Page Banner

United States Department of Agriculture

Agricultural Research Service

Specific Model Studies
headline bar

From time to time researchers at our laboratory are called on to perform a "site specific" model study.  That is, a study to evaluate the performance of high-risk or non-standard structures that are being planned for a specific location.  Our lab is well equipped to accommodate these requests, housing four separate buildings in which scale models can be constructed and tested.  Generally, a model scale is chosen based on the space and flow available in the model basins.

These studies have been performed for other government agencies, such as the NRCS, and for private contractors and consultants.  Examples of these efforts are the specific model studies performed for Sugar Creek, the Randleman dam spillway, and the Franklin County Lake spillway.  A short description of the prototype design and modeling procedures for each are given below.

Physical Model Study of a RCC Stepped Spillway for Renwick Dam, North Dakota


The Small Watershed Program administered through the USDA-Natural Resources Conservation Service (NRCS, formerly the Soil Conservation Service) has provided technical and financial assistance for the construction of over 11,000 embankment dams across the U.S. (USDA-NRCS, 2005a).  The construction peak in the Small Watershed Program occurred in the 1960s with the Tongue River Watershed Structure M-4 (Renwick Dam) constructed during this era.  Many of these structures including Renwick Dam were originally constructed to protect agricultural land, but over the years, landscape changes have taken place including the building of homes and other infrastructure in the immediate vicinity of many of these dams.  Today, Renwick Dam provides flood control for downstream farms, cropland, roads, bridges, and the city of Cavalier, North Dakota.  Consequently, flood control structures like Renwick Dam have experienced a change in hazard classification and/or no longer meet current State and Federal dam design and safety criteria (Schmidt, 2006).  As a result, many dams have inadequate spillway capacity, and altering the dimensions of the existing spillway(s) to increase spillway capacity is often limited by urbanization, topography, geological formations, and/or unobtainable land rights.  To provide adequate spillway capacity for these structures, many design engineers are choosing to place roller compacted concrete (RCC) stepped spillways over the existing embankment. 


Nearly 10% of NRCS flood control structures are expected to use stepped spillways as a design solution for rehabilitation.  One advantage of stepped spillways is the significant energy dissipation created in the spillway chute.  The energy dissipation in stepped spillways allows the use of a shorter energy dissipating stilling basin as compared to conventional smooth concrete spillways.  Energy dissipation that occurs in stepped spillways with slopes of 2(H):1(V) or flatter (slopes commonly found among NRCS constructed spillways) is not readily available in literature, and few design guidelines have been developed for flatter sloped stepped spillways.  Additionally, little design guidance is available for the energy dissipating stilling basin associated with these spillways.  Consequently, the North Dakota NRCS requested a specific model study to evaluate a proposed design of a RCC stepped spillway that is being considered for the construction on Renwick Dam.  The results presented herein are intended to assist the North Dakota NRCS in the selection of an appropriate design for a RCC stepped spillway.

Big Haynes Creek Watershed Dam NO. 3 Gwinnett County, Georgia Specific Model Study

Big Haynes Creek Watershed Dam Number 3 (H-3) is located in an highly urbanized region of Gwinnet County, Georgia.  Originally constructed to provide flood protection of agricultural land, the dam was re-classified as high hazard due to the changing demographics surrounding the dam. Upgrading the structure to meet Georgia Safe Dams Program Engineering Guidelines is necessary to meet the new high hazard classification.  The Georgia Natural Resources Conservation Service (NRCS) requested a specific model study for two separate proposed designs of a converging roller compacted concrete (RCC) stepped spillway for the rehabilitation of H-3.  The results of the study  assisted consulting engineers and the Georgia NRCS in the selection of an appropriate design for a RCC stepped spillway.  This study provided the framework to expand the research in the development of generalized design criteria for converging spillways.


Site and Project Description


Originally constructed in 1963 for flood prevention purposes, Georgia Dam Site H-3 is located in Gwinnett County, south of Lawrenceville, Georgia (Golder, 2003).  The dam is situated within the northern area of Big Haynes Creek Watershed.  The watershed drainage area is approximately 1,702 acres (2.66 mi2).  H-3 was originally constructed in a rural area, but in recent years, the area has urbanized with newly built houses located near both abutments of the structure.  The land use immediately downstream of the dam is heavily wooded wetlands and uplands, with a concrete bridge located approximately one thousand feet downstream of H-3.  Due to the increase infrastructure in the area of H-3, the hazard classification has changed, which warrants the need for upgrading the structure to meet current state dam safety requirements (Golder, 2003).    


Golder (2003) provided the following information pertinent to the model study:


·         The structure was constructed with a maximum embankment height of 33-feet, and the crest elevation for the embankment as provided by Golder is approximately 954.6 feet above mean sea level (ft msl).

·         The embankment slopes were constructed three to one (3H:1V) on the upstream and downstream faces.  The proposed step heights in the 3(H):1(V) chute are 0.3 m (1 ft) high.

·         The normal pool elevation was set at 936.5 ft msl, and the flood pool elevation, at the emergency spillway crest, was set at 952 msl, which provides 539 acre-feet of flood storage.

