Episode 29: 1D vs 2D Restoration Modeling

[Music] all right welcome to full momentum and HC Raz vcast I am your host Ben KY and joining me today uh Chris Goodell as always and a very special guest Robert Chomp welcome to episode 29 Robert of the HC Raz podcast thank you great to have you we're really excited to talk some Raz today um but before we get into some technical discussions uh I do want to just let everybody know that's out there listening that um this this week has been very hard for me and my family because uh I secured last place in the company bracket challenge of for March Madness and um as somebody who watches way too much college basketball and pretends to know a little bit about college basketball it was uh it was a tough week Chris and I know you you had some success so Ben is the uh Ben is the uh he's the manager of our office bra pool and uh so he gets this thing rolling every year and hey despite your last place finish we all still think you know the most about college basketball in in our in our group in the company but uh yeah I had a I had a period in the middle there where I was like oh I think I'm I could win this there was I was bumping around first or second place and as always happens you get your hopes up and then you have a you know one of your big teams loses later in the conference tournament and um that's it and I think I ended up 15th or something like that we have the biggest problem is our company has way too many people in the Northeast so we had a lot of Yukon fans in the in the bracket challenge which makes it an uphill battle I should have known that when I was picking mine but anyway anyway it so I I want to come clean to everybody because again we do talk a decent amount about college basketball on this podcast and um please do never never listen to me about my picks because apparently that will leave you lead you down the wrong way I think that's I think that's a real thing though you can overanalyze when you do these and um and you may be very correct in picking the team that should win but that doesn't always happen in March Madness as we all know well speaking of overanalyzing we're going to have an interesting discussion today on on um you know 1D and 2D River restoration modeling um so that's a good little segue into a preview of what what's to come later on in this episode um but before we get into that I know everybody's probably thinking hey I know who Chris is I know who Ben is who's this Robert guy so Robert again we're really happy to have you on the podcast uh if you want to start off by giving folks a little bit of a background info on on yourself and how you ended up on the vcast here today perfect I I really appreciate the opportunity uh I was nervous this morning to join the ranks of people like Grady Hillhouse and Stan Gibson uh so I'm really excited to share uh what I've learned from um doing my work here at kment especially as it relates to 1D hecr modeling of stream restoration um I'm Robert sha I'm a hydrology and hydraulics engineer with Klein Schmidt and I've been with Ken Schmidt for nearly three years now uh most of what I focus on is flood risk assessment I've been supporting some water rights adjudication projects which include surveying um and digitizing that data into 1D headass models um I've also been getting into sediment transport and learning all the ins and outs of doing sediment transport both in 1D and 2D so that's been uh a very unique Challenge and um it's been very fulfilling work uh to to kind of push the boundaries of what we know about sediment transport and what we can do with the results yeah we we purposefully had Stan on before you Robert so that we didn't put too much pressure on Stan to follow up a sediment transport Legend like you I appreciate it I appreciate it make the playing feel a little more even so that's a little bit about what you do here at kenm um yeah just out of curiosity how did you get into the field of engineering and Water Resources overall I know you got a lot of different interests out there so how did you kind of settle on that uh you know area of expertise yeah that's a very good question um over my my uh years in college i i attended four different colleges um the last being University of Texas at Austin uh where I got a MERS in civil engineering with a focus on Water Resources environmental engineering but my passion for water and Water Resources engineering actually started at Oregon Tech as I was a new student in the renewable energy uh engineering program where I learned a lot about hydrop power how dams impact uh streams and rivers all across the US and I was just absolutely fascinated with that type of infrastructure because it's complex there's there's no optimal solution between all of the um interested parties for those projects so I just I find this type of work uh very fulfilling because we get to be creative and get to propose unique solutions that uh try to you know accompany all of the types of interests that people have for these projects very cool well I know speaking for myself and and probably Chris as well we've really enjoyed working with you ever since you've been in clench man and we're really excited for you to be able to Showcase your experience and knowledge to the the HC Raz World here on on the podcast so thanks for taking the time today yeah Robert Robert sits right next to me in the office and so uh we naturally have a lot of convers work rated and otherwise and uh Robert has uh probably the the cutest dog I've ever seen into the office whoa come on Chris besides besides ruy of course I mean that goes without saying but uh it's just you know first of mine so that's all I mean ruy was the cutest when when she was coming in every day but um Marge Robert's dog is uh probably is the only uh living being in our office that has more energy than Robert oh yeah I think yeah yeah yeah good stuff awesome well let's go ahead and get into some questions that we got from the audience over the last couple weeks on on LinkedIn and YouTube before we get into those I do want to uh give a quick shout out to our sponsor um we're thankful to be sponsored by our firm Klein Schmid Associates who's known throughout the industry indry is a firm that provides practical solutions to complex problems affecting energy water and the environment you can learn more at Klein schmi group.com thank you again to Klein schmi for supporting Chris and I and for supporting Robert to be on the podcast today awesome uh first question Chris that we got from the audience that we wanted to touch on was a kind of a follow-up to the 2D Bridge conversation that we had last month um I think we we covered a lot of ground there but um there was somebody who wanted to know a little bit more about how to view specific output through a 2d Bridge uh when you have a a 2d Bridge setup as opposed to a bridge setup in 1D uh as we all know there's the ability to get uh pretty quick instantaneous output data for 1D Bridges within RZ and it's a little bit trickier to do um with 2D so why don't you share with folks your recommendation in terms of how to do that yeah yes so I mean Bridges have been around in Raz for a very long time uh since before I used Raz they've had Bridges um and they go way back in even into the heck two days um however that that's always been 1D and when 2D came out they did not have Bridges as an option for 2D areas and so we had to get a little bit creative um now we do have bridges in 2D so it's it's really nice to have that feature but it's it's still improving there could be better ways to look at output and I think that's probably the Crux of this question is hey how do I look at my 2D Bridge output because I knew how to do it in 1D but and I can look at it in Ras mapper for 2D but I don't know how to look at the bridge and section View and see the water go up and down and things like that so let me first show um and talk about how you do it in 1D just so we can kind of compare and contrast here so this is a real simple model Beaver Creek Bridge model um you get this with the example data sets there's it's a pretty schematic looking reach and you've got a single Bridge down the middle and we can look at the input for that bridge in the bridge Culvert editor there's our bridge if I if I zoom in on it here you can see all the different peers that have been added in you can look at the bridge modeling approach as well and see what equations we're using for this particular Bridge all right so that's 1D and when you run it and you want to look at results well you can always go to the cross-section viewer which is right here this button and there you