofit reservoirs in order to get them in this lifecycle paradigm and extend their design life essentially indefinitely or much longer. You broke it down into three basic categories. You can keep the sediment on the land, right, but that's very distributed and takes a lot of work and culture change. You can pass the sediment through the reservoir so it never deposits and just keep it suspended as it going through, and then once it does deposit, you can remove it. And you're best known for those latter two. Can you tell us a little bit about what are the options for either passing the sediment through the reservoir or. Stanford [00:22:20]: Removing it once it's there? George [00:22:22]: Okay. You know, I think the, you know, that is a good categorization. The thing is, you know, one is to try and pass the sediment through the reservoir or alongside it. So a way to. There are two ways you can do it. Pass it through the reservoir. The one is that you can you do what is called sluicing. So when you implement that is when the flood season starts, the flood comes in and you lower the water surface elevation in the reservoir so you can keep the flow velocities high and you try and pass as much sediment through that as possible. Okay, that's one way. Another way is in certain cases, like I said, in the case of Rogan, the character of the sediment is such that density currents form. Stanford [00:23:12]: Okay, I knew we were going to come back to this, so I was going to ask you to define that. Now, what is a density current and how is it different than the other thing you described? George [00:23:22]: So, with density current, if you have a lot of very fine material, silt and clay, and the river transports a lot of that. That water, because of the silt and clay that it contains, is denser than the water in the reservoir, which is clear. So if this river flows into the reservoir because it's denser, it goes along the bottom of the reservoir and it forms what is called a density current. Sometimes I flew over like mead once, and you can actually see the Gloria river come in and it's suddenly clear. The reservoir is clear. So there's a density current going into Lake Mead. Stanford [00:23:59]: So the sediment plunges. And so the thing that's hard for me to get my head around is you're not transporting sediment by shear stress. It's not that the river is moving the sediment by the force of the river, it's that this very dense sediment plume is actually going from a high density to low density gradient. Well, they call it a gravity current. Right. It's being transported by its own weight. George [00:24:20]: Yeah, yeah, that's right. Yeah. It's essentially, you know, I would also view it perhaps, as it's, you know, if a river flows normally, you have the atmosphere, it's also a fluid. And now if you have this clean water, it's like a fluid, but it's less dense than the river, and it goes underneath and it just carries on like gravity, gravity pulls it along. Stanford [00:24:42]: That's cool. To think about it as a three layer fluid. You have the least dense, which is the air. Then you have the middle dense, and then you have the most dense underneath it. George [00:24:49]: That's right, that's right. Yeah. So those, and those density currents travel very far. I mean, they sometimes, like in case of Xia lung d, it's 100 km. So what do you have is you have this density current containing the sediment. And then if you have a low level outlet at the dam, you try to get that current to go through. Stanford [00:25:11]: The dam, and you can pass the density current without actually drying down the dam. You just open it up. George [00:25:17]: And that is a big benefit for hydropower, for example, you can just keep your reservoir level high and generate power. And the density current comes in, you just open it and you release it. Okay. You're going to have possibly some lowering, but it's not significant, and that's a big thing. Stanford [00:25:34]: What is the difference between sluicing and routing? George [00:25:37]: Sluicing is a subset of routing. Okay, so I would call sluicing and then scan. Venting is routing. Stanford [00:25:48]: So in the Venn diagram, routing is all of the methods that try to pass the sediment through before it deposits or past it or around, including bypass sluicing and density currents. George [00:26:00]: That's right, yeah. And that brings us into the other techniques, the bypass technique, because that is, you know, like in, there's a lot of reservoirs in Switzerland, in the mountainous areas, and also in Japan, what they do is they build a tunnel, and when eye flows come in with sediment, they divert those flows into the tunnel and pass it around the reservoir and discharge it downstream of the dam. And so, and that's, they're very successful as