Stanford [00:00:13]: More than 15 years ago, my agency gathered subject matter specialists in sediment transport to start discussions about how we should manage sediment at our reservoirs. I was definitely the kid in the room. Almost everyone else at that meeting has long retired. But when we asked the question who was working on or publishing on this problem, there was really only one answer. At that time, the literature on this topic was relatively easy to work through because it was a single text, a book we just called Morrison Fan. We passed that book around the table at that meeting and it influenced our agency discussion for the next decade. Today's guest is one of the authors of that book. Doctor Greg Morris wrote the reservoir sediment Handbook with his collaborator doctor fan in the late 1990s. And it wasnt just the first book in this space. It remains the seminal text on the topic and its still my first recommendation to anyone who wants to learn about reservoir sediment management. But in the 25 years since he wrote that book, Doctor Morris has just kept applying those principles in more and more settings and accumulating lessons learned around the world. And at this point has almost certainly worked on more reservoir sediment management projects than anyone else in the field. So when I first considered a reservoir centric mini season, I knew it only made sense if I could talk to Greg. So I was thrilled when he agreed to join us and share his experience. In this episode, Doctor Morris leads us on a deep dive into the reservoir management techniques under those three broad categories of keeping the sediment out of the reservoir, passing it through, or removing it after it deposits, categories he introduced in the handbook, plus an additional category that he's added to the paradigm. We'll delve into each of these approaches in some detail, getting into the processes, challenges and success criteria, and lots of case studies from around the world. I'm Stanford Gibson, the sediment specialist at the Corps of Engineers hydrologic Engineering center. And this week, on the second reservoir episode of the RSM River Mechanics podcast, a conversation with Doctor Greg Morris. Stanford [00:02:02]: Greg Morris, welcome to the podcast. Greg [00:02:04]: Good to be here. Stanford [00:02:06]: So how did you get into reservoir management? Greg [00:02:08]: I was working on a project in Puerto Rico. We were looking at long term management of water supplies, doing a 50 year plan and I was at that time a young engineer and thinking that, well, all the major issues in water are solved. And I realized that in Puerto Rico we had a really severe sedimentation problem in our reservoirs. And at one point in time I came to recognize that, wow, this is not just a Puerto Rico problem, this is a US, a global problem. And when I went to look for literature, there was no literature on it. Stanford [00:02:43]: What's the timeframe we're talking about? Greg [00:02:45]: This was 1980s. Stanford [00:02:46]: All right, so no one's talking about reservoir? Greg [00:02:49]: No, it's a completely unknown subject. There's nothing in the professional literature to speak of. There's a few papers in the gray literature. I wrote people, of course, this is back in the time of snail mail and whatnot, typing your letters out. And I wrote people all over the world and eventually found and was replied to by Professor Fan from Beijing. And he had published a small publication through UNESCO and talked about management techniques and chinese experience. And we eventually began collaborating. I went to China in 84, my first trip to China. Very interesting because they were extremely open at Samenjiae so that I could see the reservoir at a low level. They actually held the reservoir at a low level beyond their normal closure date, so I could get there to observe it. And they brought in other chinese engineers from the area and they had a workshop. You know, people were traveling two and three days by bus and train to get there. And it was, you know, it's a different world. It's a totally different world. Stanford [00:03:57]: I recognize the name fan because the seminal work in reservoir sedimentation is Morrison Fan. Greg [00:04:02]: Same fan, yes. Stanford [00:04:03]: About what time were you in China? Greg [00:04:05]: This was 84. Stanford [00:04:06]: So you do a lot of work in South America and Asia. How did that happen? Greg [00:04:11]: I was looking to do work in the States and there wasn't much interest in the states. Stanford [00:04:17]: We just weren't thinking about reserve settings. Greg [00:04:18]: And the problem in the States is not as bad as in the Andes and the Himalaya. Stanford [00:04:22]: Why? Greg [00:04:24]: We have less sediment load here. Stanford [00:04:26]: Our mountains are not as big. Greg [00:04:28]: No, it's not that, except the geology is better. Oh, okay. You know, you go hiking in yosemite. Granite. It's granite everywhere. You know, the Sierras are granite. And you go up into the Andes and you go at 13,000ft and you've got, you know, lake deposits. Yeah, a lot of sedimentary material. You think of the Himalaya, wow. You know, big granite mountains and it's not that way. I mean, yeah, up at 20,000ft, but at 15,000 it's sedimentary material and it's a lot of weak material and so sediment loads are very high. Stanford [00:05:02]: So you done quite a bit of work in Nepal. That's also my origin story for reservoir sedimentation. Before I went to grad school, I spent some time there. I remember the Kathmandu post, they had the very simplest schematic of a reservoir and it was full of sediment just in this very simple drawing. And that was the first time I thought, oh, my goodness, all of our reservoirs are going to fill sediment. So what is the goal of reservoir sediment management? Greg [00:05:28]: Well, we have constructed these walls and rivers. We call them dams, and they, of course, obstruct and store water and we release the water. But rivers also transport sediment and we store the sediment, but we don't release it. So when you continuously store something, you don't release it. You eventually fill up, and that's what's happening. So the idea is to balance the sediment inflow with a sediment outflow. And there's, of course, several different ways to do that. You can, number one, reduce the rate of inflow. It's never going to go to zero, but you can reduce it. And this is what a lot of watershed protection work is about. Some areas it works very well. Other areas it doesn't work at all. For instance, in China and Yellow river, they've been very successful. The loess plateau is highly, highly erodible. When I was visiting there, it's the most severe eroded landscape I have seen anywhere in the world. Stanford [00:06:29]: Can you just tell us a little bit about what loess is? Greg [00:06:32]: