Modeling for Erosion Prevention

Jan. 1, 2004
It’s been said, “An ounce of prevention is better than a pound of cure,” and some engineers are taking that adage to heart. When developing new sites, they’re using computer modeling programs not only to predict where runoff will occur but also to aid in creating systems that will control or collect that runoff, which in turn prevents or minimizes erosion.Knowing Where the Floods Are
Clearwater, FL’s TBE Group, a professional consulting service that helps municipalities with National Pollutant Discharge Elimination System (NPDES) compliance, uses geographic information system (GIS) applications to determine floodplains and changes to the floodplains over time.
Flooding often can be a problem in Florida. “Our water table is close to the surface,” points out Rick Bowers, TBE Group’s director of GIS. “Three to 4 inches of rainfall in an afternoon summer storm can cause flooding because [Florida is] relatively flat and we’re surrounded by water in the Tampa Bay area. Plus, at high tide our coastal storm sewer systems are filled with sea water.”With Florida’s population growth and continual development, miles of ground have been covered with impervious surfaces, which drastically changes rainfall absorption and runoff patterns. TBE is assisting with the Southwest Florida Water Management District’s automated computer models for Federal Emergency Management Agency (FEMA) floodplain mapping. “FEMA has recognized that the former floodplain boundaries are not accurate,” says Bowers. “We’ve had a lot of heavy-storm events in the [Tampa] Bay area this year. Our GIS-aided floodplain delineations have been fairly accurate as compared to actual flood areas. This visual result is giving us increased confidence in the GIS application and allowing us to redelineate flood boundaries for FEMA’s flood insurance rate maps. We want the public to have the best information possible. If people buy a property based on an old FEMA map, they might not know they’re on a floodplain. Floodplain boundaries change over time because of new construction, changes in soils, and so on.” After collecting the land-based data, TBE Group uses GIS and hydraulic models to analyze flooding conditions. “When rainfall data come in from the field for different rainfall events, we can use this information for modeling. We use ESRI’s ArcGIS software and have developed an automated process that analyzes the data we need. We take into consideration the land use, the amount of impervious surfaces, what soil types exist within the area, and other factors. We wrote the applications in-house using various programming languages that work with ArcGIS and AdICPR [Advanced Interconnected Pond Routing, a computer model from Streamline Technologies Inc.].“From GIS we get curve numbers [CN, an indicator of the amount of porosity of a ground patch] within a watershed,” Bowers continues. “That information is fed into hydraulic models, which determine the results of 10-, 25-, and 100-year storm events. We get flood elevations and then map them to determine the flood boundaries.”Mapping the floodplain is just part of the work. “TBE Group proposes stormwater improvements to reduce flooding. Sometimes the area is just going to flood, but most of the time we recommend solutions that can efficiently and cost-effectively lessen the flooding problem. We try to improve drainage conditions on a project-by-project basis,” Bowers concludes.Tracking and Taming Water Rain falls, and the earth does its best to soak up the moisture; yet even on undeveloped ground this process varies. Some soils can absorb much more water than others can. In addition, the amount and the type of vegetation on the site have a large impact on the absorption rate. A grassed site uses less water than a wooded site. Trees need more than 5 gal./caliper-inch of water per week, meaning that just one 10-in.-diameter tree alone will absorb 55 gal. of water weekly-more if it can get it, especially in the summer.

Modeling and preparing for water movement is necessary, especially when planning for new construction. For example, the Madison, WI, firm of D’Onofrio Kottke & Associates had a recent project – the Epic System Corporation campus in nearby Verona – in a particularly sensitive area. “The campus borders on wetlands,” explains Nathan Lockwood, staff engineer, “and beyond that is the Sugar River, an excellent cold-water resource. It’s a trout source, a very clean river, so it was very important to watch the stormwater that would end up in this river.”

