Groundwater Control Trials

Groundwater control trials are an extension of the ethos of pumping tests in that they are large-scale in situ tests to determine the hydraulic properties of the ground. Furthermore, a carefully planned trial can provide other useful information.

Instead of a single test well, a groundwater control trial involves pumping from a line or ring of wells of some sort (wellpoints, deep wells or ejector wells). Obviously, to specify a suitable trial, the main dewatering design must be reasonably advanced to allow selection of the appropriate pumping method, well spacing, well screen depth, etc. As with a pumping test, the trial is pumped continuously for a suitable period (typically 1 to 4 weeks) while monitoring discharge flow rate and water levels in monitoring wells. The results of such tests are normally analysed using the design methods of Chapter 13 to !back-calculate’ an equivalent soil permeability. Powrie and Roberts (1990) describe an example of a trial using ejector wells.

In addition to determining the permeability, trials provide opportunities to investigate other issues relevant to the proposed works:

Test type

Typical duration

Outline of typical test and test objective

Phases of conventional single well pumping test

Equipment test

15-60 minutes

Short period of continuous pumping. Objective is to confirm pumps, pipework, etc. are functioning and to inform setting of pumping rates for later phases.

Yield test

1-8 hours

Single well pumped at nominal constant rate. Objective is to estimate well yield.Test duration is too short to determine hydrogeological conditions as reliably as can be achieved by longer constant rate test phases.

Step test

8-12 hours

Single well pumped in stepwise fashion with increasing flow rates (typically 60 to 100 min per step). Objective is to estimate well performance, including well yield.

Constant rate pumping test

1 -60 days

Single well pumped at nominally constant rate. Objective is to assess drawdown in aquifer over a wide area and allow derivation of hydrogeological parameters and boundary conditions.Typical test durations up to 7 days for groundwater control design. Longer duration tests (>7 days) are more relevant to projects where external environmental impacts are a concern.

Other types of pumping test

Pumping and re-injection test

2-28 days

Water pumped from well(s) and re-injected to other well(s).Typically, re-injection is via recharge wells located a significant distance from the pumped wells. Objectives are same as for constant rate test plus estimating recharge well capacity and assessing interaction between abstraction and recharge wells.

Dewatering trial/ pilot test

5-28 days

A group of wells, typically forming all or a sub-section of the proposed dewatering system, is pumped. Objective is to investigate the effectiveness of the full system (see Section 12.8.6).

Cut-off wall pumping test

1-5 days

A well or sump within the area enclosed by a cut-off wall is pumped. The objective is to investigate the effectiveness of the cut-off wall. Typically, groundwater levels are observed in monitoring wells outside the cut-off walls to observe any external drawdown (see Section 18.14).

Adapted from Preene, M, et al„ Pumping tests for construction dewatering in Chalk, in Engineering in Chalk: Proceedings of the Chalk 2018 Conference (Lawrence, J A, Preene, M, Lawrence, U L and Buckley, R, eds), ICE Publishing, London, pp. 631-636,2018.

(ii) If the trial consists of a ring of wells, a trial excavation can be made inside the ring during the trial. This excavation can provide data about ease of excavation of the soil; stability of dewatered excavations; and trafficability of plant across the dewatered soil. There have been cases where trial excavations have been combined with large-scale compaction tests to assess the suitability of the excavated material as backfill elsewhere on site.

Figure 12.20 shows examples of groundwater control trials on sites where rings of well- points were installed and the area within the ring excavated during pumping. This type of trial can not only provide information on the feasibility of dewatering at the site; it can also allow battered side slopes to be cut at various trial slopes and provide useful information on the handling characteristics of the dewatered soil.

Groundwater control trials are probably the most expensive form of in situ permeability test, but for large projects in difficult ground conditions, they can reduce risks of problems during the main works, and the cost may be justified on that basis. If the dewatering system is to be designed by the observational method (see Section 13.3.2), a trial can be a key part of the scheme.