Loess is a silt, and there's a lot of loess in the midwestern United States. It's a silt which is typically windblown. So it's a very fine sediment, but it's very erodible. Your sediment starts from your smallest as clay, but clay is actually mineral, where you take a larger rock and you change the mineralogy. But silt, sand, gravel and boulders are just different sizes of the parent rock that's been aberrated down from boulder size or down to eventually silt. So clay is very, very small. And being very small, the electrostatic force is that, you know, like when you walk around and you in the wintertime and you touch a doorknob and you get shocked, that's an electrostatic force. And in clay they're so small that the electrostatic forces are thousands or a million times stronger than gravity. So they stick together, which is why you can take clay and model things with it. And, you know, as an artist or whatever, but silt is not cohesive. So because the gravitational forces are much smaller than electrostatic, because it's now a quote unquote large particle. So what happens with the silt is that it's very small, non cohesive and highly erodible. And having this hundreds of meters deep of this highly erodible soil, a somewhat semi arid environment, completely denier to vegetation. And the gullies are 100ft deep vertical walls. It's just amazing. Amazingly eroded landscape. Stanford [00:08:09]: And so that's the source material of the Yellow river. Greg [00:08:12]: That is a source material for yellow river. And in yellow river mainstem yellow river, they have measured sediment concentrations of 990,000 milligrams per liter. Stanford [00:08:25]: Wow. Greg [00:08:25]: Yeah. No longer flows as liquid. It's a newtonian flow. It looks like plastic running downstream. And that's where they built the Salman Shah reservoir on that particular location, which is why they had a lot of problems. And it turns out that that is probably the highest sediment loads of any river in the world. Their sediment loads at that location were on the order of 1.6 billion metric tons per year. Stanford [00:08:52]: Billion with a b, billion with a b. Greg [00:08:55]: Which is about more than three times load of the Mississippi pre dam because Mississippi has a lot of dams in it now. Stanford [00:09:04]: So its load is lower now. But the natural load of the Mississippi. Three times. Greg [00:09:08]: Yeah. The natural load of the Yellow river at Salman Jaeger is three times the total load of the Mississippi river. Stanford [00:09:13]: For a much smaller area, it's. Greg [00:09:16]: Yes, it's a smaller area, but it's incredibly erodible. Stanford [00:09:20]: So we're at the said Hyde conference. That's why I got to talk to you, because you're based in Puerto Rico. But we were co located and you gave one of the keynote addresses here at the conference and you talked about how a proposed conceptual approach to reservoir sustainability has three components. You want to tell us what those components are? Greg [00:09:37]: Yeah. Well, the first one is the reduction of the yield from the watershed, which is what we just talked about a minute ago. And in some places it works. And for instance, in China and the Los Plateau, it has worked very well because they've had revegetation, they've built a lot of czech dams, and of course the Chinese have a lot of labor and they did things in a very big chinese style. In contrast, where I work in Puerto Rico, it doesn't work at all because most of our sediment generation events are based on hurricanes and landslides. And we see a similar situation in Taiwan, another tropical area. In Puerto Rico we at least have the advantage of having a lot of clay soils, but in Taiwan they have a lot of siltstone. Stanford [00:10:28]: So help me understand that is the difference, that one is more episodic and the other is distributed. So you have more success with distributed? Greg [00:10:35]: Well, if you're going to do erosion control, you are typically working on the surface of the land, you're revegetating, et cetera. But if you have landslides and landslips and mass movements that are deeper than the rooting depth of the vegetation, then your vegetation is going to have some effect in terms of managing the soil moisture, etcetera, but it's not going to be able to control it. And that's what we've seen. We've seen massive reforestation just by natural in Puerto Rico since the 1940s, but we've not seen any change in the rate of reservoir sedimentation. When I started looking at the data, I thought for sure I'd see, wow, this is going to be a perfect case study of how reforestation has reduced the sediment yield. And we have reservoirs that have, like, 60, 70 years of data now. And lo and behold, there's no change. Stanford [00:11:31]: Oh, wow. Greg [00:11:32]: And it's all controlled by these big events. Hurricane Maria came through in 2017. We had 50,000 landslides. Stanford [00:11:41]: So that's one of the parts of your work that has influenced me, is I come from the classical geomorph tradition, where we think about bankful discharge as being the flow that carries the most sediment. And over time, it's your bankful discharge that you really have to think about. But you stress the importance of episodic events in some settings. Can you tell us a little bit about that? Greg [00:12:01]: Well, for instance, going back to Puerto Rico, where I've looked at the data more than anywhere, we have basically six, seven, eight events per decade. And these are one day events. Typically, because we have small watersheds, big storms come through, and we will get 30 inches of rain in a day, and those events produce six, seven, eight days per decade, 50% of your sediment load. And this, of course, is not generalized. But generally across the world, though, the largest events do account for a very large portion of your total sediment load. So that's why we can move to the second item in terms of how to manage sediment in reservoirs, is routing sediments. In other words, when you get a big event and you get all the sediment load, what are you going to do with it? You're going to put it into your reservoir again, going back to example of Hurricane Maria, we have a water supply reservoir that supplies San Juan, and we had enough water in one day to fill the reservoir ten times. And do you want all that water, plus all the sediment to go through your reservoir? Well, what you'd like to do is to flush that through and then refill the reservoir at the end of the vent with the clean water, because by the end of the vent, you've moved most of your sediment, and the end of the vent tends to have cleaner water than the first. And that's the process of sluicing. Of course, you have to have enough reservoir capacity. You can, you know, your storms are bigger than your reservoir, which many times is not the case. You could also