can see our bridge and we can animate the results of course I haven't run this yet so let me give it a quick run this should go pretty quick but once it's run then you're you're able to see the water surface elevation on the cross-section viewer plot itself which is really helpful when you want to know hey is my water getting high enough to touch the bridge or is it going over the top of the bridge am I using the right equation as a result so that's always a really um convenient and and effective way to interpret what's going on in your model the other thing you can do too is you can look at velocity distribution within a bridge and you can see hey is do I have velocities concentrated around the peers and that kind of thing that can be important for doing scouring analysis so if I open up the cross-section viewer now I can see my water level there I can in an unsteady flow model you can animate it and you can watch the water go up and down in your Bridge section there we can see it's over topping so it's uh not only going pressure flow but it's over topping as well um and it comes back down and so people were used to this comfortable looking at Bridges this way you can also look at the stage and flow hydrographs as well for a bridge so if we click on that button right there that pulls up stage and flow hydrograph and I can go to my type Bridges coverts and here I can see how the stage on both the Headwater and tail water side of the bridge changed throughout the simulation I can also see what the flow rate is through that bridge for the entire simulation you can pick out some important values like maximums right here in the table or just go right to the plot all right so everybody who's who's been doing Bridge modeling forever is really comfortable with this they understand how to look at bridges in 1D but now that we have bridges in 2D how do we look at Bridge results so let me open up um a 2d version that this is the Bald Eagle Creek data set uh so different Bridge here and this bridge happens to be in a 2d area so notice I'm using 65 um where did my Ras mapper go there it is okay so here's our bridge if I zoom out a little bit you can see more what's going on here there's another Bridge right there so there's several 2D bridges in this particular model run there's one there and there and another one there um this too is an example data set that you can get a hold of when you download the software so it's great you can see the bridge I can turn on my velocity map my velocity output I always like to look at that when I'm troubleshooting and evaluating my Model results and I can animate and this gives you a great view from a plan view perspective what's what's happening around the bridge but it doesn't tell you everything and you can't see the section view here so a lot of people ask that question how do I see the same plot that I had for 1D in the cross-section view or how do I see that here well if you go down to um and turn on your results here and make sure you turn on your geometry under your results plan okay not the geometry up here but turn it on under the results plan and turn on your sa 2D connection okay when you have that turned on now I can rightclick on it I'm going I'm going to check uh water surface as well I can right click on it and at the very bottom you see it says plot s2d con data all enabled results so these are my enabled results velocity and water surface I can plot it um of course they haven't set it up yet to look at velocity as far as I can tell so you're only going to get water surface for now but at least you have that so when I click on that it pulls open three plots it gives me a center line plot right through the middle of the bridge it gives me the Upstream face of the bridge and the downstream face of the bridge and I can I pull open Ras mapper again where where did it go oh there it is it's hiding there all right so we have the three windows open here um it's showing you Upstream Downstream and Center Line and I can animate these if I have the right things selected in here so if I go and look at Raz mapper uh I wish it would let me put these up front so let me minimize these real quick so you want to select what you would like to see now now they're not going to be able to show you velocity yet I think that's coming in a future episode or or if it's available now I haven't figured out how to use it but you can definitely look at Water surface elevation so I'll just turn off velocity but if you want to animate this in those section views you have to make sure that you select the plan name up here as the active layer not the water surface not the bridge here not the sa Duty connection but the plan name that will allow you to see the these three different plots going at the same time and so if I bring those back up okay here are my three plots this is my Downstream I'll put this over here because that feels like more of a downstream side and that's upstream and there's a center line now I can animate and we'll see these water levels go up and down okay with the simulation so in this particular example it doesn't hit the bridge deck but maybe if we ran it longer it would um but you can clearly see how it's interacting with the peers and the nice thing about 2D obviously over 1D is you can have variation in the water surface going across here which you can clearly see uh you wouldn't get that with 1D um especially when you have peers and and those kinds of things in there they can have a big impact on the lateral water surface elevation and velocity distribution as well so that's a quick way to take a look at your Bridge output 2D Bridge output in section views what do you guys think yeah I think that's that's really helpful um I think it's a good tool especially the animation part of it because it's a little bit tricky to find right now um I know Chris though one of the other things that I was thinking about when I heard this question was the actual numerical output for the bridge so for instance if you go back to the 1D um Bridge Raz project that you had open um obviously you're familiar with the bridge detailed output um which is that table next to the full detailed output so not that one but the one next oh you want the detailed yeah yeah so that is obviously great numerical data that you can get for any onedimensional Bridge um gives you a lot of different numbers to potentially use for design or or analysis and whatnot is there a way that you can get similar numerical results for a 2d Bridge yeah there is it's not as plentiful as you have in 1D um if I let's see what happens if I right click on here you can see plot structure variables that allows you to get into a hydrograph plot there it is on my other window and so similar to what I showed you in 1D you can bring up a stage and flow hydrograph plot that way and that allows you to also look at that data by table but that's really just going to show you Headwater and tail water stages as well as flow now what you were talking about is all the other stuff in the detailed output table that we can look at for 1D all of this stuff like you know the que discharge down into the different parts Bridge versus Weir um you know bridge opening area velocity total you all these other things in here um that's not yet available easily in Raz 2D um a lot of this stuff you can get to if you dig into the hdf file um but that takes a little bit of work to do that so um yeah and this is one of the things I was talking about that they're still improving and and adding new features in that sort of thing so hopefully in future versions we'll be able to do something more like this detailed output table in 2D cool all right next question uh that we had from users was around um editing Manning's end values for a 1D model and this particular user was having some trouble doing that and so obviously it's hard to troubleshoot for somebody without seeing their model but I thought we could just go through a really really brief uh overview of the different options that are available for modifying Manning Zen in 1D yeah yeah so first of all if you're in 1D your n values they reside in the cross-section editor so I've clicked on that and there's a couple different ways that you can Define your end values one is called kind of the traditional three subsection method and the other is horizontal variation what we're looking at here is horizontal variation and I know that because when you do horizontal variation it adds this third column right here and you can put as many n values in your crosssection as you have station elevation points if you want to I would never do that that would be silly and way Overkill