1 Introduction

Land disturbance caused by human activities, such as construction, development, or agriculture, typically involves the removal of vegetation cover and topsoil disturbance. As a result, stormwater runoff and erosion rates are significantly increased. The sediment loss rates from construction sites can range from 50 to 500 ton/ha, which is 10 to 20 times the losses from agricultural lands, and 1000 to 2000 times greater than for undisturbed forested land (EPA 2000).

Erosion and sediment transport from construction sites have both onsite and offsite economic effects. Scouring of the foundations of hydraulic structures, roads, or other structures are examples of onsite effects caused by soil erosion. Offsite effects are a result of onsite erosion, as the soil particles detached from the construction site are transported by the erosive agents, such as wind or running water, and deposited over farming land, storage reservoirs, channel beds, rivers, floodplain, and lakes (Morgan 1995). Along with economic effects, uncontrolled stormwater runoff from construction activities causes degradation in the aquatic habitat. High sediment concentration reduces the sunlight reaching aquatic plants and clogs fish gills (EPA 1995), which, in turn, has adverse effects on other aquatic flora and fauna. Landphair et al. (1997) estimated that erosion from construction activities in the USA deposits approximately 3.5 billion metric tons of sediment into water bodies annually.

Despite the increased importance of incorporating erosion and sediment control best management practices in construction sites, there is still a substantial lack of information on product performance under standardized evaluation protocols (Wolfe and Peters 2009). This situation creates difficulties for engineers, designers, and contractors in choosing appropriate technologies to mitigate the discharges of sediment laden water from construction areas. Performance data is often difficult to compare and interpret due to differences in testing conditions and evaluation procedures. Additionally, data available for sediment retention performance and effluent quality is frequently incomplete or partial. Although some prior studies have quantitatively compared performance of various products, these are typically performed using protocol and environmental conditions specific to that testing facility and may not be applicable to other geographic regions (Theisen and Spittle 2006; Wolfe and Peters 2009; McFalls et al. 2010).

The sediment removal efficiency of five different sediment retention devices was presented by McFalls et al. (2010) using a testing protocol based on the American Society for Testing and Materials (ASTM) D6459 Standard Test Method for Determination of Rolled Erosion Control Product (RECP) Performance in Protecting Hillslopes from Rainfall-Induced Erosion (ASTM 2007a). Water samples were collected at the inlet of the channel and at the outlet to measure suspended solids concentration (SSC) and compute the sediment removal efficiency. Treated wattle provided the best sediment removal efficiency and ranged from 54 to 63 %. The untreated wattle sediment removal efficiency was between 45 and 46 %, the Geosynthetic dike removal efficiency was between 17 and 20 %, and the silt fence removal efficiency ranged from 14 to 16 %. Rock check dam was also evaluated, and its results showed a relatively poor sediment removal efficiency of only 2 %.

The results from a study by Theisen and Spittle (2006) using an alternative testing procedure to the ASTM D7208-06 standard method (ASTM 2010) showed the Fiber Filtration Tube to perform much better in terms of sediment retention and turbidity (98 %/300 nephelometric turbidity unit (NTU)) than four other tested products: straw and coconut fiber roll, compost sock, straw wattle, and excelsior fiber roll. The straw/coconut fiber roll exhibited higher retention than the compost sock but similar turbidity performance (98 %/4500 NTU and 70 %/5000 NTU, respectively), while the straw wattle and the excelsior fiber roll performed similarly under both measures (68 %/7000 NTU and 65 %/7500 NTU, respectively).

To address the paucity of research on product performance in the area around Illinois, this study developed and implemented evaluation criteria and testing protocols for sediment control products under typical Illinois weather and soil conditions. This study also served to provide guidance to the Illinois Department of Transportation (IDOT) in the installation and maintenance of sediment control devices, as well as providing a source for quantitative data to help assess whether specific products should be permitted for use in IDOT projects. The presented field-scale testing protocol evaluates the sediment trap capacity and deposition analysis for the evaluated products. Previous studies (Theisen and Spittle 2006; Wolfe and Peters 2009; McFalls et al. 2010) have based their product evaluation in the suspended sediment trap capacity efficiency. This study not only evaluates the suspended sediment trap capacity of the ditch check products to be tested but also the sediment deposition that occurs during the evaluation. The sediment deposition analysis was included in the study using a laser distance meter to create digital elevation models so that the soil deposition and displacement can be analyzed.