empty the reservoir and just empty it completely, not during a storm and erode sediment. But that has a lot of downstream consequences. In some areas you can do it mostly high mountain areas, but it has a lot of downstream consequences that make it not generally applicable. You can take turbidity currents because sometimes in some reservoirs, the sediment laden flow comes in, creates a turbid current that runs across the bottom of the reservoir, and you can release it because, you know, the reservoirs are stratified. Typically they're temperature stratified. In temperate zones, they're seasonally stratified. And the tropics are stratified year round. But you get sunlight on top, so it's warm. And on the bottom you have cold water. And of course cold water's on the bottom cause it's more dense. Sediment gives a lot of density to water and it will run across the bottom of the reservoir and you can release it. Stanford [00:14:36]: This seems almost magical to me, that it's essentially another non newtonian fluid where you're getting the finer material and it's running under a density gradient, not necessarily a pressure gradient. Greg [00:14:46]: Well, not even a density gradient, because what it is is that the difference in density gives you two distinct fluids, and the density current is running by gravity. So it's a gravity driven current, but it has, of course, friction on bottom boundary, friction on top boundary and the bottom boundary, it can deposit sediment as it goes along, particularly if it's not fast enough to maintain the turbulence. And the top boundary, it can have clear water that will leak into the turbidity current and cause it to dissipate. So by losing sediment and by being diluted with clear water, the density current, if it doesn't keep going and is not continuously supplied with more water, it'll fade away. Stanford [00:15:32]: So you have these density currents that can, in some cases, reach the outlet of your reservoir? Greg [00:15:38]: Yes. Stanford [00:15:39]: What's the advantage of that? Greg [00:15:40]: If you release that water that has a sediment in it, you're releasing sediment. Stanford [00:15:45]: And because it's fine, you don't need a separate outlet for it. Can you just do it through the turbines? Greg [00:15:50]: You can run it through the turbines and many plants. Do you know, turbine engineers, mechanical engineers, they don't want to see any sediments. Stanford [00:15:58]: In their sediment at all. They want blue water, they want clear water. Greg [00:16:02]: And they don't like that. But the option is that if you fill your sediment with this fine sediment that you could run through a turbine. If you fill your reservoir with that, behind that comes a delta. Stanford [00:16:14]: Yes. Greg [00:16:15]: And the delta contains sand. And I guarantee you that if you don't like the fine sediment, you are going to hate sand, because the abrasion by sand is astronomical. It is really, really intense. Stanford [00:16:28]: So the idea is, if you can vent some of the finer sediment through these density currents, even though it's not ideal for your turbines, you're saving volume and time before the sands, which will really tear them up constantly. Greg [00:16:38]: Yes. I mean, there are plants in northern India that replace their runners two or three times a year because of abrasion. So, you know, if you want to get into a two and three time a year replacement cycle, don't worry about the sand. Stanford [00:16:54]: Those situations, they're not venting turbidity currents. They. Greg [00:16:57]: Well, their situation is that turbidity current is not going to solve problems. Stanford [00:17:03]: So the coarser material has already arrived. Greg [00:17:05]: Yeah. Stanford [00:17:06]: So one of the things that I've learned from you about turbidity currents is that their effectiveness for reservoir management can change with time. Greg [00:17:16]: Yes. When you manage a reservoir, you're really managing the geometry. And that geometry changes with time, because as you start to deposit sediment, the original geometry changes. Now, a turbidity current, just like a river, it flows down the channel. It's a gravity driven current, so it seeks the lowest level in the reservoir. And these turbidity currents are one of the reasons why reservoirs tend to fill from the bottom up. You don't typically get a continuous and constant level of sediment across the entire cross section. So your reservoir doesn't fill up and you still see the river channel. Once it's half full, those inferior levels, the deeper levels, fill up more quickly because the turbidity currents will deposit there. And when you do fill that up, the turbidity current no longer has a channel through which it can run. It's no longer a compact cone of sediment, and now it's spread out across the flat bottom, and it has a very large frictional boundary on the bottom and large boundary on the top where it can interact with the clear water. So instead of having a compact flow, now you have a very thin layer of flow and it'll dissipate. So that's why release of turbidity currents is something you can typically do early in the life of a reservoir. You don't have to draw it down. All you have to do is have the correct type of outlet works. And you don't even have to have a low level outlet because you can have like a turbidity siphon, and you can lift the turbid water from the top and just release it at a higher level. Stanford [00:18:54]: That's remarkable. Greg [00:18:55]: It's just like, it's like in a regular reservoir, you have multiple level outlets, and you can select your deeper water, your shallow water, so you just select the deep water that has turbidity current and release it. Stanford [00:19:07]: And so you're kind of managing the relic geometry of the historical channel to. Greg [00:19:12]: The extent that you can, because what you want to do is to eventually reach the point where you have a sediment inflow and a sediment outflow that are balanced, and you want to have a geometry that gives you your storage, that you want within the operational pool level. Stanford [00:19:30]: Do you have an example of a system you've worked on where venting turbidity currents has worked really well? Greg [00:19:36]: I have not worked with a system where we have been able to implement that. We've worked in the reservoirs on the Vasque river in Tajikistan where it would work, but we couldn't convince the client to do it. Stanford [00:19:50]: I see. But that would be a case where if you built a siphon, which working on the us side, where to do any of this, we have to retrofit dams. It gets expensive because we're interacting with the dam safety people. They don't want us to touch the dam. But a siphon, it seems, sounds extra. Greg [00:20:04]: Consider this way. A siphon is only a flow guide. It doesn't have to be a