but technically you could do it but by doing horizontal variation that frees you up to have more in nvalue definition than just the three subsections which you see right here okay now to switch between those you would go to options and right now we have horizontal variation I can uncheck this and it's going to give me a warning that I'm going to lose all this stuff but I'll just say yeah that's fine and now it's assumed what I want to have for my left over Bank my main Channel and my right over bank so now instead of having these four end values like you see right here when I apply data now I just have left overbank main Channel and right overbank so this is where you see the end values and this is how you do different ways of Distributing end values I always like to tell people if you can get away with just using the three subsections then do that that keeps things a lot simpler especially when you start getting into interpolation of cross-sections because your n values also get interpolated and when you have horizontal variation and end values you can end up with interpolated sections that have a lot of end values broken up in there that are not necessarily very accurate anyway and it gets really messy so I always like to encourage you to try to do the three if you can but you know hey if you've got different roughness out at the edge of the flood plane versus kind of near the bank then you should account for that and you can do that by going to horizontal variation in N values yeah one of the and one of the other things I thought would be worth pointing out is um oftentimes reviewing and modifying Mannings end values that you have for a model um especially if there's a lot of cross-sections in it can be pretty overwhelming to do that one cross-section at a time so one of the other helpful um tables that I reference quite a bit is the Mannings end table which is accessible under the geometry data editor under tables Manning Zen um if you pull that table over it'll show you um if you have just a simple left overbank right overbank channel distribution for all of your cross-sections you would just see three uh n values in this case you can see most of these have spatially varied end values associated with them but this is a really quick and easy way to qaqc um and it also highlights the values that are kind of defined as in channel here um so you can review make sure that your channel values are consistent because often times they should be and then you can have some variation in your flood plane based on land use characteristics and other things so um this is a really really helpful table for references um one of the other follow-ups to that question Chris was in regard to um values Manning and values of negative 9999 which is kind of a a default basically null value that you can get um for Manning and cross-sections is that right yeah absolutely sometimes you'll see that um I see that occasionally um well I don't want to say occasionally but when I'm doing pre-processing of my cross-sections in Ras mapper and but I don't do my n values over there I just I don't have an nvalue layer over there and so if I bring my cross-sections into the geometry window Raz doesn't know what to do with n values and so it'll drop agga 9999 at the first cell and then all that means to you is that hey there's not a real n value here yet you haven't input it so let's go ahead and do that so it's just more or less I don't know why they pick that number other than it's probably so not an end value that you would use that it makes it obvious that you need to fix that and so um so if you ever do see that that just means hey you haven't set your end values yet so go ahead and do that yeah and again I think that's most common when you generate the cross-section features within Ras mapper um which is something you can do you can get the shape of the cross-section you can also get the MS in for the cross-section if you have a land cover layer established in Ras mber but like Chris said if you don't you're going to end up with kind of those no values for your cross-section so just be aware of that as you're pre-processing your cross-sections and Ras mapper so great question um and thanks Chris for for giving us some visual answers to to both of those um I would fully expect there to be some uh continued Improvement in terms of the 2D output that's available because I know it's an extremely popular feature within Raz I've got a hey I've got a I've got a question for you guys okay um with respect to end values what is your default go-to end value for a river just I think everybody everybody has a number that they start with and then they look at okay well what's going on with this river is there a lot of vegetation is it steep a lot of Boulders or is it really smooth what's your starting default because I have one and and if you don't have one then do you have like Jeopardy should be I think we we just we'll go around see how close we are to each other so Robert why don't you go first and then I'll go and we'll let Chris finish it out I usually try 0.04 first okay that's my mine is 0.045 ah my starter so mine is 0.035 and I just realized that we're three the three biggest nerds right here talking about this but it is interesting to know where that came from and and like for me I don't know where that came from other than maybe my very first Raz model the river was 0.035 I don't know but um yeah it's kind of interesting right I think in my the reason that I think the reason that I use that value is because I think 0.04 or 0.035 is probably a little more realistic but I always tend to use a little bit of a higher Ming Z value so my model's more stable so it's a little bit a learned um crutch I guess you could say um but that's that's typically where all higher end values higher end values do give you more stability that's for sure yeah yeah any that's a good that's a really good question Chris that's fun so yeah and it's it's funny how we all start there but then you know you kind of adjust as you go and depending on what what features you've noticed in your River and stream but yeah yeah well I I think that's a great one to solicit some feedback Maybe can answer have that as a poll question this week but if uh if you are so inclined it'd be really great if you could comment on this video in the YouTube channel or on LinkedIn with what your default River value is for Manny Z value be interested to if it's similar to what Chris Robert and I use kind of around that 04 or maybe it's something a little bit different so go ahead and comment on that and we might uh uh elicit a poll question here to to get a distribution of values that'd be kind of a fun exercise yeah I bet people who work up in the mountains probably have a much higher default yeah for sure cool awesome well couple quick announcements before we get into the primary topic of discussion today uh first on the class front so as a reminder Chris and I have a couple of upcoming classes that you can sign up for today um the first is our in-person class which is going to be held in Atlanta Georgia uh um September 10th through the 12th I believe that's a Tuesday through Thursday um and it's looking more and more like that we're going to have our 2024 Pub and Grub uh attached to that event so if you're interested in uh learning about Raz and then attending a a night of Raz conversation presentations and a few adult beverages um it's a really really good time we've had some some really great turnout for those the last couple years so mark your calendars if you're in the Southeast September 10th to the 12th we will be in Atlanta I hope it's not going to be hot at that time it's not not dead dead middle of the summer but yeah it's Atlanta could be hot but I love Atlanta it's a fun city it's a big city and we're going to be right off of the Georgia Tech campus that area of downtown Atlanta so I I guess it's probably not technically called downtown but it's pretty close but um yeah come join us if you uh if you want to have some 2D training this would be a good one to come to it's going to be a lot of fun great and then uh our next online class offering is going to be our standard six week course which is one day a week of lectures with workshops in between uh that's going to run from October 9th to the November 13th that is also available for for signups so um if the in-person class isn't your speed or maybe you can't pull off work for uh get out of work for three days then uh that's going to be a good option for you too as always if you have interest in training in an area near you or for your company you're welcome to reach out to Chris or I