2 Materials and methods

The evaluation of ditch check products for sediment control performance was conducted at the Erosion Control Research and Training Center (ECRTC) in the Agricultural and Biological Engineering South Farm, part of the University of Illinois at Urbana-Champaign. The total area of the demonstration and research site is 1.6 ha and contains an elbow-shaped berm, a detention pond, and three channels. The test channel used for the evaluation was approximately 61 m with a 4 % slope, and water was supplied by pumping from the detention pond. Three different ditch check products were evaluated in this study: Sediment Logs, GeoRidge plastic berms, and Triangular Silt Dikes. All products were evaluated according to the following methodology.

2.1 Field site soil and channel details

The testing channel had a parabolic shape that simulated the typical channel profile found in construction sites and roadside ditches, with a top width of 3 m at the upstream end and 7.9 m at the downstream end. The side slopes were 2(H):1(V) throughout the channel profile. Three soil samples were collected randomly from the channel to determine the soil texture and particle size analysis. The particle size analysis was done following the hydrometer method described by Gee and Bauder (1986). This method was a modification of the Day (1965) and ASTM (1985) methods. The analysis results indicated that the testing channel was composed of silt-loam soil as defined by USDA soil texture classification, with 13.0 % sand, 61.7 % silt, and 25.3 % clay.

The testing channel was divided into two zones: the discharge zone and the testing zone. The discharge zone received the water from the pumping station. The discharge zone measured 11 m in length and was stabilized with a turf reinforcement mat (TRM) and vegetation to minimize erosion and sediment concentration of the flow prior to reaching the testing zone, which consisted of the remaining 50 m of channel. To measure the flow rate, a Plexiglas 90° V-notch weir was installed across the width of the channel between the discharge and testing zone. The V-notch weir was installed to a depth of 1 m, stabilized by approximately 30 cm of cement, and then covered by compacted soil up to the channel bed. A diagram of the testing channel is presented in Fig. 1.

Fig. 1
figure 1

Diagram of evaluation procedure (not to scale)

Full size image

2.2 Testing protocol

The evaluation methodology required the following components: a water source and delivery system, a test channel with 4 % bed slope and 61 m length, soil stock pile, earthmoving and compacting equipment, scanning equipment, and camera.

Prior to product installation, the channel bed was prepared to provide uniform testing conditions. The bed was first loosened to a depth of approximately 10 cm and then smoothed and compacted along the ditch check installation area. Once the channel was prepared, two of the same ditch check products were installed in series in the testing channel according to the manufacturer’s guidelines. An area 2 m in length downstream from each product and along the entire wetted width of the channel (Fig. 2) was scanned using a laser distance meter (Leica 3D Disto). The scan pattern consisted of a rectangular grid of 10 by 10 cm spacing. Photographs were also taken during testing performance to evaluate potential product failure.

Fig. 2
figure 2

Elevation measurement area, 2 m in length downstream from each product and along the entire wetted width of the channel (not to scale)

Full size image

Each ditch check product installation was tested using three consecutive 30-min applications of the same flow rate, 5, 7.5, and 10 l s−1. The flow conditions can be reproduced in the field setting using commercially available pumps and were selected based on the peak flow-generated construction site of varying size in central Illinois, and the 5, 7.5, and 10 l s−1 flows corresponded to peak flow generated from 5, 7.5, and 10 acre construction area from 10-year rainfall event. Grab samples were collected at 5-min intervals from the upstream and downstream sides of each ditch check. Once the test was completed and any remaining water drained out of the channel, the second test under the same flow rate was performed, followed, in turn, by the third. All tests for the same flow rate were performed with no soil bed disturbance or product removal between them.

After the product was evaluated three times under one flow rate, the channel was completely drained, and the post-scan was performed. The post-scan operation followed the same scanning pattern as the pre-scan. For each additional flow rate, the channel was prepared again and a new product of the same type was installed.

Total solids concentration for each sample was measured based on the procedures in ASTM D3977-97 Standard Test Methods for Determining Sediment Concentration in Water Samples (ASTM 2007b). Turbidity in NTUs was measured for all grab samples, and the relative reduction in NTU between the upstream and downstream sides of each ditch check was computed. The total volume of sediment retained by the downstream ditch check was computed using the surface scans. The commercial software SURFER (Golden Software, LLC, USA) was used to plot the scanned surfaces and to estimate the total volume retained.

2.3 Statistical analysis

Welch’s t test was performed to compare the performance among the products (Welch 1947). This test allowed comparison of the performance between paired product data to gauge how likely two products’ results were drawn from the same population and determine whether the products performed the same. A test at the 90 % confidence level was conducted to compare performance between evaluations.