hydrostatic structure like the intake intake on a dam. You have gates that you close it. The siphon, you can have just an opening at the bottom and just guide the flow to your outlet. So it's much simpler in that respect. Stanford [00:20:27]: So we move by sluicing pretty quickly. Let's go back to that a little bit first. Can you just state again the difference between sluicing and flushing? Greg [00:20:35]: Well, flushing is basically when you empty the reservoir, not necessarily coincident with a flood event. In many reservoirs where they do this, they release through a bottom outlet that wasn't necessarily designed for this in the first place. But sluicing, you, by definition, have to release large floods. So you need large capacity gates, which are either low level outlets of large capacity, which is what we're designing in a lot of hydro sites in the Himalaya are being designed like that. Now, you have a hundred meter tall dam, and near the bottom you have these large outlets that can release your floods. And the other option, of course, is to have a deep radial gate. Your depth is limited. I think the tallest one I've seen is 22 and a half meters at Ziberi on the Mekong delta. But if you're going more than 20 meters deep, then you have to go to a bottom outlet. Stanford [00:21:37]: And so can you tell us a story of a place where sluicing is working? Greg [00:21:42]: Sluicing is working, for instance, at Salman ja in the Yellow river, which we began talking about, heaviest sediment load of any reservoir in the world. And there they do the drawdown sluicing during the season, they have a flood season that is late summer, early fall. And so the first part of that season, they'll keep the reservoir basically empty and pass the floods through, and then they'll close the gates and fill the reservoir with the clear water that comes at the end of the flood season. It's not, obviously totally clear, but much lower sediment load. And they do this on an annual basis. That's the same procedure they use at three gorges, which, of course, is the largest hydro project in the world. I think it's got 22,000 separate powerhouses and 600 kilometer long reservoir. So, yeah, it works in Nepal, Kaligandaki small project, five kilometer long reservoir, only like 40 meters tall, and it works. Stanford [00:22:47]: And it sounds like you're utilizing the natural hysteresis in the sediment load, which is something we've talked about a little bit on this podcast, where the. The rising limb of the hydrograph tends to have more sediment for the same flow than the falling limb. Greg [00:23:00]: Yes. The whole idea of sluicing is that you want to capture as much of the clear water as possible and exclude the dirty water. And one of the other strategies that you can do is using a bypass tunnel, where you have a tunnel that can actually bypass your sediment laden flows around the storage pool. Some of the smaller hydro sites in Nepal were designing for a combination of flushing plus bypass, because that reduces the frequency of flushing. And you don't want to flush too frequently, because if you're flushing, you're not generating power, and you're going to flush during the monsoon, when the plant would normally be operating full power. And if you're generating a million dollars of power every day, you obviously want to reduce the number of days of flushing. You can also put the reservoir outside of the river, an off stream reservoir, and we have designed and built a couple of those in Puerto Rico, and we basically are able to reduce the sediment load in the reservoir 90% to 95%. Stanford [00:24:10]: Oh wow. Greg [00:24:11]: So we only get less than 10% of the load of a conventional dam. So the life that this reservoir would have normally would be 100 years. And if you do this other strategy, by putting it off channel, your 100 years becomes well over 1000 years. And then because you have sized the reservoir only for your useful pool, you don't have a dead storage pool because you don't have to store sediment. You can come back and dredge this thing, a million cubic meters every 500 years. I mean, we can figure that one out. I think the Luisa reservoir I mentioned for the water supplies for San Juan, we're doing a dredging project right now, 2 million. It will recover capacity equivalent to about seven years of sedimentation. Stanford [00:25:08]: That's expensive water. Greg [00:25:09]: You cannot count on dredging to hold and recover reservoir capacity, you have to combine methods. Which brings us to the third means, which is to remove sediment. And flushing is basically a sediment removal process, whereas sluicing is a sediment pass through process. Stanford [00:25:30]: Let me just review the three. Because we've been talking linearly here and we talked a little bit to George Annandale about this. And he credited you with this kind of conceptual model, which has been my conceptual model for years because I learned most of it from your book, is you can keep the sediment out of the reservoir by managing the watershed you can pass it through during the flood that brings it or through or around or around. Excellent. And that includes sluicing and turbidity currents and bypass. Yes, or if it gets in there. Now we have these other methods that you can use to remove it once. Greg [00:26:03]: It'S already deposited, which is basically two methods. Stanford [00:26:05]: Two methods. Greg [00:26:06]: You can use hydraulic method, which you can flush it out, empty it and let it erode. The limitation there, of course, is that if you look at a river, river has a certain width and you build a dam and then you start flushing and the river is going to go back to its original width. So if you're in the high mountains and your reservoir is very narrow, yeah, you can clean the whole thing out. But if you're not in the high mountains and you have a wide reservoir, the part you're going to clean out is going to be the size of the original river. And all the stuff on the side is not going to get cleaned out. But the interesting thing is that when you do create this flushing channel, a large part of your inflowing load will get concentrated into that channel. And there is a lot of work done at the Cachi reservoir in Costa Rica some years ago, where they did very detailed documentation, and they do annual flushing there. And with the flushing channel, they maintain the turbidity currents flowing down. Their turbidity currents pass through the turbines and these areas that are outside of the flushing channel, very low sedimentation rate. So the flushing channel, it's much more than just, I'm going to excavate this v shaped thing in my reservoir. It's that you are creating a geometry which actually tracks the sediment to it because your turbidity currents will flow into that. And so this has a dual purpose in that, yes, you can recover