and we can always set up something um you know just for you customized uh depending on what you or your company needs are uh also on conferences a reminder Chris in just a couple weeks here will be at ussd in Seattle Washington that's going to be April 22nd through the 26th uh if you want to swing on by the Klein Schmid Booth say hi to Chris ask him some nerdy Raz questions tell him what you use for your default uh River Mannings end value um Chris will be there along with some of our other H&H co-workers so come do the uh do the 5K with me I was um I was informed by our conference committee lead that I will be doing the 5K so I'll be running um yeah we'll see how that goes very good uh and then I will be at uh asfpm National Conference in Salt Lake City uh in middle of June uh giv a presentation there on some HC Raz work that we've done here at kenm so I'll be there with a coworker if you're going to be at asfpm it'd be great to to see you and connect at that where's that one again Ben Salt Lake City Salt Lake City nice yeah going to I already got I already got my uh times I'm picked out to go to Red uana which is the best Mexican food uh me best Mexican restaurant I've ever been to in the United States so really yeah it's pretty phenomenal have you been to Red uana Robert I have and it's the only place I've been in Salt Lake City we it up on a a road trip and it was it was incredible yeah Robert has a a real connection with the wakan region of of Mexico so I don't want to say that it's probably the best mole that he's ever had but it's definitely the best M I've ever had so yeah cool um and then lastly trivia question I told you I was gonna get you guys today because Chris has got me a couple times um but this is the softball I think I think you guys can get this so there is only one Raz file all all the files associated with Raz have extension right so the restart file is rst the output file is do. O right HTF file um there is only two there are only two files that have the exact same extension than Raz do you know what they are oh I I see where you're going with this do Robert knows okay Robert I'm gonna give you the accolades here I believe it is the prj files that is correct ding ding ding game nice well Robert you can go ahead and leave now you you uh you're 100% on Ras trivia so it's only downhill from here Chris did you get that were you gonna get that one too yeah that that was gonna be my guess too um it's it's really it it's always comical that you know they have the same projection and how often people try to open a projection file or they try to use a Raz project file to set their projection but yeah so my question is who who has claims to being the first one to use that project or that extension you think do you mean who what do you mean who has claim oh I the project it's got to be the projection because people finding spatial references long ago yeah but but were they doing it electronically like because Raz goes He Rass goes back to the uh to the early mid 90s and we weren't doing I mean maybe some people are doing gism but not a lot of people are that's an interesting one it' be interesting to Ask Gary that I think they should fight it out to see who gets to claim prj and make the other one change it I know right well actually I see CRS used a lot for projection files too so maybe they should just change it to CRS makes more sense because it's more than just a projection yeah yeah agreed good well well done Robert that was probably a little bit too easy I should have got you on that one but uh what no that's okay I'm I'm recently refreshed because as I get into sediment transport I had to relearn a whole new batch of file extension names so I'm always talking to Chris and others that Klein Schmid about uh what's in this model what are these files so yep yep very good well let's get into the primary uh topic for today um which were uh you know the reason primary reason we had Robert on for this conversation is because some of his ique experience that he's gained over the last year so working on a cool project um that involves doing some restoration modeling but in 1D which is not something typically that you'll see it's less and less common now with u the robustness of the 2D Raz engine uh and kind of some of the detail you can get within 2D modeling you see a lot more people kind of going that direction for uh for restoration modeling but just like anything I think there's still a place for 1D modeling when comes to Restoration and so we thought it'd be prudent to have Robert on to discuss his work some of his experiences and then we can get into some of the conversations around you know when 1D modeling is still going to be a good tool to use for this type of project work uh before I turn it over to you Robert to share kind of your experience and what you've learned um I did want to share uh the results of a poll question that we asked related to this uh topic and that was um you know which how do you represent restoration features within Raz um we had a number of different options up there some of them 1D some of them 2D some of them energy dissipation related like Manning's end some of them physically related like ter terrain modifications and the answer to this surprised me a little bit 75% of the people that answered that poll said they simulate restoration features using the newer terrain modification tool that's available in Raz so actually making those physical um changes to your terrain to account for uh the loss the form loss right that's that's you can get within 2D Raz uh so that's interesting we'll maybe talk a little bit about the downsides or upsides to to that particular approach but Robert if you want to kind of give a little bit of background and and what you've learned with doing restoration modeling with with 1D going back a little bit in time certainly I appreciate it yeah the project I've been working on includes uh many examples of river restoration conceptual designs um most of it has been done in 1D for two main reasons uh a model that we received previously was built in 1D and it's very robust and it is extremely detailed so it was important for us to continue that work um especially because it was calibrated and reviewed you know tens if not hundreds of times so it was important to continue using that same 1D model the other main reason is uh my mentor one of my my mentors here at Klein Schmidt primarily Works in 1D he has many decades of experience and is very very comfortable with 1D modeling and I think that has benefited me because a lot of what you can look at in 1D uh you can back calculate or quickly find the source or reasoning of why the model is doing what it's doing not saying that that can't be done in 2D but I feel like it's readily available in 1D so it's been a very good learning experience where all of these simplifications we're making to make the 1D model work um has been informative to me and in how you represent stream restoration elements in a in a river yeah one of the big challenges that Chris and I talk always talk about with 2D modeling is the fact that it's a little bit of a black box right and that's a challenge for people um especially as they get into troubleshooting and digesting the results of their model um it's kind of hard to explain how that's computed right versus with 1D like those are established equations that you can like you said Robert back calculate using hand calculations to check your check the results of the model check the work um and I think there's some real benefits to that approach especially for you know a younger engineer who's kind of growing his understanding of fundamental Hydraulics right um jumping straight into 2D might be a little bit of a disservice and I know that's happening a lot and share my screen here all right so for everyone who is uh watching this as well there will be many ways that this information is visualized so if uh this first initial CAD drawing isn't compelling uh because it looks like a bunch of lines and information I promise I'll be showing uh versions in terrain format as well as satellite imagery so it'll be reproduced in many different ways um what we're looking at here is an example of a conceptual uh River restoration design with many different elements and features that you might see in other River restoration uh projects so we have a river moving um here from the left and up to the top we have multiple different elements right here is a Boulder Field Boulders being maybe like 2 3T in diameter to increase roughness we have a log Jam with piles um acting as an obstruction to flow um on the inside of this Bend and we also have embedded