actually minimizing sedimentation that way. Stanford [00:26:32]: There's a flood risk management project in Seward, Alaska, where the town is built on an alluvial fan. And so instead of trying to keep the reservoir open, they're trying to keep the channel open, and they actually bypass the sediment down through another channel and form the fan and somewhere else. George [00:26:47]: Oh my God, I must go and look at that. My son lives in Alaska. I don't know that. Yeah. Stanford [00:26:53]: Okay, so density currents bypass and sluicing. And my question with sluicing is, we've talked about hysteresis with a couple of guests on this podcast, and a lot of times you have more sediment on the rising limb of the hydrograph. Is that something you can take advantage of with sluicing? George [00:27:11]: Yes, yes. You know, and I think just as an example of a success of sluicing is Sinar Dam in Sudan in the Nile river. So the Brits built it in 1925, they finished it, and they have 80, 80 low level gates in it. Stanford [00:27:30]: I can't even picture that. George [00:27:33]: So the rule there is, as the flood season starts, in the beginning, you sluice that flood through. So that's got most of the sediment in it. Stanford [00:27:42]: Okay. George [00:27:43]: So you lower it and you sluice it through the reservoir. Stanford [00:27:47]: And that's important because then you can use the back end of the hydrograph to fill the reservoir back. George [00:27:52]: Exactly, exactly, yeah. And obviously, the Nile river, you have these huge, long flood seasons, so that's ideal to do that. And they've been very successful with that since 1925 to 1981, the reservoir storage loss was 0.4% per year. It's almost nothing. Stanford [00:28:11]: How many years is that? George [00:28:12]: It's about 56 years, 0.4% per year. Okay. So it was probably over the 50 year, it's about 20%. But then in 1981 to 85, there was one of the district engineers decided not to implement that management policy, and the reservoir filled up at 8% per year per semester. So over the five years, it more than doubled. The previous 50 years went from 20% to 60%. Stanford [00:28:45]: Oh, wow. George [00:28:46]: So, you know, it does work. There's no question about it. So. And then, obviously, then other techniques would be like off channel storage. Gregory Morris designed some really successful projects like that where you have a reservoir off stream and you divert the clean water into it, and if a sediment flow comes, you divert the flow past the reservoir. So that is. That is the routing techniques. Stanford [00:29:14]: So then we have this other class of techniques where we didn't manage to pass the sediment during the event, the sediment deposits, and we have to come and afterwards move the sediment out. And what are some of the options there? George [00:29:27]: Okay, well, that is the most well known option, as far as hydraulic removal is concerned, is flushing. Stanford [00:29:33]: Yes. George [00:29:34]: Okay. So the difference between flushing and sluicing is in flushing, you implement during the dry season sluicing, you implement the beginning of the flood season in lower reservoir elevation, and it flows through. But flushing, you essentially empty the reservoir. And what you aim to do is to create river like conditions in the reservoir so that you can erode that sediment out and pass it downstream. That's really the difference, you know, and an example of that, this kabidim dam in Switzerland, they flush that on an annual basis. I've been there. It's really a very successful operation. And then, you know, other techniques. Obviously, dredging is what everybody knows, you know, so that is something that you can do on a regular basis. And Cogswell Dam, for example, another technique is to empty the reservoir, flood control reservoir, for example. And you send a lot of earth moving equipment in there and you dig it out and you put it in a landfill somewhere. Stanford [00:30:32]: Oh, wow. George [00:30:33]: And then there are another technique, dredging technique, like this hydrosuction. There's an engineer in Norway that does a lot of that. He's old companies based on that. He seems to be pretty successful with that. And the idea there is that you essentially use the siphon power over a reservoir to use instead of a pump. So you just create the suction and you move the sediment out. Stanford [00:30:56]: The magic of the Bernoulli principle is that you have this head difference, this giant head difference. You could potentially use that instead of gasoline to power your dredging. George [00:31:08]: Exactly, exactly. Yeah. Stanford [00:31:09]: So the first time I heard about this problem, and I think that probably if some people are listening and maybe it's the first time they've thought about this