some capacity, but it helps you pass sediments downstream in the future. Stanford [00:27:44]: So, for example, if you look at your flushing geometry and it's only 30% of your historic sediment, it might be a much larger percentage of your future sediment because your geometry is funneling the sediment into that path that now will be removed in future flushing events. Greg [00:28:01]: Correct. Stanford [00:28:02]: So you've mentioned a couple of times that there are some downsides to flushing and is not a universal solution. What are some of those downsides? Greg [00:28:10]: Big downside is you're trying to release, let's say you flush every year. So you're trying to release a year's worth of sediment in three days. So if you just open the gates and let it go, you're going to see 400,000 milligrams per liter. Stanford [00:28:26]: Oh, wow. Greg [00:28:26]: We've seen this in a number of. Stanford [00:28:28]: Sites around the world, which is, again, 40% solids. That's approaching a non newtonian fluid. Greg [00:28:32]: Yeah, it is. Newtonians, typically in China at least, they're about 350 milligrams per liter. 350,000 and above, they get newtonian. Stanford [00:28:42]: Okay. Greg [00:28:42]: So what happens, though, is that, as we mentioned, that in this flushing channel, you're concentrating your deposits. So when you open the gate, you have this flushing channel that's full of poorly consolidated, recently deposited a lot of fine sediment. So you open the gate and this mud just comes out. So one of the things that you need to do to mitigate downstream impacts is a couple different things. You need to control the rate of release, and not only initially, but during the release, because you can open and close the gate, you gradually vary the gate of. And as the water level goes down, you're releasing more sediment. But if you're releasing too much, concentration is too high, close the gate a little bit and it will not be releasing as much sediment. If you have an adjacent gate a little bit higher that you can release clear water to mix with it. That also helps. And also you would like to flush not once a year, but several times a year. Because if I'm going to release, let's say, 4 million tons of sediment a year, which in the himalayas, like lots of places, have more than that. Stanford [00:29:59]: More than that. Greg [00:29:59]: But if you want to release 4 million tons a year, it's better to release it four pulses of 1 million each rather than one big one. Stanford [00:30:09]: So you're describing something that we've been thinking about but I've never heard anyone say out loud, is that flushing works better as an adaptive, managed solution where you have a tight feedback between observations of downstream concentration and your operations. Greg [00:30:27]: Yes. And initially, when you start flushing, you're going to have to dig a channel. So there's going to be a lot more sediment release. But over time you can bring the system into, let's call it a more stable situation. It's like learning to drive a car. You're 16, you get in the car and you learn to drive and you don't know what you're doing, but after a little while you figure it out and then you're okay. And your grandparents are no longer frightened to be in the car with you. But it's the same way. There's a learning process. You have to understand how things work and there will always be some sort of problem that's going to come up. But it's like a medical procedure. You don't go. Any medical procedure has no risk. I mean, you go and you drink a Coca Cola and it has sugar in it and there's risk, and you cross the street and there's risk. You know, you have these risks and you're always weighing risk. But here we have the risk that. And it's not just a risk. I mean, we know that the reservoir is going to go, we know we're going to lose it. So we're having to find the path by which we can minimize our adverse impacts, yet maintain the benefits on which we depend. I mean, think about water. How much water do you use? I mean, everybody thinks, well, we use a lot of gasoline for our cars, but compare the volume of gasoline to the volume of water on the average, like an american house, your water use is enough to fill your house completely in a year. Stanford [00:32:02]: Wow. Greg [00:32:02]: I mean, do the numbers. Yeah, it's a lot. You know, you look at how much gasoline you use in a year, which is the liquid that we probably think of using the most, and it's peanuts compared to water. And then you turn around, you're talking about, yeah, well, how much water do we use for irrigation of crops that we eat and how much water do we use for manufacturing processes? And all of a sudden, we live in a hydraulic society. I mean, hydraulics is the basis of modern society. Stanford [00:32:34]: So you described a situation where it sounds like if you're gonna flush the reservoir, the first flush has by far the most pain and by far the most cost and is gonna be the least popular, and then the next few, you're kind of dialing it in. But you can get to a situation where the impacts are lower. But it would seem to me like the sooner you do that first flush, the less pain there's gonna be. Greg [00:32:59]: Yes. And this also brings us back to sluicing, because sluicing doesn't have those big impacts, because sluicing events naturally have a lot of water. It's a lot of dilution. They naturally carry a heavy sediment load anyway. So you're releasing sediment downstream consistent with the natural cycle and at concentrations that aren't too much different from the natural cycle. Because downstream impacts, you have a couple different types of environmental or ecological impacts. One, you have the impact of being out of sequence. You know, fish spawn at certain times of year. And if you have spawning gravels or spawning areas, for instance, a salmon, a spawning gravel, you have to have water flowing through these gravels because that water brings the oxygen that keeps the eggs alive. And if you have a reservoir and you flush out of sync and you put a lot of sediment onto the spawning beds with the eggs, then you can obstruct the flow of oxygen through the bed. Also the issue, you need to have occasional flushing flows that will actually mobilize that bed to clean it out. So it's complex, but you can work with the system. You need to understand it. And this is why sediment management is very multidisciplinary. And engineers need to understand biology, ecology, not what species are and not, I know, what fish that is, but how the ecosystem works. And unfortunately, engineers never study aquatic ecology. It's very foreign to them. Stanford [00:34:40]: Yeah, we at least need a team member who has that skillset. Greg [00:34:44]: I guess I have a lot more sensitivity to that because I did study systems ecology as a master's degree, and I thought it was very