EROSIONAL STABILITY OF SEDIMENT DEPOSITS IN HAMILTON HARBOUR Bommanna Krishnappan, Ian Droppo, Niels Madsen and Robert Stephens, Environment Canada, Canada Centre for Inland Waters, Burlington, Ontario, Canada ABSTRACT Erosional stability of sediment deposits in Hamilton Harbour, Ontario, Canada was studied in a rotating circular flume. Sediment cores from the Harbour were collected using a coring device developed specifically for collecting a layer of sediment near the sediment-water interface, and were tested in the flume by mounting them in a special device that exposed the sediment to the flow near the flume bed. Critical shear stress required to erode the top layer of sediment and the erosion rate were obtained by operating the flume at different bed shear stress levels until the erosion of the sediment layer was initiated and continued. The critical bed shear stress and the erosion rate values obtained from the flume deviated from a set of measurements carried out using an in-situ erosion flume carried out in a previous study. Possible explanations for the for the discrepancy are explored. RÉSUMÉ Nous avons étudié l érodabilité des dépôts de sédiments du port de Hamilton, en Ontario (Canada), dans un canal expérimental circulaire. Nous avons prélevé des carottes de sédiments dans le port à l aide d un dispositif de carottage spécifiquement conçu pour la collecte d une couche de sédiments près de l interface eau-sédiment. Nous avons ensuite soumis ces carottes à des essais dans le canal en les montant dans un dispositif spécial qui permettait d exposer les sédiments à l écoulement près du lit du canal. La contrainte critique de cisaillement nécessaire pour éroder la couche superficielle des sédiments et le taux d érosion ont été déterminés en soumettant le lit du canal à différents niveaux de contraintes de cisaillement jusqu à ce que l érosion de la couche de sédiments s amorce et se poursuive. Les valeurs de la contrainte critique de cisaillement et du taux d érosion, obtenues dans le canal, présentaient un écart par rapport à un ensemble de mesures relevées lors d une étude antérieure à l aide d un canal d érosion in situ. Nous recherchons de possibles explications pour justifier cet écart. 1. INTRODUCTION Remedial actions such as dredging, capping and in-situ treatment using stabilizing chemicals for cleaning up of contaminated sediment sites require knowledge of erosional stability of sediment deposits and the erosion rate of sediment for a range of bed shear stresses. With the present state of knowledge, such data can only be obtained empirically using measurements either in the field or in the laboratory. In this study, a methodology developed for testing stability and erodibilty of sediment deposits in Hamilton Harbour, Ontario, Canada is described. The methodology consisted of collecting samples of sediment deposits from the harbour using a sampling device suitable for collecting an undisturbed sediment core at the sediment water interface, and testing the samples in a rotating circular flume located in the Hydraulics Laboratory of Environment Canada in Burlington, Ontario, Canada. Details of the sampler, flume and the experimental results are given in this paper. 2. DESCRIPTION OF THE SAMPLER A drawing of the sampler is shown in Figure 1. Figure 1. A drawing of the sediment core sampler used in this study. As can be seen from this figure, the cylindrical sampler consists of two sections. The top section, which is made out of plexiglass, is 2 cm in diameter and 5 cm long. This section is for sampling the overlying water, and it connects to the bottom stainless steel section which holds the sediment core. The length of the 1123
stainless steel section is 1 cm and it has a slot for inserting a knife to cut the sediment core. The sampler has to be operated by a Scuba diver. After pressing the stainless steel chamber into the mud, the diver inserts a knife edge into the slot in the cylinder to cut the sediment deposit and to hold the deposit inside the cylinder. The sampler is then brought to the surface, when the knife is removed and a plug is inserted to hold the sediment core inside the cylinder until the sampler is brought to the flume for testing. The sampler fits inside a sample holder in the flume (see Figure 4). There is a provision in the sample holder to push the sample up into the flow as the sample is eroded under the influence of the flow field in the flume. 