3 Results and discussion

3.1 Ditch check product evaluation

The ditch check product evaluation was performed using three different flow rates: a high flow of 10 l s−1, a medium flow of 7.5 l s−1, and a low flow of 5 l s−1. The products were tested under different flow conditions to evaluate how each one might affect the ditch check product performance. Higher total solids concentration (TSC) was expected for higher flows, which was observed for two of the products tested: the Triangular Silt Dike and Sediment Log. The trend reversed for GeoRidge, however, with higher TSC (and hence higher channel soil loss) found for lower flows.

Each product evaluation for a particular flow condition was repeated three times. The TSC values calculated from the 5-min grab samples appeared to show a trend of declining concentration with each performed test. The sediment concentration of grab samples seemed to consistently stabilize for all products after the first 15 min of each test.

The average TSC values computed from grab samples for the Triangular Silt Dike under the high flow condition are displayed in Fig. 3. The results showed a declining trend in TSC during the first 10 min of the evaluation, after which the TSC stabilized until the test was completed. Samples were taken at the upstream and downstream sides of each of the two ditch checks installed in series. The TSC value after 5 min is not shown for the downstream side of the downstream ditch check, however, because the flow at this location did not reach steady state until approximately 10 min after the test began. This behavior was only observed for the Triangular Silt Dike and was due to the specific characteristics of that product. In contrast to the Sediment Log and GeoRidge, the Triangular Silt Dike permeability was very low, which resulted in a significant flow barrier and created a series of ponded areas between the ditch checks installed along the channel. This diminished the energy slope and, in turn, the shear stress along the bottom of the channel, which prevented erosion in the channel bed and enhanced sediment settling.

Fig. 3
figure 3

Average TSC values for Triangular Silt Dike for 10 l s−1 flow rate

Full size image

The TSC values under the 10 l s−1 flow rate for the GeoRidge product evaluation showed a similar trend to the results obtained for the Triangular Silt Dike. In this case, steady-state flow occurred approximately 3 min after the test started; TSC values decreased during the first 10 min of evaluation, thereafter stabilizing to a constant value. The manner in which this product retained sediment and prevented erosion in the channel bed differed from the Triangular Silt Dike. The GeoRidge ditch check primarily reduced flow velocities, causing sediment to settle upstream of the ditch check. Reduced water velocity also resulted in decreased downstream erosion.

The Sediment Log results under the 10 l s−1 flow rate showed substantially higher TSC values throughout the entire test. The Sediment Log retained sediment and prevented channel bed erosion in a similar manner to the GeoRidge product. The Sediment Logs exhibited a lower permeability than GeoRidge, which dissipated more flow energy, augmented sediment settling upstream of the check dams, and minimized erosion downstream of the ditch check. The discrepancy in higher TSC obtained for the Sediment Log than GeoRidge, however, can be mainly explained by the flow undercutting that occurred during evaluation, which is discussed in section 3.3.

In order to compare the performance of the three ditch check products, the total soil loss was computed. This was calculated using the average TSC value over each 5-min period during testing, while discarding the first 5 min of each test to ensure steady-state flow conditions. The resulting values were used to compare total channel soil losses between each product. The average soil losses for the three different ditch check products are presented in Fig. 4. The difference in total soil loss between the Triangular Silt Dike and GeoRidge were statistically insignificant at a 90 % confidence interval, but the Sediment Log soil loss was significantly higher than each of the other two products.

Fig. 4
figure 4

Average sediment yield for three ditch checks under 10 l s−1 flow rate

Full size image

The average TSC results obtained from the 7.5 l s−1 medium flow exhibited the same trend as that observed for the high flow rate. Steady-state flow on the downstream side of the downstream Triangular Silt Dike occurred approximately 10 min after the test began (as in the high flow test), while for the GeoRidge and Sediment Log products, steady-state flow was reached around 3 min after testing started. The sediment yield for the 7.5 l s−1 test was significantly smaller than the yield for 10 l s−1 flow for all three products.

The last set of tests was performed under low flow conditions of 5 l s−1. Steady-state flow occurred after approximately 15 min for the Triangular Silt Dike, while it was reached in less than 5 min for the GeoRidge and Sediment Log.

The fact that steady state was not reached until after 15 min considerably affected the final results for the Triangular Silt Dike; grab samples could not be taken during the first 10 min, and the stable TSC trend observed under the previous two flow conditions was barely observed at the downstream side of the downstream ditch check. Failure to adequately consider this behavior could lead to erroneous interpretation of the results.