problem, it seems like there's a pretty easy answer. You open up the dam gates and you let the water blow the sediment out. The flushing, it seems very straightforward. What percentage of dams does that work for? What kinds of dams does that work for? And what are the biggest obstacles to that? George [00:31:39]: Yeah. You're referring to flushing? Stanford [00:31:41]: Flushing, yeah. George [00:31:42]: The thing is a flashing. You know, I think what this is, first start, there are two kinds of flushing. There's pressure flushing and then draw down flushing. Pressure flushing is when the reservoir is water, surface elevation is high. And you open these gates. And essentially what you do is you just scour a cone around your gate and that's okay for, you know, if you have, for example, an intake just above the gate, just to keep that clear of sediment. But for drawdown, flushing experience has shown that it's really relatively small reservoirs. You know, there's a lot of work that's been done by Professor Sumi in Japan. And he got all this data and I developed this one graph really using his data. So flushing is really for relatively small reservoirs, relatively short. And then also it has to be have a linear shape, not the multiple legs and stuff like that. Stanford [00:32:36]: If it's narrow, you can open the gates and essentially you'll clear the footprints of the gates in a little bit. But what happens if the reservoir is really wide? George [00:32:44]: It's very wide. You're just going to create a channel through the sediment. The rest is going to sit there. Stanford [00:32:49]: So the sediment can deposit much more broadly than it can, erode. George [00:32:54]: Yes. So that is where Gabedan dam is a good example. A very narrow channel and they can flush out all that sediment. But I think you've done some work here in. I think it is in Nebraska. Stanford [00:33:06]: Yeah. Spencer Dam. George [00:33:07]: What I saw from the videos that you guys developed, it's essentially a channel. Stanford [00:33:11]: That's right. George [00:33:12]: That you scour out within the sediment deposits. Stanford [00:33:15]: So they had to do it twice a year just to get enough head to run that. Run that power plant. George [00:33:20]: Yeah. Yeah. So that, you know, you need some special, special features there. And that's actually where res Con is quite nice as a preliminary assessment because we used work that was done by Wallingford. Doctor White. He's retired now. But he's got some very empirical techniques that you can quickly assess how successful you're going to be. Stanford [00:33:40]: This was my first experience with Rascon is I was in a workshop with you. The whole purpose of the workshop was to analyze the different sedimentation management possibilities for tuttle reservoir. I just thought it has low level outlets. We should be flushing this thing. And I remember you brought up Rescon and said well let's see. And you typed in a few numbers and said, you know, it's just too wide. We'll only remove a very small percentage of this sediment. And it just was a very powerful tool. George [00:34:09]: I'm glad. I'm glad to hear that. Stanford [00:34:12]: So in my experience, one of the obstacles to flushing is that because of this old design life approach to dam design. And a lot of our reservoirs don't even have low level outlets. You can't actually open the dam to bring it down to run of river. What's your experience with that internationally? George [00:34:33]: Let me put it this way. There are requirements for dam safety purposes to reduce the water level at a certain rate. But what you find is if the dam becomes very high, gates that you can get off the shelf can take about 120 meters pressure that's often not built in, I would say internationally. Most dams that I know of do not have low level outlets, even if they are within the limits of gates. It just doesn't seem to be something that design engineers value such except for dam safety purposes. Stanford [00:35:12]: And low level outlets would be useful for multiple of these tools. George [00:35:15]: Yes. Yes, yes, yes. Stanford [00:35:17]: What are some of the recommendations that you tend to make for new dam to get them into this lifecycle design paradigm? George [00:35:23]: Changing minds is difficult. This project we had with the World Bank. I may be digressing a bit, but this is important because in a big way I diverted myself from the technical stuff in the last several years more towards policy. And I developed a policy document for the World bank with this last project. And then while we were doing that, I discovered that the University of Virginia has a PhD program in sustainability and cognitive psychology. Perfect. We tried to get a contract going with him, but money ran out before we could do that. I'd