interesting. My bachelor's was at Naval Academy. Stanford [00:34:54]: Oh, really? Greg [00:34:54]: And I got a medical discharge when I graduated. Didn't know it, but was discovered that I had hemophilia. Stanford [00:35:00]: Oh, wow. Greg [00:35:00]: Mild case. So I went from Naval Academy, University of Florida. And it's interesting because the same mathematics that we use for feedback systems and missile guidance is the same math that's used in systems ecology. Stanford [00:35:16]: Oh, interesting. Greg [00:35:17]: So that was a very, very interesting influence. So I got into systems ecology. It was pretty cool. Learned a lot. Stanford [00:35:24]: And so what was your schooling from there? Greg [00:35:26]: And then I went PhD. I was in environmental engineering. Okay, great for masters and PhD. And then when I got to Puerto Rico, I got interested in sediment. Stanford [00:35:35]: So we've talked about these three methods. Reduce the sediment, pass it through, remove it, and under removal. We've talked about. You said there are two options. We've talked about flushing. Greg [00:35:43]: Talking about flushing. And now we come to dredging. You know, we're doing a dredging project right now. We're working actually on a couple of them. One we're working with in Strontia Springs in Colorado and another one in Puerto Rico. And in some places, dredging can be a solution and it can be a long term sustainable solution. Surprisingly enough, at the Tarbella reservoir on the Indus river, sediment load of about 150 million tons per year. Huge amount of sediment. Dredging is the solution. Stanford [00:36:13]: Is it really? Greg [00:36:14]: Very surprising to me when I came to that conclusion, spent months over there working on that. And why does it work there? Because the other option they are considering is bypassing tunnels, ten tunnels, each of them 10 meters in diameter, to allow the reservoir to be drawn down for an annual flushing event. And this takes a 4800 megawatt power plant offline for almost 30 days. Stanford [00:36:46]: Oh, wow. How often? Greg [00:36:48]: Every year. Stanford [00:36:48]: Every year. Greg [00:36:49]: And their power production cost is less than one cent a kilowatt. Stanford [00:36:54]: And this is in Pakistan. Greg [00:36:55]: Pakistan. Because its project is paid for, the replacement cost is close to $0.12 or more per kilowatt. Stanford [00:37:03]: More than order of magnitude. Greg [00:37:05]: It's more than order of magnitude. So when you start looking at, I'm gonna pay twelve cents a kilowatt every day, I have to replace this energy from somewhere else. It looks like you've got four to $500 million a year you can spend on dredging. And with $500 million a year, you can do a lot of dredging. And the sediments at that reservoir are ideal for dredging. It's silt, sand, fine sands, coarse to medium silts, and it's perfect for dredging. So that particular case, because of that situation, it works. And also on the Indus river, you have the problem of moving the sediment downstream. Now, this is another problem that people don't really think about. Erosion is a geologic process. The eroded material gets into the rivers and moves downstream and goes to the ocean. And that's the way geology has worked for a billion years. Stanford [00:37:59]: As long as geology has existed. Greg [00:38:01]: We're not going to stop it. And you're going to put these little tiny dams, you know, we think they're big, but, you know, in geological context, they don't even count and geologies can overwhelm them and the sediment continuity from upstream to the ocean will be returned to its original condition because there's no way we can stop it. We can control watershed erosion, which is basically trying to undo the damage we've already done to the watersheds. So we can undo some of the damage that our poor land use practices have done. But we're still going to have a lot of sediment. And in the Himalaya you have a lot of background natural sediment yield that's very high because glaciation, very little vegetation. The mountains are continuing to grow. You know, it's very active tectonically. So you have this reservoir that you really need to have the reservoir to supply irrigation because there's like 80 million people who farm on the end this floodplain and depend on this for irrigation. Stanford [00:39:13]: And travella. Greg [00:39:15]: Yeah, yeah. So, you know, you look at the benefit of 80 million people and this big reservoir and it's filling with sediment, you have to do something and the sediment's going to go downstream. How are you going to get it downstream? Stanford [00:39:27]: One of the ways that I say is the mountain will win. Greg [00:39:30]: The mountain will win. Stanford [00:39:31]: And so we have to minimize our footprint on the way to the ocean. Greg [00:39:34]: Yeah. And the problem in the Indus river is that all of their infrastructure has been set up on the basis of all the sediment is trapped. So that's a non sustainable solution. I work in Pakistan quite a bit and it's a continuing topic of discussion. But it's like in this river to Pakistan, it's like the Nile river is to Egypt. The Nile river in Egypt, they don't have as much of a problem in the near term because the Aswan dam can trap several hundred years of sediment and now they have upstream dams going in, which is creating other issues. The Egypt issue is not going to be sediment, it's going to be water availability. Stanford [00:40:15]: So as we mentioned, you co wrote this reservoir sedimentation handbook. Yeah, this is my first introduction to these ideas, like maybe 15 years ago. There's a PDF of that online. Greg [00:40:24]: People can go to reservoir sedimentation.com and you can get that and you can get several other publications. Stanford [00:40:30]: Fantastic resources. But in your keynote address, you said that over the years people have come up to you and told you what are the big things they've gotten out of that book? And one of the surprising things was how diverse that was. But it left me thinking, yeah, but are there some trends? Are there some things that people. Greg [00:40:45]: No trends. Stanford [00:40:46]: No trends. Greg [00:40:46]: No trends. It's like after the keynote, I had several people come up and say, wow, I really like the way you talked about the rating curves or about elephant butte or whatever. Everybody got something different. Stanford [00:41:00]: So maybe let's flip the question and say, what are some of the big things that you would like for people to get out of that book? If you were to say, here are three or four big ideas that someone should get out of a text like that? Greg [00:41:13]: The big idea is that we are basically taking for granted a resource system, you know, infrastructure system, I should say, that's not sustainable. And when you really become aware at