The total channel bed soil loss was significantly lower for the Triangular Silt Dike when compared to the GeoRidge and Sediment Log. Soil loss was also lower for all products when compared to the results obtained under higher flow rates. The total sediment yield was significantly smaller, because steady-state flow was reached approximately 15 min after the test started; hence, the total sediment yield was only computed for 15 min, while in all other evaluations, it was calculated for the last 20 min of testing. This prevents a reliable comparison of the Triangular Silt Dike results under 5 l s−1 flow with the other two products evaluated.

The average soil loss obtained was also observed to be significantly smaller under low flow than medium flow for all three products. There was no significant difference between the performance of the GeoRidge and Sediment Log, but comparison with the Triangular Silt Dike could not be performed. The total average soil losses from the product evaluations with three replications per test are presented in Table 1.

Table 1 Average channel bed soil loss for all products and flow rates
Full size table

3.2 Statistical analysis of ditch check evaluation

The low degree of variability among the replications permitted an accurate statistical analysis of the results obtained from the product evaluations using Welch’s t test. The p values obtained for the paired Welch’s t test are provided in Table 2 for 10, 7.5, and 5 l s−1 flow conditions.

Table 2 Average soil loss p values for pairwise comparison under all flow rate conditions
Full size table

Applying Welch’s t test to results from the 10 l s−1 flow condition indicated that the Triangular Silt Dike ditch check product performed significantly better than the Sediment Log ditch check product, while there was no significant difference when compared to the GeoRidge. Comparing results from the GeoRidge and Sediment Log evaluations indicated that GeoRidge products performed significantly better. Hence, one could confidently state that in terms of total sediment retention, the performance of the Triangular Silt Dike and GeoRidge was similar and significantly better than the Sediment Log.

The statistical analysis results from comparing the three products for 7.5 l s−1 flow rates differed slightly from that of the high flow evaluation. In this case, there was no significant difference between the Triangular Silt Dike and the Sediment Log or between the Sediment Log and the GeoRidge, but Welch’s t test did indicate that the Triangular Silt Dike performed significantly better than the GeoRidge. Even though the average soil loss obtained from the GeoRidge evaluation was smaller than that of the Sediment Log, the statistical analysis showed a significant difference only between the Triangular Silt Dike and GeoRidge and not the Triangular Silt Dike and Sediment Log. This was due to the larger variance obtained from the Sediment Log evaluation and the small variance from the GeoRidge evaluation. In order to overcome this issue, a larger number of replications would be necessary for the statistical analysis.

The results obtained from the evaluation under 5 l s−1 flow rates were inconclusive since the Triangular Silt Dike ditch check product could not be compared with the other two products. The total soil loss computed from the Triangular Silt Dike was less than the soil loss for the other two products largely because steady-state flow was not reached during the initial 10 min of the Triangular Silt Dike evaluation, which prevented collection of grab samples and produced a lower sediment yield. The statistical analysis of the other two products, however, showed no significant difference between the Sediment Log and GeoRidge performance.

The total sediment loss computed for the three ditch check products showed a noticeable difference between the different flow rates used during evaluation. Hence, the average TSC values after stabilization was reached for the different flow rate conditions were compared and discussed individually for each product, and the overall results are displayed in Table 3.

Table 3 Average TSC, after stabilization, was reached, for all products and flow rates
Full size table

Welch’s t test with a confidence level of 90 % was used for comparing average TSC values between products (Welch 1947). The results of these statistical comparisons are provided in Table 4. The p values obtained after performing this statistical test for the Triangular Silt Dike evaluation results revealed no significant difference in average TSC values between the different flow rates. Hence, the flow rate did not seem to significantly affect the average TSC values for that particular product.

Table 4 Average TSC p values for pairwise comparison of the three ditch check results under all flow rates
Full size table

For GeoRidge evaluation, the resulting p values indicated no significant difference between the average TSC under the 10 and 7.5 l s−1 rates, while there was a significant difference both between the average TSC under the 10 and 5 l s−1 flow rates and between the average TSC under the 7.5 and 5 l s−1 flow rates. The result indicates that the product is more effective for 5 l s−1 flow rate compared to higher flow rates. The statistical analysis of results for the Sediment Log showed no significant difference in the average TSC values either between the 10 and 7.5 l s−1 rates or between the 7.5 and 5 l s−1 flow rates. There was a significant difference between the values under 10 and 5 l s−1 flow rates, however, which indicated that the average TSC under 10 l s−1 was significantly greater than that under 5 l s−1. It can be inferred from this result that this product is more effective in TSC reduction for 5 l s−1 flow compared to 10 l s−1 flow.