Logs with Boulder ballasts that are positioned here on uh what is a gravel bar this Gravel Bar here is defined by kind of the uh these are topographic lines um so this is an area where the the river is split into two main channels so Robert yeah I I got a question for you so you've got you've got three different restoration element types in here you've got the uh The Rock outcropping the Upstream end you've got uh some log jams it looks like on the Left Bank and then some some trees laid out and ballast by rocks right in the middle what are some of the things that you're looking at with the existing stream to make you want to put those there yes that's a very good question yeah versus somewhere else or or you know what you know what dictates one type versus another definitely yeah there's a there's a lot of uh Engineering in tuition that goes into choosing the elements for stream restoration sites um this one is specifically and as I identify pattern as I work on more and more projects a lot of what you're trying to do is slow the water down you want gravel to deposit in areas that are currently being scoured it's obvious here that at this site it's we're going around almost the you know maybe uh 110 10° Bend so it's important to slow that water down as it goes around the corner um to create uh habitat and refugia for uh migrating salmon juveniles going Downstream and adults uh returning um in the fall so some of this uh in this split stream is we're trying to create a slower um part in the the split channel so that we have velocities that are more conducive for um sa salmon traveling up and downstream In this River we also are trying to uh reduce some bank erosion that's not our primary objective the primary objective is to increase spawning habitat in this River um but as a byproduct we're probably going to uh reduce erosion on this outer uh Bank of this Bend yeah and you definitely don't want to put in some restoration features and then come to find out they initiate some sort of erosion event and right there at the outside of the bend would definitely be something to keep your eye on right because you would expect more erosion over there so yes that's a very good point um and we'll actually see that that's a good tea up for uh when I start showing the results you'll be able to see how some of these features affect the velocity and water surface elevation of this site yeah and how they even though it's 1D and it's all Backwater calculations 1D still can represent some of the interactions between these elements um at least in a simple way so let's now go to hecr I'm assuming the audience is familiar with um what a Terrain visualization looks like where here we have white being higher elevation and these lighter Blues being the uh main Channel bed um this is just a representation of how we've gotten the cad design into the hecr RAS mapper format um just by using shaped files and spatial references so that we could see now how do these elements that have been drawn you know in their own space in CAD and how do they line up with some of the one-dimensional cross-sections that we already had in our model one question Robert on that just I just thought of just now you you mentioned the fact that this is an existing 1D model you wanted to keep kind of as as as much as you could because of the quality the calibration and whatnot did you end up actually adding any additional cross-sections for the restoration site itself or were you able to kind of capture everything you needed to with some of the methods you're about to describe yes it was important um to retain the cross-sections in the exact same format um both for you know the Integrity of the model that we received but also because as you can see in this image this uh digital elevation model is essentially hydr flattened or the liar couldn't Pate the water so this River although it has varying depth across the channel um you can't see that in the terrain or at least the terrain that we uh were able to receive um so the these 1D cross-sections actually have surveyed um one-dimensional um station elevation transects so we needed to keep that information in the model and try to best represent it um through other cross-sections that said of course the number of transex was much less than what we needed to represent the density of these these ribber restoration elements so we did add interpolated cross-sections to the model in order to try to capture um the features as they might appear hydraulically in the channel makes sense thanks yeah so let's look at one more visualization this is the um satellite view of this site not much to see cuz there's Shadows but um one thing I want to call out to attention is that Gravel Bar I was discussing on the first couple slides now you can see it how that might exist um in you know satellite imagery so to give an example of how we implemented these elements in the 1D model and what kind of results we are able to get out of it I would wanted to focus on this specific cluster of elements being the uh log with Boulder ballasts and the log Jam structure with a flood fence um and see how that specifically is modeled along this transect that we've put into the model so this is another uh this is a sharper view of the cad drawing the conceptual design we can see those um these two elements here and what they end up looking like in the drawing design is you have your Log Jam here with crisscrossed um you know you have your bowl and your your trunk or your root wad combined and then you have piles interspersed between those logs uh amongst that structure and then also with the embedded ballasted logs we have a log that's embedded uh with the bowl side down into the the gravel we have a boulder on the um tip of that log to wave down and then we also have boulders to weigh it down where it breaches the surface and actually adds this is the part that adds roughness to the channel this is the part where I'm most interested in um you know its effects on the river as a as a restoration element so the the trunk is twoth thirds of that is just to to Anchor it into the ground right correct and the the parts sticking out is just for for roughness and flow complexity right correct yep exactly you you hit it roughness and flow complexity and flow in this scenario is going from right to left for both of these um so Flo is hitting the root Rod side of the uh trees first now Robert I look at this and you know right away especially that drawing on the left like that's a very complex uh detailed design element right and so um my you know hecz 2D lizard brain is going off saying how can you possibly model that in 1D that's so complicated there's no way you can account for all the complexity of that with a 1D cross-section um how did you guys approach you know that particular problem which I'm sure a lot of people kind of come upon when they're deciding whether to do 1D or 2D yes that's a that's a very good question I'm sure as you have seen and the audience has seen that uh this is a Hot Topic on research and it has been for a while but there's been so many uh papers coming out about these you know multilog Jam features and how flow interacts with them how they add roughness to the the river um what you're about to see next how it's implemented in the model will be probably the simplest version that you'll see them implemented but there is such a wide range of ways to model it even um you know in 2D uh Chris and Chris has shown us that you can model it with piles you know actually trying to get the the features as indiv elements you can model it with Manning n as a uh Mannings end refinement as many people do in the workshop three uh of the heg grass class um and then people also have cfd computational fluid Dynamic models uh to really get into the nitty-gritty of this um the question is is in terms of cost and time efficiency what is good enough right for example part of my project uh this spring was looking at the calculations of uh drag and buoyancy on these structures and many of the equations that you can find both modern and kind of the classical equations have factors of safety of three or four uh because there's a lot of variability in these trees in the water surface elevation you know construction methods once you start working in a channel it becomes increasingly difficult um so to say that what we attempt in 1D isn't good enough is difficult when you have a factor of safety of three being applied to your your calculation um so a lot of it is a is a optimization of cost efficiency um and accuracy in into in what you're trying to achieve that's a good question there's so many ways to do it