3. DESCRIPTION OF THE FLUME The laboratory flume used in this study is a rotating circular flume with a mean diameter of 5. m, depth of.3 m and a height of.3 m. The circular flume is supported on a rotating platform, which is 7. m in diameter. A counter rotating annular ring fits inside the flume with close tolerance (~1.5 mm gap on either side). When the flume and the top cover are rotated in opposite directions, a nearly two dimensional flow field is generated. A schematic view of the flume is shown in Figure 2. Figure 3. Bed shear stress distributions of flows generated in the rotating circular flume. In this figure, the bed shear stresses are plotted as a function of transverse distance across the width of the flume with rotational speed, N f as a parameter. The ring speed, N R had to be slightly larger than the flume speed in order to minimize the secondary circulations. For the tests shown in this figure, the ratio N R to N f was maintained at 1.167. The flume bed surface was smooth and the depth of water in the flume was.12 m. The points are measured values and the lines represent predictions from a three dimensional, hydrodynamic model called PHOENICS (Rosten & Spalding (1984)). It can be seen from this figure that the model predictions agree very well with the measured data. The model is used as an operating tool to predict the relationship between the bed shear stress and the rotational speeds of the flume assembly under both smooth and rough bed conditions (Krishnappan & Engel (24)). 3.2 Description of the sample holder Figure 2. Schematic view of the flume used in the study. The flume is fitted with a sample holder to accommodate the sampler. A hole in the bottom of the flume was cut with a diameter equal to that of the sampler, and a housing containing a piston to push the sample into the flume was mounted underneath the flume. A top view through the hole and the piston assembly are shown in Figures 4 and 5 respectively. 3.1 Flow characteristics in the flume The flow characteristics generated with this flume were studied experimentally (Krishnappan, 1993) and theoretically (Petersen and Krishnappan, 1995; Krishnappan and Engel, 24). The bed shear stress distributions across the width of the flume are depicted in Figure 3. 1124
Sea to Sky Geotechnique 26 3.3 Description of sampling sites Sediment deposit samples were collected at two locations in Hamilton Harbour (Figure 6). Figure 4. A hole cut in the flume to accommodate the sampler. Figure 6. Location map showing sampling sites. Site 1 is a shallow water site with a depth of about 2 m. Site 2 is a deep water site with a depth of about 6 m. Sediment in this harbour is contaminated and remedial action plans are underway (RAP Report, 1992). A photograph of the sample collected at one of these locations is shown in Figure 7. Figure 5. A view of the sample holder assembly. Figure 7. A photograph showing the sample collection process. 1125
3.4 Testing of sediment cores in the flume The sediment cores were tested in the flume for erosional stability and erodibility. To begin the test, a sediment core sampler with a sample was inserted into the cut-out in the flume. The flume was then filled with water from the harbour until the plexiglass section of the sampler was fully immersed. The plexiglass section of the sampler was then removed taking care that the disturbance to the sample was minimized. Fill up of the flume continued until the depth of water in the flume was 12 cm. The flume was then operated at different shear stress steps. At each shear stress step, the concentration of sediment in the water column in the flume was measured by collecting water samples at regular intervals of time. The results for the two samples are shown in Figures 8 and 9 respectively. 4.1 Critical shear stress and average erosion rates The data shown in Figures 8 and 9 are useful for determining the critical shear stress for erosion of the top layer of the sediment deposit and the erosion rate as a function of shear stress when the shear stress exceeds the critical value. For example, both figures show that that there is no eroded sediment in suspension until the bed shear stress reaches a value of.21 Pa, and hence the critical shear stress for erosion of the top layer of the sediment deposits can be considered as.21pa. An average erosion rate can also be computed from the data shown in these two figures. Based on mass balance considerations, an average erosion rate can be computed as: ( C C ) V ( A t) E = / (1) f in Concentration in mg/l 18 12 6 Concentration Shear stress.5.45.4.35.3.25.2.15.1 Shear stress Pa where C f and Cin are concentrations (in mg/l) of the eroded sediment in the water column at the end and at the beginning of a shear stress step respectively, V is the volume of water in the flume in litres, A is the cross sectional area of the sediment deposit in square metres and t is the duration of the shear stress in seconds. The average erosion rates computed using Equation 1 are plotted for the two sites in Fig. 1..5 4 8 12 16 2 24 28 32 36 4 Time in Minutes Figure 8. Concentration of eroded sediment as a function of time for different shear stress steps- core from Site 1. Concentration in mg/l 36 3 24 18 12 6 Concentration Shear stress. 4 8 12 16 2 24 28 32 36 4 44 Time in Minutes..5.45.4.35.3.25.2.15.1.5 Figure 9 Concentration of eroded sediment as a function of time for different shear stress steps-core from Site 2. 4. RESULTS AND DISCUSSION Shear stress Pa Average erosion rate in Kg/m^2 sec.9.8.7.6.5.4.3.2.1 Site 2 Site 1.1.2.3.4.5 Shear stress in Pa Figure. 1. Average erosion rate as a function of bed shear stress for the two sites. From this figure, we can see that the erosion rate variation for the two sites shows different behaviour. For Site 1, the average erosion rate is fairly constant at around.1 kg/m 2 s, whereas it shows a large variation for Site 2. At lower shear stresses, the average erosion rate is slightly lower than those for site 1, but for large bed shear stress (around.4 Pa), the erosion rate increases considerably reaching a value as high as.8 kg/m 2 s. The spatial variability in the erosional characteristics of the sediment deposits in the Harbour may be due to differences in the water content or dry density of the sediment deposits. These 1126