use now an example. I'm currently working with an engineer in Greece, Doctor Ftamio. He finalized this new reescon program and we working on a dam in Colombia. And Colombia is one of the few nations in the world, or countries in the world that actually require lamb owners to develop reservoir sedimentation. Stanford [00:36:32]: Oh, really? Oh, wow. George [00:36:33]: Yeah. So we are working with them and there is a low level outlet on this particular facility. But that's not the most effective way manage the sediment. We found out, you know, this whole project that we're doing is we found that the combination of dredging and a bypass tunnel is the most effective. Stanford [00:36:54]: Interesting. George [00:36:54]: Yeah. Yeah. So it's very. It's very project specific. Like I say, in the case of Rogan, it is density current is only thing you can do. And, and we've seen that also in Eurek dam downstream over there, they are density currents and they have to vent them. You know, what do I recommend? I try to always talk about this old story about what we need to do as engineers. And actually recently I gave a keynote at USSD last year to try and shame people into doing this. I said, well, what is the ethics around this? And I refer back to the. The fundamental ethical principles of the American Society of Civil Engineers. And I was actually. I think I proposed to USSD that we should actually develop ethical principles for United States society of dams as well. There are four fundamental principles of ASCE. And the first one is to create a safe, resilient and sustainable infrastructure. Number one, sustainable. They didn't talk about standard development, but I think that's what they mean. And then to treat all persons with respect. And then third one, to consider current and anticipated needs of society. So that's where the sustainability becomes sustainable development. And then lastly is to use our knowledge to enhance the quality of life of man on earth. So, in essence, those ethical principles support and impress the importance of sustainable development. If I try to convince engineers to look broader than what they're doing at the moment, we have a lot of uncertainty in designing dams. I mean, it's tremendous. And we've developed geological techniques and we do seismic analysis and we do various things to make sure that this dam stands up. But what we forget is that the dam's purpose is the reservoir, to keep the reservoir. So I say, in principle, there are only two failure modes in a dam reservoir system. The one is the dam fails and you lose your reservoir. You cannot use it anymore. Or the dam stands up and you fill your reservoir of sediment and you can't use it anymore. The whole facility is useless. So we need to think broader than just the dam and go to this whole concept of the system as a unit. It's a dam and a reservoir. Stanford [00:39:26]: Speaking of failure modes, one of the things I'm a little bit obsessed with is reviewing projects that haven't gone well. I think one of the best ways to kind of learn is by counterexample. And I suspect with a number of reservoir management studies that you look at, besides the primary failure mode of not considering it, or this kind of design life design strategy, which is kind of itself a failure mode, what are some of the failure modes that you've seen in attempted reservoir sediment management? Or can you give us a couple of examples of places where this has been tried and hasn't worked very well? George [00:39:59]: In Bhutan, there's a number. They also tried it like a flushing. You know, you can only have a certain amount of sediment removed. I think that is probably one fatamotive. I can call it that. Another one is like, for example, in Rosera's dam upstream of Sennaar Dam. It was built after Sennaar was probably in the seventies. And they provided sluicing gates. But there's a few. I mean, I told you, Sennaar has 80 of those. 80, they have six. The six gates are here and the hydropower intake is over here. Stanford [00:40:34]: So it's on the opposite side of the. George [00:40:36]: I mean, this plant filtered up so quickly. I think what it is, is not really understanding what you are aiming at and not doing the necessary studies to ensure that it works. Well, I'll give you. This is a good one. This is Natbayakri in India, and it's a huge project. I mean, I think the head is 700 meters. And the way that they tried to manage sediment, this is now, not to abrade the turbines, was to build huge sedimentation ponds in the mountain. I mean, these are massive asset. The first year when, after they commissioned it, within nine months, all those turbines were done. Stanford [00:41:12]: Oh, wow. George [00:41:13]: And it was just essentially silt, but they have like 80% quartz in that silt. Stanford [00:41:18]: Oh. So the residence time wasn't high enough to strip out the silt, but the silt was too high of quartz content. George [00:41:25]: So they got a lot of the sand out, but not the silt. And the silt had a very high quartz content, so it just screwed up the turbines. So the way they solved it was to redesign the turbines so that you have less wings and the flow batten is more gentle. Stanford [00:41:42]: Oh, so it's actually a mechanical engineering engineer. One of the things that I've enjoyed about, about this conversation so far is it really does illustrate your idea that reservoir sediment management solutions are very science specific, and I would have never thought of a mechanical engineering solution as part of that. George [00:41:58]: Yeah, yeah, absolutely. Stanford [00:42:00]: So one of the things that we run into a lot is, you know, our government has spent billions of dollars trying to keep sediment out of channels. And so there's this been the sense in the United States for many years that sediment is a pollutant. We have tmdls. And so the idea that some engineer is going to want to open the reservoir and send a bunch of sediment through, if you were, say that in a room with, say, some resource agencies, you get some interesting reactions. What's been your response to concerns about sending sediment pulses downstream, and particularly environmental concerns? George [00:42:35]: Well, the United States is mixing it right there, one sediment in the rivers, and I think that's where Matt Kondoff is also doing a lot of work. And you yourself is to try and educate people to say, you know, to keep this river healthy, you have to have sediment in it. Yes. Now, that is where Colombia is also a step ahead of us, because they don't have these tmdls. Every river is on its own. And like I said, this reservoir, we're working on the whole argument, okay, we need to get enough sediment downstream to keep the river rivers geometrically happy. And again, it's one of those mindset changes that is going to take years to change, to get regulators to understand that sediment is not a pollutant unless it is. I mean, you get polluted sediment, but on average, it's not a polluted, polluting element. You need it in your rivers. Stanford [00:43:30]: This podcast is a product of the core's regional sediment management program. And one of the big ideas of regional sediment management is to treat sediment as a resource, and that if you kind of hoard it upstream of the dam, there are infrastructure and environmental consequences to downstream communities, both human and non human. George [00:43:50]: Yeah, yeah, yeah, absolutely. It also is one of Matt Connor's ideas is that he had some experiments. We take sediment out of the reservoir and place it downstream of the dam, and when a flood comes, it just takes it down. Let's go to the Mekong game. That is the big concern in the Mekong river. Yes, because all those dams, once they built, 95% of the sediment will be gone. The Mekong delta is going to degrade, and that is the rice bowl of Southeast Asia. Stanford [00:44:15]: Let's actually talk about Mekong because that is what I want to talk about next. So the Mekong rivers, it's a system that you and I have both worked on. I've been more upstream, but it's one of the big biodiversity hotspots in the world, and there are a number of large and dozens and dozens of small dams. George [00:44:32]: 300. Stanford [00:44:32]: Yeah, 300 dams that are planning to go in place where Matt Condolf has done a lot of great work. But you were part of a team that got involved and did one of the most remarkable pieces of work I've seen with Sambord Dam. Can you just tell us the story of what was Sambord dam going to be and what was your proposal? George [00:44:53]: Yeah, you know, this is an interesting story. The original Sambo dam design was just a straight, long embankment, very high, with a covered water. I think it's like 100 km upstream. Stanford [00:45:07]: And where was it? George [00:45:08]: If you go in the Mekong river, the last proposed dam in the Mekong river is in Cambodia, so most of. Stanford [00:45:14]: Them are upstream and Laos and China, but this is the downstream most. George [00:45:19]: That is the most downstream. And if you built that, it was going to take out a lot of sediment. Stanford [00:45:24]: It doesn't matter what you do with sediment management. Upstream or the downstream most dampen. George [00:45:29]: Exactly. Yeah. So what we did there was actually, it's interesting. If you look at that part of the river where Sambur dam is from, in a, from an aerial point of view, it looks like a braided river. But then if you go in a boat and you go up there, you see it's not alluvium, it is braids in rock. So what we then said is, okay, well, let's see if