The company used HydroCAD, a computer modeling program based on the United States Department of Agriculture’s (USDA) Technical Release 20 (TR-20) and Technical Release 55 (TR-55) hydraulic programs. HydroCAD contains most of the soil and storm-type data that users once had to input by hand. “I use it for detention basin design, channel design, and stormwater routing,” says Lockwood. “We have some engineers here who have never done stormwater modeling before; it’s easy, user-friendly.” Lockwood explains how the model operates: “The end result is graphs, and you create flow charts to show the water flow. With the charts, you can show where water will come out; when you add the rainfall statistics the graphs will show you how a storm will react throughout your system. A lot of the data you’ll need for the charts is built into HydroCAD; it contains information and backgrounds developed by government agencies, such as USDA and SCS [the Soil Conservation Service, now the Natural Resources Conservation Service]. A lot of what’s in the program is published material, but with HydroCAD, it’s all in one place; you don’t have to look it up and input it yourself.
HydroCAD sample applications“You then put in your default typical rainfall amounts,” he continues. “Yearly rainfall patterns are the same-research is so intensive. You get the national figures and then input extra rainfall data from your general area. HydroCAD also tells you how this storm plays out; for example, for a Type 2 storm, the program has all the information regarding this type of storm.”To meet the local stormwater ordinance, D’Onofrio Kottke used the model to incorporate catch basins and other systems to the Epic site. “We’re required to make sure that, for one- to 100-year storms, the peak rate of runoff is the same as it was before the site was developed. That’s normally how it’s done,” Lockwood explains. “However, in Verona, they require a ‘pre-people’ standard – you have to make sure you don’t have any more runoff or speed of runoff than would have been in a virgin prairie meadow-even though the site had been a farm before it was selected for the Epic System campus. Fortunately the permeability information for a prairie meadow is an embedded set of information within HydroCAD.”How much water would the site absorb? “You have to determine the curve number,” he says. “Pavement has a 98 CN; a farm field’s CN is about 70; a prairie has a 58 CN. All the calculations happen through the model. You have the CN assigned to the area, then you incorporate the time of concentration [Tc], which is how long it takes a raindrop [to hit] the most hydraulically remote point-the farthest-away point in the site, which varies from site to site [and is] never static-and you’ll perhaps measure the site from corner to corner, however the water flows. Of course, you may have a peak in the middle of the site; survey data and topographic maps help you figure this.”As Verona wanted drainage to match the prairie’s low CN, the farm field alone needed work, not to mention what measures would be required once construction was finished. “Then you start making some engineering judgments,” Lockwood says. “Do you perhaps need a channel? You put channel characteristics into HydroCAD. For example, maybe on a prairie surface, the water would take a half-hour to run off; on pavement, this would take maybe five minutes. You would look up your soil type-each has an assigned soil group. It so happens that every soil in Wisconsin is a Type B, and the CN for B-type soil is 73. So you see, compared to the prairie’s typical CN, Wisconsin soil has a lower permeability. Of course, ordinances are theoretical, a generalization-they’re not really based on what’s going on on-site.”Lockwood needed to “run all the numbers” for the Epic site. “We did modeling; we looked at the existing situation, the topo map, and the defining watersheds. We input all characteristics of the site into HydroCAD to identify the Tc and so on.” Once the model is set up, he says, “you then physically type in rainfall amounts for the area. A two-year storm equals 2.9 inches. This will give you volume of runoff, peak rate of runoff, time-length of the storm. Ordinances are most concerned about peak rates of runoff, but most ordinances don’t evaluate volume of runoff. We calculate for one-, two-, 10-, 25-, 50-, and 100-year storms. For the Epic site, we planned to be able to accommodate a 100-year storm, which is fairly common around this area, especially where there’s a closed basin. You want to make sure there’s a safe overflow, especially for the Sugar River.”Rainfall isn’t the only moisture to take into consideration; Wisconsin has long, snowy winters. “You do need to be concerned about frozen ground,” Lockwood points out. “For example, if you had a storm at the end of February when the ground is frozen, you have to keep this in mind; when it’s still frozen, the ground is impervious.”After determining the figures for the “virgin prairie” first, Lockwood used HydroCAD to discover how to meet the same criteria. “We get another map with the site marked on it, along with its proposed grading. In our case, we used open channels to move the runoff. Let’s say a 50-acre watershed that was prairie now has 4 acres covered with concrete. We’d be draining off a parking lot into a grass channel into the outlet point. We use detention basins; the model allows you to put a pond at the outlet of the site. In fact, we put in a pond for peak control and sediment control, yet it’s also for aesthetic reasons; Epic will probably put an aerator in it.“We didn’t need any draining pipes,” Lockwood continues. “We used grass swales, which you can model in HydroCAD’s pond node, and with gabion baskets we made a weir structure. That’s how the pond gets drained. That drainage rate is what you compare to the preexisting rate. The ponds have been built, and the construction crews are now working on everything else. It’s part of the ordinance that stormwater features have to go in first. The site will use some permeable parking areas; for example, [it will use] Geoblock on some access drives for the fire department.”
Not only did Lockwood have to deal with the runoff, but he also had to control its temperature. The site was thermally sensitive. “The Sugar River is cold water, so we had to make sure the heat of the runoff water stayed a certain temperature. Controlling temperature is a budding science. We had to run the water underground, where temperatures are 55°[Farenheit], and hold the runoff there for the five- to 20-minute interval when the asphalt or concrete was giving off its heat during the first minutes of a rainstorm. We used a rock trench. Water flows into that to hold it for a while, to dissipate the heat.”In the past, D’Onofrio Kottke used USDA’s TR-55 and the Virginia Tech/Penn State Urban Hydrology Model. “Neither of the previous programs was as user-friendly,” Lockwood notes. HydroCAD also generates the reports needed for a plan’s approval process. “We print out reports to take to Verona with drainage maps to explain how we got to the answer, which they would then pass on to their reviewing engineer, who would look it over and make comments. We also had to give reports to the [Wisconsin] Department of Natural Resources, the [Dane] County Regional Planning Commission, and the Dane County Land Conservation Board. The approval process used to be a quick look by the state or county; now it’s months of review. With the increasing number of regulations, agencies, and individuals getting into the stormwater issue and making it a concern, HydroCAD helps us meet the criteria and helps us explain processes to approval committees.”Holding-and Hiding-the Stormwater
“In two recent projects, we needed to design infiltration chambers according to NPDES Phase II regulations for an office building in Redhook, New York, and a 12-apartment building for seniors in Poughkeepsie,” reports Wendy Berger, P.E., CPESC, of Poughkeepsie, NY’s Berger Engineering and Surveying. “The drainage system was required to hold a two-year storm. You want to know storm and drainage peaks so you’ll know what you will need to build to hold that whole storm.”
Accounting for different types of storms once required more calculations. “We used to have to put in all the coverage of the area if we wanted to print out two-year storm reports. Then we had to start all over again if we wanted to show projections for a 10-year storm,” Berger explains. “With HydroCAD 6.4, we can do it all at once; we run figures for different storms simultaneously. This is especially useful because, although we’re required to treat the two-year storm, when we’re able to-due to the budget, for example-we hold the 10-year storm as well.” Recent upgrades have made the model easier to use. “We used to have to give a diagram of the site, one we made by hand, along with the flow charts. Now if I want to describe sites in more detail, instead of ‘node,’ I can mark it as ‘shed one, shed two,’ et cetera. HydroCAD does have limited nodes on versions, but you can reimport the data and make ‘shed A, B, C, D,’ and so on if the area requires more nodes. You just tell the program: ‘Import those same values that I had from pond five.’ You’re not losing the data.”