do you guys have any thoughts on you know what you might use well I'm I'm I'm Gonna Save my thoughts until we see some of the the 1D representation you have these so yeah well the results that we used were informative for our designs but they weren't calibrated so I don't have an ultimate you know um um comparison but we'll we'll we'll talk about what that looks like you know what's interesting Robert about the the image in the upper left where um you're showing the the wood the the flood fence is that what the orange circles represent is that the fence or or is the whole thing considered a fence typically uh the flood fence would be just Upstream of this elements um so a flood fence would I believe the main purpose would be to reduce scour that would be occurring at the bottom of the structure um okay so I believe I believe a flood fence would be implemented in you know future designs or as an alternative to this type of structure this is generally The Log Jam part of it what I being a HEC RZ modeler what I really like about that design is that's easily modeled in Raz because we can put in peers in as terrain modifications right to represent the the vertical anchor logs and now in the current version of Raz we can put in little tree modifications or big tree modifications for that matter and so we could lay those out in that Matrix pattern like like you see right here and replicate this pretty closely in Raz right now with terrain modifications well I'll push back on that a little bit and say I don't know if it can easily do it you might be able to easily add the modifications the challenge becomes doing that for multiple proposed features having small enough cells to capture all those elements individually yeah and having a model that still runs efficiently and I think that you know Ka in a little bit with what Robert's about to show Us in terms of you know the 1D representation so yes prepare to be shocked you've never seen this before so how we represent those features um in this iteration of the conceptual design is using the obstructions feature um each of those two log jams are two being these two individual um with one kind of being higher up the slope and one being a little bit lower on the slope uh we represent with just obstructions uh built straight into the 1D uh transet that would be the the ond version of terrain modifications right yeah good point yeah obstructions yeah you're just taking a an a slice and of course how these obstructions operate is it completely blocks flow there's no flow moving through it when it does its computation which is a conservative result so um when you talk about computational fluid dynamics and even 2D um it is just still conservative to say that no water is moving through that that feature um in combin in combination with the uh obstructions we increase the Mannings end value to 0.1 to indicate an increase in roughness in this area um to represent the embedded log structures we do not um represent that with obstructions because they are a little bit more you know open to flow moving through them and amongst them than a log Jam feature would be especially if there's debris especially if there's debris get that gets caught up and stuff right exactly exactly yeah and a log Jam feature is similar to what you would actually see um occurring naturally in places and eded bed log you know maybe a little less frequent but definitely it it uh it can occur and but they're more Sparks right when you have just a huge old gr type tree just embedded in a in a river was there anything was there anything to the value of 0.1 for the Mannings that you guys chose there is that just based on experience or any sort of calculation it was partially based on experience um my mentor that's been helping me uh with this project chose the the value of 0.1 I think it's reasonable um but I'm sure that there's an academic basis to using 0.1 but we already we just had the discussion on what Mannings Z do you use to start with a river 0.1 might be the might be his value that he uses for debris well it's interesting because yeah because 0.1 you know kind of in that range 0.08 to point one two is kind of the numbers you're using for a forest right a right PA for and what are we doing here we're putting trees into the river so we're essentially doing something similar right so so that makes a lot of sense to me the point one it's a good point Chris perfect maybe it's a little thicker than your typical Force because you've got them all kind of woven together uh well although some of them aren't so anyway yeah definitely yeah so you can see that how we've kind of implemented those two types of features at least along just this transect here um one thing I wanted to bring up as we were discussing at the beginning of this vcast that uh the use of lateral VAR laterally varying Mannings end I do that pretty much almost always especially for transex here the entire Valley Wide so there's got to there's generally usually something changing uh for this location we're using 0.45 for the Channel bed so we're using Ben's number for the immediate overbank areas and particularly on this uh right Bank side of the transect we're using 0.06 because that's farmed land it's closer to like tilled or maybe have some low growing vegetation on it and then once we get up in a out of the flood plane here we go to 0.08 which is kind of like an open forested area Andor maybe a home with some structures and you know farming equipment so so I do want to make sure Robert that we leave enough time to discuss some of the 2D benefits of restoration modeling so um what sort of results did you get from this analysis definitely so here I have a quick visualization of what those results might look like so this table is values of the difference of running the hydraulic model without the restoration elements and then with the restoration elements and then taking the difference in the velocities um here I picked out that transect we've been mentioning where you have your log Jam feature which is causing quite a bit of velocity reduction in the channel which is the point of adding those elements to this section as well as these um embedded logs is decreasing the velocity part of that though of you can see that velocity is increasing between those elements as it squeezes past and rushes through that section so that's something to pay attention to especially for um considering scour in this design really quickly just because I'm sure there's some people that are that are thinking about this how did you get so many slices in a cross-section of velocity because a lot of people just know how to get a leftover Bank a channel and a right over Bank velocity but obviously you get a lot more detail here how did how were you able to do that definitely so in the 1D calculations it only calculates one velocity for the channel for the transect um but there is an option in I believe it's in the computational options or in the 1D um plan where you can ask hecz to slice your channel your left over Bank your right over Bank into a number of slices so it still applies that one computed velocity for that channel but it will use the hydraulic properties of those for in this example 20 slices to um extrapolate what that velocity might be for that slice cool thank you yeah so and added visualization to this and I'm a very visual person so Wy can actually be a struggle for that I I put those uh those tabular results uh directly laid out onto the map here where you can see uh darker red is reductions in velocity due to the elements and green is increase in velocity and once you start looking at how those 1D um values are placed spatially on the map you can see that it's working relatively well you're you're capturing the elements um in their in their space and how water is moving amongst them um and then from that another part of the calculations that we're interested in for these conceptual designs is um its overall effects to channel velocity and to to um the difference in water surface elevation due to the elements and you can see here these last five rows are rows that have stream restoration elements implemented and you can see that the water surface elevation is rising due to due to the increased roughness Channel velocity is decreasing for the same reason and then you can see that effect perpetuating to the other transects that are Upstream you know with kind of diminishing returns uh to see how that affects your stream reach in general you know on a more of a reach wide perspective rather than just transact by transact the bottom line is I think if you've got a lot of experience um in doing these kind of