differences can arise because of bioturbation and/or accumulations of gases. 4.2 Dry density profiles and depths of erosion To examine the variability in the density profiles of the sediment deposits, density measurements were made for the two cores using an ultra sonic high concentration measuring instrument manufactured by the Delft Hydraulics Laboratory in the Netherlands. The dry density profiles measured for the two samples are shown in Figure 11. Distance from bed surface in mm.2.4.6.8 1-5 -1-15 -2-25 -3-35 Site 1 Site 2 Dry density in gm/cc extrapolated Figure 11. Dry density profiles for the two sediment deposit samples. The density in the top 5 mm of the deposit was not measured because of the limitation of the instrument and hence it has to be extrapolated as shown in Figure 11. The density profiles for the two cores are very different. Site 1 core shows more or less a linear variation of density as a function of depth while Site 2 core exhibits a more complex behaviour. It shows an increasing trend upto a depth of about 1 mm, and then it decreases before it starts to increase again resulting in a hump in the density profile. Knowing the density of the deposit and the amount of the eroded sediment, it is possible to calculate the depth of erosion at each shear stress step. Such calculations were carried out and the results are summarized in Tables 1 and 2. Table 1: Summary of computed results for Site 1 Shear stress (Pa) Dura -tion (min) Amo -unt (gm) Density (g/cc) Cum.21 3 2.21.39.2.2.27 5 1.4.4.8.1.33 5 6.62.41.5.15.39 5 1.4.42.9.24.46 7 6.2.43.5.29 Table 2: Summary of computed results for Site 2 Shear stress (Pa) Duration (min) Amount (gm) Density (g/cc) Cum.21 3 3.9.39.3.3.27 5 3..4.2.5.33 5.47.4..5.39 5 8..4.7.12.46 8 119.43.97.19 From these two tables we can see that the computed depth of erosion values are only a fraction of a millimetre and are significantly smaller than the values measured in a previous study using an in-situ erosion flume (Krishnappan and Droppo, 25). In that study, depths of erosion of the order of millimetres were measured for the same range of bed shear stresses. Possible explanations for the discrepancy may be 1) the sample might have undergone consolidation during the transport from the sampling location to the flume, 2) degassing of the sample might have occurred thereby consolidating the sample, 3) inactivation of organisms responsible for bioturbation and 4) possible biofilm development. More work is needed to confirm these hypotheses and to collect a representative sample for testing in the flume. 1127
5. SUMMARY AND CONCLUSIONS A new sampling device to collect undisturbed sediment cores in a water body was deployed in Hamilton Harbour, Ontario, Canada. The sediment cores were tested in a rotating circular flume of Environment Canada in Burlington, Ontario, Canada. The test results were useful to determine the critical shear stress for erosion of the top layer of the sediment core and the erosion rate as a function of bed shear stress for two cores collected at two different locations. The erosion rates determined for the two cores were different. This was attributed to the difference in the density of the sediment deposits. The difference in density can arise because of non homogeneity in bioturbation, gas accumulation, biofilm development/degradation etc The erosion rates obtained in this study are somewhat lower than the values obtained in a previous study using an in-situ erosion flume. This may be due to a possible consolidation of the sediment cores due to transport of the cores from the sampling site to the flume, degassing, inactivation of bioturbation and biofilm development. This hypothesis needs to be tested in future studies. 6. ACKNOWLEDGEMENTS The authors would like to acknowledge the assistance of Mr. Bruce Gray and Mr. Ken Hill of the Technical Operations Service of Environment Canada in collecting the sediment cores. References Krishnappan, B. G. 1993. Rotating Circular Flume. Journal of Hydraulic Engineering, ASCE, 119 (6): 658-767. Krishnappan, B.G. and Engel, P. 24. Distribution of bed shear stress in Rotating Circular Flume. Journal of Hydraulic Engineering, ASCE, 13: 324-331. Krishnappan, B.G. and Droppo, I.G. 26. Use of an in-situ erosion flume for measuring stability of sediment deposits in Hamilton Harbour. Water, Air and Soil Pollution (in press). Peterson, O. and Krishnappan, B.G. 1994. Measurement and analysis of flow characteristics in a rotating circular flume. Journal of Hydraulic Research, IAHR, 32(4): 483-494. Remedial Action Plan (RAP) for Hamilton Harbour, 1992, The Remedial Action Plan: Goals, Options and Recommendations, Stage 2A Report, Volume 2, ISBN Number -7778-533-2. Rosten, H.I. and Spalding, D.B. 1984. The PHOENICS reference manual, TR/2, CHAM Ltd., Wimbledon, London. 1128