Modeling has helped Berger solve another stormwater-related problem. “Infiltration ponds were becoming liability problems for property owners, so we started putting in infiltration chambers, usually under a parking lot,” she explains. “We make sure the chambers under the parking lot are secure. We build for H-20 loading [as defined by the American Association of State Highway and Transportation Officials, a vehicle weight of 36,000 pounds per axle]. Just don’t skimp on your stone or on the height of soil over your infiltration chamber-as most chambers today are made of plastic-and there should be no cave-ins from the parking lot, even with all the weight [chambers] carry.”

Once a project is completed, however, Berger’s work is far from finished. “The irony is you can’t really predict erosion. You put in what you think will best hold the soil, but then you need to field-check from time to time; where you think the water will go-sometimes it flows someplace else. We’re constantly checking and correcting sites we’ve worked on. We try to keep everything as stable as possible with vegetation, although on higher slopes we’ll use Curlex or polyester-lined channels. For a really high slope, we’ll use riprap or line the ditch with interlocking grid pavers.”Underground Retention Consulting Engineer Edwin Minnick, P.E., in Berwick, ME, has used quite a few modeling programs in his long career. “After 32 years with the Soil Conservation Service, I entered the public sector, joining a consulting engineering firm located here in Maine. When I began my consulting services in 1992, I used the SCS’s Technical Release 20 project formulation hydrology for all my drainage review modeling. Somewhere in the mid 1990s, I was introduced to HydroCAD and have been using it almost exclusively ever since.”

For a recent project in Salem, NH, Minnick says, “We were putting in underground stormwater storage – like a retention pond but underground – which consisted of 20 to 30 pipes stored full of water. We had to have an outlet for this water, and HydroCAD made it easy to plot this; we couldn’t have been able to do it with TR-20. HydroCAD handles split or divided flow [taking water to one pond or another], whereas with TR-20, you have to work yourself ragged. The law of hydraulics is built in, as is weir flow, pipe flow, and so on. The model calculates the discharge.