Overall, the results did not show a strong relationship between flow rate and sediment concentration over the 5 to 10 l s−1 range. For the same comparison performed with a 95 % confidence level, there was only a significant difference between the average TSC values for the GeoRidge product under the 10 and 5 l s−1 flow rates.

In order to account for the sequential testing of replications for a given flow rate without channel repreparation, Friedman’s statistical test was performed for total soil loss (Hollander and Wolfe 1973). In this analysis, the unique pair of each flow rate and position in the testing sequence was treated as a unique block with a single observation. The null hypothesis was that no significant difference existed between the effects of all three products (apart from block effects), and the p value obtained was 0.00432. Hence, the effect of at least one of the ditch check products was different from the others at the 95 % confidence level.

Performing pairwise Friedman tests for the three products showed significant difference between the Triangular Silt Dike and Sediment Log and between the Triangular Silt Dike and the GeoRidge, with p values of 0.0027 and 0.01963, respectively, at the 95 % confidence level. However, there was no significant difference found between the Sediment Log and GeoRidge effects at even the 90 % confidence level, with a p value of 0.7389. These results confirm the previous statistical analysis, which showed the Triangular Silt Dike to perform significantly better than the other two products in terms of total soil loss.

3.3 Field observations and discussion

Sediment transport is a process that involves both suspended sediment and bed load transport. Ditch check products are intended to provide channel stabilization until vegetation can provide long-term channel soil protection. Prior to vegetation establishment, the soil in channels is highly erodible, and ditch check products are installed to prevent soil disturbance and reduce soil displacement.

Therefore, ditch check products should not only prevent sediment transport out of the construction site but also ameliorate the negative effects that soil displacement has on long-term channel stabilization. After channel disturbance, ditch checks are installed and the channel bed is seeded to provide long-term stabilization; if the soil is displaced from its original position, however, it will carry the seeds along with it, and the areas where soil displacement occurred will not be able to generate a vegetative cover for long-term channel protection.

Test observations, photographs taken before, during, and after product evaluation, and total sediment retention analysis were used to determine the product effectiveness in terms of channel bed disturbance and potential product failure during test performance. The upstream sides of both ditch checks in series were scanned to quantify the total sediment retained by the ditch checks. The scan covered an area measuring 2 m upstream of the ditch check and along the entire wetted width of the channel. Even though these results supported the visual observations, the results could not be used for product comparison. Ideally, the scan should be performed over the entire channel so that a complete mass balance can be performed to determine the total sediment leaving the area; this would also allow calculation of the volume of sediment displaced from its original position and permit comparisons between different products. Profiling the entire channel was not feasible in this study due to equipment and time constraints at the time the product evaluations were performed. However, software for the laser scanning distance meter has subsequently improved, and profiling the entire channel can now be accomplished in future evaluations.

The pre-test and post-test scans were used to compute the total accumulated sediment volume in front of the downstream ditch check. The upstream ditch check was scanned as well, but only the downstream one was considered for analysis, because it was more representative of those found at a typical construction site. However, it was still considered important to take photographs and perform scanning for the upstream ditch check to evaluate any potential product failure.

The total accumulated volume in front of the downstream ditch check was interpreted as the volume of soil displaced from its original position. This volume was calculated using SURFER software. The laser scanning distance meter was used to take topographic measurements on a 10 by 10 cm grid of the area prior to the start of testing and after completion of each test. The grid of elevation measurements was then interpolated onto a surface using kriging. Once the surfaces of the scanned areas were obtained, the volume was computed by overlaying the surfaces.

Very little soil disturbance was observed under the three flow rates for the Triangular Silt Dike evaluation, and sediment accumulation in front of the Triangular Silt Dike was barely noticeable. On the other hand, soil disturbance and sediment accumulation were easily observed for both the Sediment Log and GeoRidge products under all three flow rates. The calculated volume of accumulated sediment in front of the downstream ditch check supported the visual observations recorded during testing. Therefore, the total volume of accumulated sediment was used as an estimate of channel bed disturbance, and results for all products and flow rates are presented in Table 5.

Table 5 Total volume of accumulated sediment (cm3) in front of downstream ditch check for all products and flow rates
Full size table

Photographs of the front side of the downstream GeoRidge and associated surface scans before and after testing under the 10 l s−1 flow rate are displayed in Figs. 5 and 6, respectively.