features um walking streams knowing Rivers knowing about geomorphology you can look at your 1D results and look at the plan form of the river and the Tendencies of the river to scour and it and you can make some pretty good assessments about where to put these features what kind of features to put in and how frequent and then what your expected benefit would be or potential um negatives too would be to those features and then you use this 1D output to kind of verify your assumptions and maybe give you a little bit more of a quantitative feel to what's going to happen or what you might expect but um ultimately I think too a lot of this stuff stuff and tell me what you think Robert but I see this as a lot of hey let's get it in here um we're going to do our best to design it to stay in place and and hit the the criteria we're trying to hit but the reality is some of this stuff may eventually wash away in into the river and uh maybe temporary yeah one question Robert um for you because again looking at this it might be a question that folks have uh so there are a couple features in here particularly The Boulders at the Upstream end of this image that don't have a cross-section associated with them but you can still see the velocity impact that those are having um at your cross-section there or slowing down velocity how did you account for restoration elements that don't run through uh a cross-section and I was talking about the so when I said the I should have said Downstream at the downstream end was the B here there yeah how do you account for that without having a cross-section go through them yeah it's definitely a lot of a lot of what I learned from the project um because you're trying to utilize so many different parts of the project keeping it similar to what it was so that you have some frame of reference to other computations and other uh model Alternatives but you're also trying to add resolution and add details to this model um so that you can see it working appropriately um in this specific example each of these uh 1D results has been applied to the trans that's just Downstream of it flow is moving from left to right in this image um so we are we are taking some Li Liberties and using some intuition to try to make the best of the transex um that we have and can enter into the model um and apply those features to them yeah and then being aware that we need to maybe not take the exact spatial uh definition of these transacts at face value but we need to reinterpret you know what that would look like for our site sure yeah that makes a ton is there is there any particular reason why you didn't want to interpolate a section to go through those Boulders on the downstream end yeah a lot of it so a lot of what we're trying to do and I bet you could talk about a lot more about best practices is that we want the transects to be relatively equally spaced that's just um a good good practice I suppose so in that in that effort we equally space these transects but that leaves those Boulders pretty much smack dab in the middle of the two so once we started applying these elements to these transects um it was apparent that we kind of wanted to represent that Boulder Field on the transect that's Downstream of it um but but primarily we uh set up the transects first you know and then put the elements on it to try to prioritize the modeling best practices versus like absolutely capturing all these elements um exactly where they're drawn yeah awesome well this is great Robert thanks for sharing this this is really I love this this figure in particular I think it's very visually easy to understand how really well the one-dimensional modeling is performing for capturing some of the hydraulic effects that these restoration features have um I do want to go through a rapid fire here just between the three of us to brainstorm some situations where again maybe this 1D approach is is is warranted maybe limitations of a 1D approach when it comes to Restoration um and you know when you might want to move into a a 2d space for for this type of design so um from what I took from this conversation you know one of the biggest drivers for uh this particular project and maybe other projects that people are are going to elect to use 1D modeling for would be you know a lack of bimetric data right if you're limited to bom metric data at particular cross-sections um you know that might limit the effectiveness and accuracy of a 2d model in that space I think that's certainly a a reason to stick with 1D another one that you mentioned that is is uh right on with this project is the fact you had an existing 1D model right so there's a lot of projects out there that are budget or time constrained might not have the ability to go through and develop an entirely new model but if you have a 1D model uh Robert some great ways how you can still answer some of these questions around restoration design with that modeling tool without going to 2D um Chris I'm going to throw it to you what are some things uh off the top of your head that are limitations potential limitations of of a ond model when it comes to Restoration design yeah well first of all I think another another reason why you know Wy could be used confidently here is you had you you have a team that has experience doing this work and knows how to interpret 1D results and apply those to a very three dimensional issue which is restoration features in a stream but some of the downsides with 1D you know you can get into the computations and and um you know how Raz discretizes the terrain in 1D it uses cross-sections cross-sections are um they produce a single water surface elevation across the entire cross-section and it calculates a single average velocity for that cross-section now as Robert showed you you can break that up into your subsections or even finer with these vertical slices so you can get a feel for how the velocity would be dist distributed but that is purely based on conveyance there's no momentum associated with that distribution it's purely just a conveyance weighted um distribution no velocity so the way Raz 1D calculates water surface and velocity is a little bit of a disadvantage there because these restoration sites are close in you've got flow patterns moving in and around them in a very near field so a lot of two-dimensional effects there that are just not going to be picked up with 1D but this is where the experience comes into play in my opinion where you can look at your 1D results you can look at how you're setting up your 1D model and you can use that to basically project that into a 2d space based on how you understand water to flow in a river like that with and without these features how you understand the science of geomorphology and how that's going to be affected by these features I think um plays into it so and then besides the computations now you've just got the the feel of the software and and how you can look at things and results uh you know if you're used to 1D obviously you're going to be drawn to that more there lot of people today especially younger engineers and modelers who've never done 1D they only do 2D and so they're not going to want to do 1D here 1D is actually really hard if you haven't done it 2D is a lot easier if you if you've done neither I think 2D has an easier um level of Entry but um but 1D has the history too so you have a lot of different output options we kind of looked at that earlier um with respect to Bridges and that you've got that that detailed output table uh that's been around forever but it gives you a lot of information about bridges well same thing goes for restoration features you may want to use that detailed output table you may want to use the profile output table or all the profile plots are already created and welldeveloped in 1D uh Ras but are still kind of trying to catch up in 2D um so I think those are some of the the things that come to mind for me anyways about 1D versus 2D uh personally I think you know we're moving towards more and more 2D and I think that's going to be more and more common for restoration work and it's um obviously one of the big downsides of 2D is the time it takes to run these models and when you get into restoration work especially on the project you're working on Robert where you've got a long stretch of river if you try to do that all that in 2D your model may take hours or even days to run especially if you want to have any sort of resolution around these features meaning small cell sizes right so yeah I we're modeling 20 miles and we have 10 uh sites within those 20 miles and we are really interested in how they interact as well you