“In the old days, this was all done with a calculator and a spreadsheet. It could take weeks to do one pond by longhand,” Minnick recalls. “When the program TR-20 came out, you could do that work in a day, but now with HydroCAD, you can do it in a minute.”He points out that the model makes available, for automatic input, all of the runoff curve number data contained in the SCS land-use tables from its TR-55, Urban Hydrology for Small Watersheds. “HydroCAD has numerous methods for computing the time of concentration. It includes hydraulic analysis of just about any structure you could imagine-with the option of inserting your own stage-discharge figures if you so desire-and it offers several methods for computing the stage-storage data.” Information on all soil types is included. “Type A is droughty soil, very pervious, little runoff. B soil doesn’t perk as fast. In Type C soil, you’ll find hardpan at 30 or 40 inches, and a Type D soil is a wetland.” Even a model, he notes, doesn’t do everything for the engineer. “In New England, you also have to determine runoff from snowmelt. You can put snowmelt in with a base flow. What I usually do is assume this snowmelt will happen at the worst time-say the ground is saturated or frozen. As this isn’t in HydroCAD, you have to do some reasoning on your own. In Maine, the worst time is spring. There’s a lot of snow on the ground, so if you get a couple of warm days, you could have a problem with runoff overflows.”Minnick cites just one drawback to using a modeling program, no matter how sophisticated: “It can make an engineer who thinks he knows it all look better-and, as with any computer program, if garbage goes in, garbage comes out.”Keeping Basements Dry
“We’re a civil engineering company, usually hired out by the actual developer from a survey standpoint for the planning boards,” explains Don Tatem, system project manager for Beals Associates in Exeter, NH. “For state permits, New Hampshire requires site-specific reports for existing flow conditions. Say we do a subdivision; we calculate flows to the property lines, adding the impervious stuff that will be added-driveways, streets-and calculate the swales we will need to put in. We try to make the runoff water flow at the amount it was before development or less.”
For example, Tatem continues, “Let’s say we get a 3-inch rainstorm in 24 hours. We plug those data into HydroCAD. As ground cover only dictates how much water goes in or stays on the ground, you note in the program what the site consists of: tilled ground, pavement, grass, woods, or underbrush. Of course, you do have to know what kind of hydrological soil type you have, and here in New Hampshire we have 90% C-type soils. If you’re just starting out, you’d have to ask your soil scientist what group of soil it works into.

“You plug in the parameters-for example, C-type soil, woods, and underbrush are what’s on the ground-and HydroCAD will tell you the flow rates for different rain events. Then you plug in the specifics on each building. Each residential home lot has an average of 0.1% impervious on it-the driveway, foundation, and so on. The model will tell you the flow rates with these changes.”

Tatem uses as an example the problem of sizing various stormwater structures. “Suppose you have a culvert down the way that floods a lot; you can find out how to size a retaining pond. You can quantify cubic storage of catch basins and culverts. You can also quantify the speed of the water. If it’s a 9% pitch, will the grass hold that flow? So we can tell the developer how big the culverts on driveways or streets have to be.”Fitting the Bill for Infiltration Richard Phillips, P.E., of Landmark Design Group in Williamsburg, VA, has a good perspective on the benefits of modeling programs and how they’ve evolved along with requirements and standards. “I started using HydroCAD in 1983 when it first got started,” he says. “Before that time, you could do some charts with the TR-55 manual or the SOTP-seat-of-the-pants-method. When I got started in this business in the 1960s and 1970s, when you wanted to get rid of water, you’d just build a ditch; erosion was not a consideration. Have a 24-inch stream? Dig a 24-inch culvert-that’s how we used to do it.

“But in the early 1980s, before we started getting [government] requirements, we started thinking about property damage and decided we needed hydrology software so we’d know how to route runoff through and around [property],” he continues. “SCS provided the tools to make some of these calculations, but only the government used to have computers to do such computations. Data from a number of sources are compiled into HydroCAD-data which we once had to go to the library to look up.” More features have been added through the years, he says, noting that the program used to be contained on a 3.5-in. floppy disk. “It was that compact.”

In “the old days,” creating retention ponds was not always so scientific, allows Phillips. “In the 1960s, we used to make retention ponds around parking lots, but all they collected was paper trash and carts-no water-because the only methods we had to use were computation methods.”He recalls a 20-year-old project on which software provided a definite advantage. “We were working on condo projects on hillsides by ski slopes, and we didn’t want to cause damage to the surrounding area. This site, on 40 acres of drainage area in New York County, Virginia, had a Type 3 retention pond [less than 2 acres] and a manmade wetlands. With HydroCAD, we got very good hydraulics information on what would happen with this pond-and we needed to know that, as there was going to be a destination hotel with water slides on the site. When we knew where the water was moving, we could make sure we weren’t creating erosion channels.“For our condo project,” Phillips continues, “I could project storm events and predict which storm would have an impact on water quality. Plus, we didn’t want to have flooding and liability on these expensive condos. For the runoff from each driveway, we thought of having a small pond. We didn’t want huge retention ponds at the end of subdivisions. We had a pipe underneath each driveway, and we could model that with the program. That was almost 20 years ago, and the site still doesn’t need large retention ponds.”He notes that every project involves not only runoff or water-quantity issues but also water quality. “As you develop a site, you’ll get compacted areas where the water runs off quicker. You’ll need to capture that, slow that rate of runoff. There’s also a water-quality aspect. You’ll have to get whatever sediment is in the water to settle out.“There are still some handbook methods of doing this kind of work,” he adds. “Governing bodies don’t want to ‘force’ you to use computer programs.” A modeling program can help engineers demonstrate to a governing board that they’re meeting all relevant criteria. “You can test different alternatives, consider two or three of them, and get the best results.”