the level of the general public and politicians that it's a real issue, it's going to maybe be too late, or maybe not too late, but very expensive to solve. Stanford [00:41:45]: The longer we wait, the more costly it'll be. Greg [00:41:47]: Exactly. And the second item is that there are a lot of solutions. And we've talked about reducing yield from the watersheds. We've talked about passing sediment around or through the storage. We've talked about removing sediment. And there's a fourth item that we need to mention, and that is what we might call adaptive strategies. For instance, in Puerto Rico, we have this reservoir that's losing capacity. It's being overdrafted. It has a firm yield of 60 some odd million gallons a day, and they're withdrawing 80 to 90. We have a drought, and you turn the water off and you open the faucet and there's no water. So what do you do? Well, here we have a solution that we can use a large well field in the north coast, aquifer, limestone aquifer, and use this on an intermittent or standby basis. So when the reservoir level drops to a certain level, you turn your pumps on and thereby not only can you maintain the supply, but you can also augment what's available from the reservoir. So this conjunctive use strategy, using a surface water and a groundwater resource, allows you to overcome many types of problems. This is being used increasingly around the world. Part of what they're doing, for instance, in California, is allowing or encouraging farmers or whoever has excess water during a wet year, put it underground. So you know, you have an underground storage reservoir and you have an above ground reservoir. So you can manage both reservoirs in a conjunctive manner. And that's, you know, one method that you can, you can use and a. Stanford [00:43:32]: Nice thing about underground reservoirs is they don't fill with sediment. Greg [00:43:35]: They don't fill with sediment. Stanford [00:43:36]: They're already full. Greg [00:43:37]: There are a lot of different strategies. But don't look at the sedimentation problem as, oh, this is a reservoir problem, it's a water supply problem. It's a flood control problem because you use that storage for water supply, flood control for hydropower. Hydropower is interesting because I've worked with lots of hydro projects in Colombia and Columbia, is 70% supplied by hydropower. Stanford [00:44:01]: Oh, wow. Greg [00:44:02]: So it's a very important resource there. And what happens is that if you have a large storage reservoir and you lose that capacity, you can still retain a small reservoir pool by sluicing and the other methods, and you can maintain that forever, but you can still capture most of the hydropower energy. It's just that it's going to be time shifted. You're going to have instead of a year round supply that you could deliver at almost constant rate. Now you're going to be more into the wet season. I see, which is a bit of a problem because if all your power plants are producing in the wet season, but then, you know, maybe we can offset with photovoltaics. Because in the wet season it's cloudy. Stanford [00:44:51]: That's right. Greg [00:44:51]: In the dry season it's not so cloudy. Stanford [00:44:53]: And in that case, you don't have a reservoir problem. It's an electricity problem. Greg [00:44:57]: Yes. Comparable to water. You look at how everything fits together and then this hydro reservoir that doesn't have so much capacity is still highly useful. In fact, essential as a power peaking or as intermittent supply during the night when the photovoltaics obviously aren't working because it's nighttime. Stanford [00:45:20]: Let's pause on this point a little bit. The difference is that for flood risk management and water supply you need volume, but for power generation you'd need head and some volumes. Yeah, eventually you need volume to have head. Greg [00:45:35]: That's true, because once you have volume where you can do delivery during peak hours, that volume becomes very expensive because the difference in energy price between your baseload and peak hour, it's typically about double. Stanford [00:45:50]: So you want it to be time shifted and you need volume to time shift it. Greg [00:45:53]: Yes. Yeah. Stanford [00:45:55]: I think it's pretty fair to say that you have worked on sediment management and more reservoirs around the world than anyone else. We haven't talked about any in the United States. What has been your experience with reservoir sediment management in the lower 48? Greg [00:46:12]: I guess nascent would be the word. You know, I've worked with the corps and with the Missouri sedimentation Coalition on Gavin's point, looked at a couple of reservoirs in the Hells Canyon area and did a project with a corps of engineers. Very interesting. They have like 380 reservoirs. And we got a publication that finally was published this year on that. But we looked at all the core reservoirs and, you know, looked at key reservoirs in different parts of the country. And it's interesting because the corps was not looking at that at all at the time. And the first thing you need to do is you need to put your data together. So they're starting to do that. Stanford [00:46:56]: Yeah. In the next episode, we'll talk to Paul Boyd a little bit about that. Greg [00:46:59]: Yeah. You know, I met Paul back maybe 20 years ago when we were looking at Gavin's point and Spencer Dam, which is no longer there. Stanford [00:47:08]: That's right. That's also how I came to know. Greg [00:47:09]: Paul because Spencer was really interesting. It was a small dam, but I. But it's intermediate between the size of having a laboratory flume and having a big reservoir. Stanford [00:47:20]: That's right. Greg [00:47:21]: Like Gavin's point and Lewis and Clark Lake, which had Spencer, which is very small. And you could observe and see and monitor what happens at a scale that's a lot larger than your hydraulic laboratory. Stanford [00:47:34]: So I mentioned where it said hide. So I'd like to introduce a new podcast segment, which is a q and a segment because I taught a class in reservoir sediment modeling, the front end of this, and I asked if anyone had any questions for Doctor Greg Morris. Would you be game for a couple of those? Yeah. Okay. The first one is what are the most effective reservoir sediment management techniques that you've seen implemented around the world? Greg [00:48:00]: The most effective is offstream reservoir. Without a doubt. It doesn't completely resolve the problem, but it's good. The other one that's highly effective is sluicing. Stanford [00:48:11]: Okay. Greg [00:48:12]: If you have an off stream reservoir, eventually you're going to have to dredge. The volumes are going to be very much reduced sluicing. If it's done right, you can probably not ever dredge. Or if you have, for instance, coarse grain sediment, you