know mile by mile it takes 10 seconds to run 20 miles it's incredible yeah yeah and it helps us uh choose elements right the person I'm working with you know puts his thumb out and he has Decades of experience to say this is what it should look like but I might be trying 100 things and using the results to inform what might be best so yeah the computational efficiency is huge yeah I mean one of the rules of thumb that Chris was talk about in our 2D class is to start simple add detail where needed that's really catered towards a 2d kind of mindset is how I've viewed it but I think it applies in this case as well right um I think there's GNA be a lot of models especially in again a situation where you have a an awesome calibrated existing 1D model right what's the simplest thing you can do with that keep it the same right and so doing that and then you know potentially adding additional detail or potentially adding some 2D areas later on um you know that's the approach that should be done I think far too often like especially with younger inexperienced modelers that only know 2D they're going to jump straight to that tool and I think they're they're causing themselves to potentially walk into some inefficiencies that are unnecessary um particularly if you can be working with somebody that has so much experience you know processing 1D results um a couple other things off the top of my head uh in relation to this discussion one uh we didn't talk about super elevation at all yet um which I think is a is an important idea um and really not just super elevation but any sort of water surface elevation constraints um obviously you can only get one water surface elevation per computed cross-section in 1D Raz which is a real limiter in certain situations now I think this project Robert you were describing there weren't any real sensitivities around water surface elevation um in terms of like having to M you know stay under a certain water surface or having to have at least a certain water surface elevation yours was much more of a generalized where am I slowing up water how is that going to result in change to my geomorphology a little bit more qualitative of a 1D analysis I think in certain situations if you're doing restoration design um and you're having to make sure that that ation design doesn't impact the low cord of a bridge Upstream of it or something like that where there's going to be some definitive qu quantitative results that you need to analyze um it's going to be a little bit tricky to get that with 1D just because we know there's going to be some variation in water surface elevation around these elements and how the how that kind of extends Upstream in the same way um you know if we have water surface elevations that are sensitive around a bend in the river uh and we have restoration elements there uh you're not able to capture that super elevation that's going to occur as water moves around uh those features you will be able to do that uh in a 2d model if you're using the full shallow water equation so those are a couple things that that stand out as you know things to consider what is the restoration because not all restoration models are the same right we talk about River restoration as this General topic but there are river restoration models like what you're working on Robert that are more generalized a little bit more qualitative um we'll call them maybe more art than science right um and then there are river regation models that are going to be you know used for permitting where you have to say this is going to maintain a water surface of X and depending on what those objectives are you're going to have different tools that you're going to want to use maybe a combination of the two but you're going to you're going to want to make sure you're using the right one um the last thing I'll mention which is another important one and right at my alley is sometimes these restoration features are designed for um regulatory Rivers where Fe Regulatory authority over them and you're going to have to show a noise or uh submit a conditional letter m r Vision to show that you're not impacting properties adjacent to your project site uh that's going to be a lot simpler to do in 1D especially with encroachment analysis so if you are doing River restoration design happens to be in a regulatory area that could also be a reason to maybe stick with 1D for that particular project so a lot of different things to consider certainly pros and cons all around um I think my biggest take away from this conversation uh and conversations that we had leading up to the podcast is just the importance of um your comfortability with the tool as an engineer right um if you are really comfortable and really confident in the results of a 1D model and you have experience interpretating those interpreting those results you know that's a good tool to use same with 2D right if you're really confident and really competent content with 2D modeling that's going to be a good tool to use where we start getting into a little bit of trouble is when we start extending ourselves into tools or methods that we're not comfortable with and that's when you can make some mistakes Miss some things um so I think that that level of comfort is is something that's really really important to consider when making these decisions Chris anything to add in closing out on the conversation of ond 2D restoration modeling I think you really hit on on an important thing that we talk a lot about in our class which is Raz is a tool um you have to know how to use the tool you have to know how to interpret the results from the tool to make it a useful tool and um I kind of liken it to I was just thinking about this like you know if you're framing up a house you could do it with a hammer and nails you could also do it with a um 18volt battery power drill with a u you know um a hex drill bit on it and you could get the same thing accomplished right but somebody who's been framing houses with a hammer they're going to be really skilled at that they're going to be able to do it really quick even though the drill is maybe a fancier tool it's more modern um maybe even holds together better than nails long term I don't know but the the thing is if you you know how to use it you're going to do an equally good job I think in a lot of cases and I think this certainly this 1D 2D discussion for restoration kind of fits into that area um so awesome yeah really good Robert I appreciate you sharing that with us that was that was really cool I like your the way you did velocity distributions I think uh I think I'm gonna steal that yeah I think I think that's also a key takeaway for folks you know if you're going to be doing 1D restoration modeling to make sure that you have those velocity distribution broken up in your cross-sections to to maximize the utility of that as much as you can because yeah but and going back to the you know know the tool discussion is know what that is that's that is not a true two-dimensional representation that is just a velocity distribution based on conveyance and so that's that's different it gives you a pretty good idea but it's not the same as a a 2d model based on the um the full sanon equations yeah well it sounds like we another topic maybe Chris that we can talk about do a little comparison of 2D and 1D velocity distributions that might be a topic all right um great well thanks again Robert this has been this has been really awesome I do want to preview our next episode so uh coming up next month uh we will be talking about nature-like fishways uh 2D nature-like fishways how to model those what end values to use for those types of designs nature like fishways are one of those projects that are becoming more and more common as we're kind of adding uh robust fish passage facilities to hydro projects around the us we're going to have some guest speakers for that podcast so kind of similar to this we're going to bring in people that know a lot more than Chris and I uh on a particular topic and um yeah hopefully they do as good a job as you did Robert this was this was great having you on today yeah thank you awesome thanks Robert Chris any closing uh closing words come check out our classes and uh if you happen to be at ussd or asfpm come find me or Ben respectively uh we'll be there uh hanging out and um having a good time awesome well thank you again everybody for listening today hopefully you found this informative uh this has been full momentum in HC RZ vodcast until next time

Channel: The RAS Solution

Views: 650

Keywords:

Id: 8siKJmu7Nyo

Length: 77min 15sec (4635 seconds)

Published: Thu Apr 18 2024