Fig. 5
figure 5

a Photograph of downstream Triangular Silt Dike prior to testing for 10 l s−1 flow rate. b Associated scanned profile for front side of downstream Triangular Silt Dike

Full size image
Fig. 6
figure 6

a Photograph of downstream Triangular Silt Dike after testing for 10 l s−1 flow rate. b Associated scanned profile for front side of downstream Triangular Silt Dike

Full size image

Product failure was not observed for any of the three products tested under the selected flow rates. Undercutting was only noticeable for the Sediment Log under the 10 and 7.5 l s−1 flow rate conditions but was not severe enough to cause product failure. The undercutting observed for the Sediment Log ditch check product during the 10 l s−1 flow rate evaluation is displayed in Fig. 7.

Fig. 7
figure 7

Undercutting observed with Sediment Log product evaluation for 10 l s−1 flow rate

Full size image

3.4 Relationship between total suspended solids, total sediment concentration, and turbidity

The direct measurement of TSS cannot be performed in the field, as it is necessary to collect grab samples and then determine TSS in the laboratory. One of the less expensive and more efficient methods utilized to predict TSS is the turbidity. Turbidity measurements have been utilized in diverse environments. Patil et al. (2011) developed a linear regression model to predict TSS from turbidity measurements for each primary particle class. Packman et al. (1999) developed a model to determine TSS from turbidity measurements in urbanized streams in the Puget Lowlands. Data collected from nine streams with both urbanized and rural drainage areas showed a strong log-linear relationship between turbidity and TSS. However, turbidity is also affected by many other parameters other than particle concentration. Water color may be affected by dissolved solids and temperature (Aiken et al. 1985), as well as the shape, size, and mineral composition of particles (Clifford et al. 1995; Gippel 1995), all of which can significantly affect turbidity readings.

While other authors (Packman et al. 1999; Patil et al. 2011) have developed regression models to estimate TSS from the NTU measured, the suitability of development of a regression model to estimate TSC from the measured NTU was analyzed. The purpose of this analysis was to develop a simple regression model to evaluate whether turbidity reading could also be used to accurately estimate the total solid measurements for the ditch check, analogous to those previously developed for TSS discussed above. All ditch check data is plotted in Fig. 8 with a semi-logarithmic regression model. The semi-logarithmic model seems to be a more convenient model for broader ranges of data. Different particle sizes along with other water properties can influence the linear relationship between the NTU and TSC and being only valid for a smaller range of the data.

Fig. 8
figure 8

TSC and NTU data for ditch checks with semi-logarithmic regression model

Full size image

The statistical analysis showed a strong relationship between TSC and NTU with R 2 value of 0.85. Therefore, a relationship can be proposed to determine TSC from NTU values as given below:

$$ \mathrm{T}\mathrm{S}\mathrm{C} = 133\ \ln\ \left(\mathrm{N}\mathrm{T}\mathrm{U}\right) + 129.3 $$

This equation can be used to quickly get sediment concentration information from a construction site with just NTU measurement, which is not difficult to obtain.

4 Conclusions

Field-scale evaluation protocols for sediment retention devices were successfully developed and implemented at the ECRTC at the University of Illinois at Urbana-Champaign. These evaluation protocols were developed based on current studies and were designed to be reliable, easily replicable in similar testing facilities elsewhere, and amenable to any adverse weather effects. Three different products were evaluated under three flow conditions, and based on observations made during the study, the Triangular Silt Dike generally performed better than the GeoRidge or Sediment Log.

The surface scans only provided information about the sediment being retained but did not permit a complete mass balance analysis to quantify the total sediment loss in the channel. A complete scan of the area between the two installed ditch checks was not feasible at the time of testing, but the equipment software has been upgraded recently so that such operation can now be performed. Including these scans will permit accurate estimation of both the total soil loss and soil disturbance in the channel bed in future evaluation. Finally, the ditch check products were evaluated under different flow rate conditions for fixed soil type and channel slope conditions. It is recommended that future studies also test the products under different soil types and slope conditions that are commonly found in construction sites.

Additional product evaluations and testing protocols for sediment retention are under development for future studies at the ECRTC as well. This study also provided guidance to the IDOT in the installation and maintenance of sediment control and erosion control products. These protocols will continue to provide reliable product performance data to help IDOT select and modify approved erosion and sediment control products for Illinois construction sites.