know, gravels and cobbles coming in, go upstream and have a commercial sand and gravel extraction operation. And that, that will typically work. And that's used at a number of sites around the world. Those I think are the most. Two most effective. Obviously dredging is quote unquote effective, but it's also very, very expensive and that makes it ineffective for most purposes. It's effective but not sustainable. It's just economically too much. Stanford [00:48:54]: Then the second question, as dam projects become larger in scale, for example, Kyrgyzstan, do you have concerns about managing larger and larger volumes sediment? Is there a high end of like the volume of sediment that we're going to be able to manage? Greg [00:49:09]: Well, as I gave the example in Salman ja, you know, they were managing about a billion tons a year. Yeah. The problem is not the size of the reservoir. The problem is the size of the reservoir in relationship to the inflow, your capacity to inflow ratio. For instance, the little reservoir we're working with in Puerto Rico, it's not little little, but it's, it's certainly orders of magnitude smaller than three gorges. In China. They're the same size from the hydrologic standpoint. Stanford [00:49:41]: Okay. Help me understand that. What do you mean? Greg [00:49:42]: So in Puerto Rico, I have a 500 square kilometer, which is like what, 100 5200 square mile watershed. And in China they have like almost a million square kilometer watershed. So you scale your reservoir in relationship to the amount of water you have out of your watershed. Stanford [00:50:02]: So they scale to the same size. Greg [00:50:05]: Yeah. Because what's important is that if you want to sluice sediment through the reservoir with a flood and you have a reservoir that will capture about, let's say 75% of the mean annual runoff. Now, your runoff varies from year to year. 75% means that some years you're not going to fill that reservoir. If you get two or three years of drought, it's not going to be full. So you will rarely ever have enough water that you can empty the reservoir flush, setting it through and then refill it. Maria, we had ten times the volume in one day and three more times the volume in the next two days. Stanford [00:50:48]: Wow. Greg [00:50:48]: So I had 13 times the volume of the reservoir. So obviously open the gates and let it go. It's easy. But if you have a flood that can barely fill the reservoir, you can't do anything. But remember, sediment will take your big reservoir and will make it small over time. So when you plan for the future, you need to plan that today we're going to do this. Maybe we'll start releasing a turbidity current. And then 50 years from now we're going to do something else. Another 50 years, we're going to do something else. And these may be additive. Stanford [00:51:26]: So sediment size is not a static property and your sediment management technique changes as it changes, changes, or you can. Greg [00:51:34]: Add to them and can be more aggressive. So whereas at one volume you would be doing a sediment sluicing event once every five years, maybe down the road, you doing it five times every five years. Stanford [00:51:50]: Okay, last question. This one's a little cheeky. Greg [00:51:53]: Okay. Stanford [00:51:54]: Do you have any advice for a young sediment geek like me to grow up to be like you? This is just the question. How do you become Greg Morris? Question. Greg [00:52:05]: I would say two things. Stanford [00:52:06]: Okay. Greg [00:52:06]: Read a lot. I use Zotero, which is a free bibliographic software. I think lots of universities use it. And I'm always taking stuff off the web. I mean, research today is so easy, you know, you don't have to get on an airplane and go to Georgia tech and spend three days in the stacks with a nickel xerox machine, which is what I did. Stanford [00:52:30]: I'm just old enough to have caught the tail end of that. Greg [00:52:32]: Yeah. So research is so easy. So you go and read what's out there. There's a lot of stuff out there. And number two, go to the field and observe. Be observant. I mean, I live in Puerto Rico and I live on the beach. So I would go to the beach and I would run the beach every weekend. And there's a place at the end of the beach has a lot of rocks. Rocks and sand. And I would watch and play with the sand and observe things. Get into a kayak. I kayak. Get into a river and a kayak. That will help you understand eddies. And if you're designing an intake, for instance, you know, intakes, you don't want them to bring sediment into the intake. You observe things. You get in your kayak and you start understanding things, you know. And I teach a class and I say, well, you know, I can go out in the middle of the river and hide behind a rock and in my boat and cross my arms and the paddle just sitting there and I'm not going to move. And the river's going to go by me. And people say what? I said, yeah, get into a kayak. You know, look at these things and be observant, be curious. It's all a question of being curious. Stanford [00:53:40]: Greg Morris, thank you so much for being on the podcast. Greg [00:53:42]: It's been a pleasure and hopefully it will be of some use to somebody. Stanford [00:53:49]: I appreciate Greg taking time during a busy said hide week to record this episode and share his wisdom. I know I found this conversation helpful and I'm confident others will to. I put links to Greg's website with the free PDF of Morrison Fans Reservoir Sedimentation Handbook on the podcast website, which is linked in the program notes. We're also going to start running some video clips from these interviews and some footage from some of these reservoir operations. And we'll start running those on the HEC sediment YouTube channel. And the podcast website. So that'll be worth checking out in the coming weeks. Next week, we're going to delve into what Greg called. The nascent realm of reservoir sediment management in the United States. And while it's true that we face some specific challenges. Because of our unique sediment history here. We're actually using most of these methods. And some innovative new technology. To manage sediment at several of our reservoirs. So I'm going to talk to the two practitioners. That are connected to more of these us projects. Than anyone else I know, Doctor John Shelley and Doctor Paul Boyd. These are informal conversations. And the views expressed do not necessarily reflect the official position. Of the US Army Corps of Engineers. Or the offices or center of the guests or host. The RSM in the podcast title stands for regional sediment management. Because this project is funded by the Corps of Engineers Regional Sediment Management R and D program. Mike Loretto edited this episode and wrote the music for the season. We'll be back in two weeks. Thanks for listening.