·         The proposed spillway crest is a 100 m (330 ft) long ogee crested weir.

·         The design discharge for the ogee crest profile is 763 m3/s (26900 cfs).

·         The proposed convergence of the spillway is 52°.


Sugar Creek Riffle-Pool Rock Chute


In the summer of 2001, our engineers constructed a hydraulic model to evaluate the performance of a new riffle-pool grade-control structure located on Sugar Creek, approximately two miles south of Gracemont, OK just downstream of the Highway 281 Bridge.  The purpose of the structure is to stabilize the discontinuity in the channel bed resulting from a 6 ft. migrating headcut.  The structure is designed as an in-stream grade stabilization measure that allows fish passage and biological enhancement.  The model study consisted of three phases, which examined the seepage characteristics, rock stability, flow velocities, and overall hydraulic performance of the new grade-control structure.


Phase 1 involved the construction of a small-scale seepage model in order to study the seepage characteristics of the proposed structure.  The seepage model was 4 ft tall, 8 ft long, and 6 in wide.  Three sheetpile cutoff depths (4, 8, and 12 ft) with three driving heads upstream of the cutoff (4, 8, and 12 ft) were examined.  The model was used to confirm results with a two-dimensional, steady state computer program that predicted seepage.

A 1:4 scale 2-D model of the deepest section of the channel was constructed in Phase 2.  The model was constructed inside a 6 ft wide and 96 ft long flume with 8 ft tall sidewalls.  Water was delivered to the flume via a supply canal, and the flow rates were measured just upstream with an 8 ft wide modified Parshall flume.  Flows were initially set to examine the rock stability and general flow performance.  The same flow rates were then repeated while measuring the flow velocity at selected locations.


Phase 3 consisted of a <1:20< SPAN Minute="20" Hour="13">scale 3-D model designed to represent the entrance conditions, the rock-drop structure, and the outlet conditions.  The riffle-pool sequences provide enhanced fish passage capability and improved geomorphological behavior within the grade control structure.  Rock material used in this model was three inches or less in diameter and sieved to collect the desired material size.  Six different test flows were conducted in the 3-D model.  The model scale allowed for testing the performance of the structure for routine flows and extreme flow events.  The hydraulic performance of this structure proved it could handle the frequent low flows it will experience as well as extreme flow events.  The rock drop structure should provide a reasonable opportunity for fish passage. 



Randleman Dam Spillway

A hydraulic model study of the proposed dam spillway and stilling basin in Randleman, North carolina was performed in the early spring of 1997. The project features a roller compacted concrete dam embankment with an overflow section consisting of a 500-ft wide standard ogee crest that transitions to a stepped spillway surface. The purpose of the study was to assist in the design of the stepped spillway and the USBR stilling basin downstream of the dam. The hydraulic performance for selected spillway convergence angles, the step height, the energy dissipation afforded by the stepped surface, and the performance of the stilling basin were all evaluated in this study.

The model study was performed in two distinct phases. A 1:40 scale two-dimensional model was examined in phase one. This 2.5-ft wide model represented 100 ft of width in the prototype. The two-dimensional model was used to examine a smooth spillway surface (one without steps), and to examine 3 ft and 6 ft spillway steps. The second phase of the study involved constructing a 1:40 scale three-dimensional model using the recommended step size from the 2D model. The three-dimensional model was used to examine the convergence angle of the spillway training walls, the stilling basin performance, the riprap stability downstream of the structure, and the influence of an inlet tower attached to the dam face. Both models were examined over a wide range of flows that scaled from 5,000 cfs to 200,000 cfs.

Franklin County Spillway

Another model study recently completed at our lab is that of the Franklin County Lake principal spillway in Mississippi. This proposed system consists of a 6-ft by 24-ft drop inlet spillway, featuring a 77-ft tall riser with an elbow transition to a 6-ft square, 373-ft long conduit on a 2.41 per cent slope. The conduit exits into a 6.5-ft high by 56-ft long section which expands to meet a 20-ft wide stilling basin. The design discharge for the spillway is 1,426 cfs. This spillway and its components were represented with a 1:20.6 scale physical model.

The purpose of the model study was to evaluate the hydraulic performance of the spillway itself, the elbow transition, two different outlet basins (USBR Type III and SAF), and the plunge pool design downstream of the stilling basin.


Although design and performance criteria for drop inlet spillways were established years ago at this lab, this particular system contains some non-standard features which could be better evaluated with a model study. The first of these was the 4:1 length-to-width ratio of the spillway itself. To complicate this atypical geometry, a transverse wall was needed to separate the weir length into two 12-ft segments for structural reasons. The second unique feature of this spillway design was the use of a double circle elbow, used as a transition from the riser to the conduit. This geometry was needed to reduce flow separation as flow entered the conduit, therefore reducing the chances of cavitation occurring at high discharge velocities.

The results of this study provided a head-discharge relationship for the spillway, recommendations for the optimum elevation of the lower edge of the transverse splitter wall, suggestions for the stilling basin, and proposed dimensions and riprap size for the plunge pool downstream.

For additional information, please contact our office.

Last Modified: 12/12/2013