FLOOD1 report: Appendix 3

Transcription

1 FLOOD1 report: Appendix 3 Summary of Laboratory Testing Programme Part of the investigation into the behaviour of groundwater in the unsaturated zone of the Chalk required testing samples to determine what variation in chalk properties were present, where this variation occurred in the rock column and how these variations affected the way water was stored and moved (mechanisms), both in the rock material (different types of chalk) and in the rock mass (fractures plus material). As with other aspects of the investigation it was uncertain what type of tests would be most appropriate and how many tests to carry-out. The approach taken, therefore, was to focus initially on numerous index tests to determine the range of variation and then, select fewer samples in well defined lithologies on which to carry out more complex and time-consuming tests. The testing program began with well established techniques and progressed towards the development of novel experiments. As in the case of the geological surveys of the catchments the testing was an iterative process of evaluation and refocusing at each stage. Figure 1. The test rigs at the University of Brighton used to perform the FLOOD 1 tests. Level 1 tests included Saturation Moisture Content (SMC %), Intact Dry Density (IDD Mg/m3), Porosity (%) and Void Ratio. In total, 390 index tests were carried out covering all the formations in the catchments. The tests were carried out in accordance with British Standards (BS 1377: 1990). The cores used in the Level 2 tests also formed part of the Level 1 index tests. 1

2 Level 1 test results (Figures 3 & 4) show that the full range of chalk lithologies, in terms of their density classification, have been tested. As density is an indication of porosity there was some concern that actual field densities might not be the same as laboratory tested densities because of the potential influence of overburden pressure in the saturated chalk. To test this idea density was measured during effective stress testing at stress levels considered typical of the burial depths of the samples in the unsaturated zone (Figure 4). The results indicate that there is less than 2% difference in density between laboratory tested samples at ordinary atmospheric pressure and samples tested up to pressures equivalent to depths of 60 m in the aquifer. Figure 2. An example of the sample frequency for testing the FLOOD 1 Chalk, the East Ilsley E12 Cored Borehole showing the horizons sampled for Level 1 index tests and Level 2 core plug tests. Despite the good correlation between IDD and SMC (Figure 3), there are samples significantly deviating from the correlation, suggesting some other factor such as texture may be involved. 2

3 Figure 3. Results of index tests (IDD and SMC) for the FLOOD 1 samples obtained from the cored boreholes at the experimental sites and the cores provided by Union Railways (North) Ltd (CTRL), Lewes CSO and the A303 Stonehenge. The CIRIA density divisions (low medium and high density) are shown. Figure 4. Measurement of the change in density caused by effective stress (i.e. depth of burial in the unsaturated zone). The in situ (field depth) density values vary by less than 2% the values for intact dry density obtained in the laboratory. 3

4 Level 2 tests included saturated permeability. Test cores were drilled and cut to the international standard length to diameter ratio (L:D) of 2:1 (80 mm x 40 mm). The test proceeded through three stages. In the first stage core plugs were saturated with de-aired water under vacuum pressure. In the second stage samples were consolidated hydrostatically at effective stresses corresponding to their depth in the ground (borehole cores). Then the samples were subjected to at least four flow increments (Figure 5) for each sample. Results were corrected for system head loss and temperature. Figure 5. Samples were subjected to four flow increments and results plotted as flow versus hydraulic gradient and corrections made for system head loss and temperature. Comparing permeability test results with Intact Dry Density (IDD) (Figure 6) shows a poor correlation which confirms the general view that permeability cannot be defined by simple index tests. Factors such as material grain size, pore structure, pore-throat diameters and micro-fractures also need to be investigated. Many of these factors are stratigraphically distributed through the Chalk hence there is some indication of permeability values for the material (not field permeability) also being stratigraphically distributed (Figure 7). The lowest permeabilities occur in the marly chalk lithologies (West Melbury Marly Chalk) and in the crystalline nodular and marly chalks in the Holywell and base Lewes chalks. The highest permeabiliteies occur in the purest white chalks with low densities and high porosities. The greatest dispersion of permeability results (Figure 6) tends to occur in the medium and high density materials. These medium and high density chalks have less pore-filling carbonate crystals than the very high density chalks but retain a stronger bonded structure than the low density chalks, maintaining better pore-throat connections (e.g. Mortimore & Fielding, 1990). Hence texture appears to be a significant controlling factor. 4