we can use this as an advantage. And what we did is I designed a concrete dam with a lot of low level gates. If you look from the upstream to downstream, on the right side, there's a big braid going out. And then the powerhouse was in a braided next to it, and then there's another big braid that ran down to the side. And what we had to do is two things, we had to pass through as much sediment as possible. And secondly, we had to pass fish. Stanford [00:46:21]: And this is important because there's many migratory fish. George [00:46:24]: Ooh, more than 300 species. I don't even know if they really know how many species there are. Stanford [00:46:29]: And an enormous amount of Cambodia's protein is fishing. George [00:46:33]: Oh, absolutely, absolutely. That is a prime protein there. Yeah. So what we did is then we have the new Sambor dam. In the one braid we have a little, another small braid here. We had the power station and then connected this large braid with a 300 meters wide channel downstream. And that was intended for fish passage. So I mean, it's like a natural fish bass. It's not a step thing or whatever. It so happened that the river on the left side is slightly higher than the raids on the right. So I could just do it naturally like that. Stanford [00:47:09]: So you had an. In a branching channel with competent geology. George [00:47:14]: Yeah. Stanford [00:47:14]: And you dedicated two of the channels to hydropower and one to sediment transport and fish passage. And so in your book you talk about how we used to cite dams starting with where can I put this dam to get maximum power? But that maybe we should start with a different premise. George [00:47:32]: Yes, yes. And that is the premise we use there because, I mean, the dam that was originally designed was to maximize power. This one could generate two thirds of that power. And an economic analysis that we did, it was still economical. So you don't have to go to this maximum extent. And that's the same. You get that all over. In Zambia, also in the Zambezi river, there's another dam they wish to build upstream of Kariba. It's called Patoka and the same thing, they just try to maximize the energy. But there's a lot of reasons why it should lower the dam for tourists to use upstream and things like that. So that is a problem in terms of optimizing. You have to look wider than just the money. And that's where I think whether it's economic analysis or financial analysis, the difference between those two becomes important because I think a lot of people don't understand that, you know. Stanford [00:48:26]: Okay, so I want to wade into that. I just have to tell you that I'm an economics fan, but I feel like I'm going to have to ask you to define and repeat because I feel like this gets very, very ethereal. Like you said, it's philosophy with some numbers. But let's just start there. What's the difference between the finance and the economics of a dam? George [00:48:49]: So what do you do is if you start a project, what you should do is you do an economic analysis. And economic analysis is about weighing value. I mean, you may have an old stingray out here and it's very old, but because you value it more, it's more than what you know, someone else would be. Okay, so the value is different to price. So in economic analysis, you look at value and it's not just dollars, it is value of the resource, but also value to the community, benefits for the community. This benefits to the community, environmental benefits. So it's a completely holistic view. You look at now when it comes to reservoir sedimentation management, often what will happen is perhaps say you use sluicing or perhaps flushing and you have to empty the reservoir and fill it again. You're losing some benefit. And if you just look at it in the short term, then it may not be economically viable. But if you look at the long term, what the benefits may be to future generations, then it can become economically viable. So the first thing is an economic analysis is that you look at value and not at prices. And secondly, you look at how you use your discount rate. Because all these economic analysis and finance techniques use a discount rate to discount value to the present. So you can compare things easily. But if you use an interest rate, say of 10%, then in 30 years, whatever happens after that is not important. And that's why an economic analysis, you should use a declining discount rate. And as it goes more into the future, discount rate becomes smaller and smaller. So those values or benefits and cost way in the future is also accounted for here. So that is the economic analysis. Stanford [00:50:47]: I'm going to need you to help me understand this because the discount rate is difficult for me to grasp. But it is very deep in the. George [00:50:54]: Economic analysis world comes down to how you value life. The theory is that I would rather have a dollar right now than give it to me in five years time. So it's how you value time rate of stuff. Right. Stanford [00:51:09]: And so that difference between would I rather have a dollar now or in the future is the discount of the future dollar. George [00:51:16]: That's right. That's right. Stanford [00:51:17]: When we build projects, when we think about future costs or benefits, we discount them with the same idea. George [00:51:26]: Same idea. Yeah. So we're saying is. What we're saying is, and that's where we get into conflict with the concept of sustainable development. We say is that I'm developing this thing and I want the maximum benefit right now and I don't care about the future. That's essentially what it is. But if you change your analysis technique and you increase your discount rate as you go on the diminishing discount rate, then you at least look at what happens to the future generations, to their cost and benefit. And is it worthwhile? Stanford [00:51:57]: Okay, so this seems to be related to the idea of renewable resources. And I know that when I sat in my fifth grade class, I was told there are renewable resources and non renewable resources. And the non renewable, they were like coal and gold and things you extract out of the ground. And the renewable resources are like wind and sun and water. We've already kind of deconstructed that a little bit. Is that, well, surface water, certainly not groundwater, but even surface water. I guess the big question is reservoir storage. Is reservoir storage a renewable resource? George [00:52:31]: Well, that's one of my. Another pity. Reservoir storage, in my view, has a dual nature. Stanford [00:52:44]: Okay. George [00:52:45]: Okay. And its nature is decided upon by the designer and the operator of the dam and reservoir. If you design a dam that has the ability to manage sediments so that you can extend the life of this facility as long as possible, preferably in perpetuity, it's possible to do that. Then you classify it as a renewable resource. But if you design a dam and a reservoir and you know it's going to fill with sediment, you're not doing anything about it, then you define it as a non renewable resource. It is a decision firstly by a designer and developer, secondly by the operator. Stanford [00:53:25]: What's the hoteling principle? George [00:53:27]: Essentially what hoteling says is that if we work with a non renewable resource, then the value of that resource should increase with the discount rate. So in other words, if I discount it, it remains constant. That's the essence of it. Stanford [00:53:45]: Let's talk about gold. So if we're going to talk about reservoir storage as if it's a non renewable resource, we think about gold. If I take the gold out of the ground now, or if my grandchildren take the gold out of the ground, it should have the same value to each of those individuals. George [00:54:05]: Yeah, at that time. Point in time, yeah. I think actually it's a good example of gold because if you see how the price of gold keeps on increasing. Stanford [00:54:14]: That's right. And as you made the case in your book, water is not a constant value. It's actually going to be more scarce and more value. George [00:54:21]: Yeah, yeah, yeah, exactly. Because it's going to be scarcer and because it's bigger demand, it's going to become more valuable. Stanford [00:54:27]: Well, George Annadale, thank you so much for being on the podcast. George [00:54:30]: It's absolutely my pleasure, Stan. Thank you very much for inviting me. I appreciate it. Stanford [00:54:38]: I really appreciate George taking the time. Stanford [00:54:40]: To introduce this series for us and to lay out the motivations and options for reservoir sediment management. Stanford [00:54:45]: Next week we'll delve into those options in more detail and frankly, I cannot think of a better guide to walk. Stanford [00:54:50]: Us through the options. Stanford [00:54:51]: Than next week's guest doctor, Greg Morris. If you are in any way adjacent to reservoir sediment processes. You will not want to miss that conversation. These are informal conversations. And the views expressed do not necessarily reflect the official positions. Of the US Army Corps of Engineers. Their partners, or the offices or centers of the guests or host. This project has been and continues to be funded by the Corps of Engineers Regional Sediment Management program. Dave Perkey leads that program. And Tate McAllenhenne Alpin coordinates the inland contributors. Which is where the reservoir initiatives live. Mike Loretto is back editing the second season. And wrote our theme music. Stanford [00:55:22]: It's fun to be back. Thanks for listening.