5 Figure 6. Permeability compared with Intact Dry Density (IDD) shows a poor correlation suggesting factors other than simple index properties of chalk are involved. Figure 7. Permeability distribution from core samples in the Patcham catchment, Brighton. 5

6 To investigate further the influence of texture on permeability and mechanisms of flow in the chalk, novel laboratory experiments using tensiometers on core samples were developed (Figure 8). The idea was to develop techniques that could measure the rate of movement of water through different types of chalk and on surfaces such as fractures. To do this, some way of measuring pore pressure changes as samples were progressively saturated needed to be manufactured. Trial tensiometers became available in the latter part of the research programme. Working with the manufacturers of the tensiometers, Wykham Farrance, the technique was applied to standard sized chalk cores to see if they would work in this environment (Figure 8). As this was a new technique using new instruments it was not surprising that some malfunctioned. Sufficient results were finally obtained to provide a unique insight into the way water moves in different chalks in relation to critical pore pressure controlling air entry under draining conditions and development of a water film on a surface during saturation experiments. Figure 8. Laboratory tensiometers attached to either end of the SC7C core sample. The graphs show the results from testing mass loss compared with time (upper graph) and a drying test curve (degree of saturation compared with porewater pressure) for an open tensiometer compared to a tensiometer covered with a membrane (lower graph). Figure 9. Experimental set-up used to measure mass change as porewater pressure changed. 6

7 Figure 10. Results from the experiment to measure mass changes compared with suction pressure. Figure 11. Using the mass change results divided by the surface area it is possible to show the water-film thickness reduction as pore-water suction pressure increases. Figure 12. Results from comparing degree of saturation with pore throat diameters in sample SC7C illustrating the air entry point at a negative pore pressure of -75 kpa. 7

8 Figure 13. Two plots of results from tensiometer results used to (1) measure surface pore-throat diameters (2) show that water-film thickness on the surface of the sample, generated at negative pore pressures, appears to be related pore size (or pore throat diameter) rather than surface roughness of a fracture. Figure 14. Drying curves for U.K. FLOOD 1 Chalk samples showing increasing negative pore pressure as degree of saturation reduces. 8

9 Wetting and drying curves are often used to provide a measure of pore throat diameters and the rates at which water will be imbibed or released in geological settings. The drying curves for the U.K FLOOD 1 samples (Figure 14), show increasing negative pore pressure as degree of saturation reduces. The curves vary depending on lithology. Summary The results from the FLOOD 1 laboratory testing carried out by RAs at the University of Brighton have identified 1) Saturated matrix permeabilities for the FLOOD 1 chalk samples have a range of 10-8 to m/s 2) Saturated matrix permeability has a poor correlation with index properties (intact dry density and porosity) because of structural and textural features in the chalk 3) Effective stress conditions (depth of burial of the chalk sample (overburden stress)) do not change saturated matrix permeabilities significantly 4) Saturated matrix permeability shows a significant correlation with the Chalk lithostratigraphy 5) Novel tensiometer tests have shown the change in pore pressures related to degree of saturation that control whether a film of water develops on chalk surfaces 6) Water-film thickness generation at negative pressures appears to be related to pore size rather than surface roughness of a fracture 7) Pore size may prove to be stratigraphically controlled and, therefore, fracture film-flow may be predictable by lithology These laboratory test results emphasise the importance of producing good lithological geological maps of the Chalk aquifer and sampling carefully in relation to the lithostratigraphy for laboratory testing. In this way the laboratory tests can be integrated into the field geology and increasingly refined models of the aquifer produced. The laboratory test results also link directly to the field tensiometer test results obtained at the FLOOD 1 experimental sites in showing at what negative pore pressures to expect air entry or to expect the development of films of water on surfaces of fractures in different types of chalk. Such measurements will help to identify the triggering of pore-water release and, combined with other data, will form an important part of future groundwater FLOOD predictions and warnings. 9