Wellpoint dewatering trial,

Figure 12.20 Wellpoint dewatering trial, (a) Ring of wellpoints installed around trial area. A 20 by 20 m rectangular ring of wellpoints has been installed around the trial area. Local excavations are made within the dewatered area to assess the behaviour of the dewatered soil when excavated, (b) Battered excavation inside the wellpoint ring with sides at various trial slope angles. A trial excavation is made within a 40 by 40 m rectangular ring of wellpoints in a silty sand. Trial batter slopes are formed to assess the stability of the soil following groundwater lowering. (Courtesy of WJ Groundwater Limited, Kings Langley, UK.)

Borehole Geophysics

Borehole geophysical methods (Section have traditionally been used extensively in hydrogeological studies for the development of groundwater resources but are used less commonly in investigations for construction projects. These methods do not generally give direct estimates of permeability values; as discussed in Section 12.5.3, they are typically used to provide non-quantitative data to allow indirect estimation of permeability and identification of zones of relatively higher or lower permeability.

However, there are some borehole geophysical techniques that can provide quantitative values of permeability. In this approach, special sensors (termed ‘sondes’) are slowly lowered into a borehole using a cable and winch (Figure 11.3). Two methods are discussed in the following sub-sections: borehole magnetic resonance (BMR) and borehole flowmeter logging.

Borehole geophysics is a specialized technique, and results are sensitive to background hydrogeological conditions and the measurement techniques used. Such investigations should be designed, executed and interpreted by suitably experienced specialists.

Borehole Magnetic Resonance (BMR)

Borehole magnetic resonance (BMR) is a technique that has been widely used in the oil 8c gas industry for more than two decades to evaluate reservoir properties. This technique is now sometimes used for investigations for construction and mining projects. As the sonde is lowered down the borehole, a magnetic field is used to cause resonance of hydrogen nuclei in the surrounding strata (Behroozmand et al., 2014). This allows the sensing method to determine porosity and the proportion of water in the pores that is potentially mobile and immobile (bound to clay particles or held in small pores by capillary forces). Processing of the data can allow direct estimates of permeability. The continuous nature of the logging process means that a continuous trace of permeability with depth is generated. In combination with other data sources, this can be useful in assessing changes of permeability with depth.

Borehole Flowmeter Logging

Flowmeter logging uses a sonde to measure the vertical velocity of water flow in a borehole, which can identify zones of inflow and outflow. The most common technique uses a flowmeter sonde with an impellor that rotates to record vertical water velocity. The sonde is lowered into and out of the well at a constant winch speed, providing a continuous record of vertical water velocity; an example log is shown in Figure 12.21a. An alternative approach is the heat pulse flowmeter, where a sonde is lowered to a specific depth and held there for a few minutes (to allow water flow to stabilize). The sonde then generates a heat pulse, and vertical water velocity is determined from the time taken for the pulse to reach temperature sensors along the length of the sonde. The sonde is then moved to a different depth and the process repeated.

A vertical trace on a flowmeter log indicates constant velocity, which implies no inflow or outflow to that section of the borehole. Changes in the velocity trace indicate inflow at that section of borehole, with the inflow rate proportional to the slope of the velocity line. As the sonde is lowered down the borehole, the velocity will reduce progressively as each inflow zone is passed.

Flowmeter logging is normally carried out after at least 24 hours have elapsed since the most recent work on the borehole (either drilling or development) - this is to allow the establishment of stable groundwater flow conditions. Logging is then done under unpumped (static) conditions, when the vertical water flow is due to natural movement only. Unfortunately, the small velocities typically observed can make inflow zones difficult to differentiate. A more useful approach is to log under pumped conditions. This requires the installation of a temporary pump in the upper section of a borehole with the flowmeter logs run beneath the pump. This gives a greater velocity contrast along the borehole and makes it much easier to identify inflow zones, as illustrated in Parker eta/. (2010,2019). The vertical flow velocities can be combined with the borehole diameter (measured by calliper log) to provide values of vertical flow rate. This can allow the borehole inflow to be assessed from the difference in vertical flow rate between discrete depths. If the overall transmissivity (permeability multiplied by aquifer thickness) has been determined (e.g. from a pumping

Examples of geophysical logs using impellor flowmeter to record vertical water velocity in boreholes,

Figure 12.21 Examples of geophysical logs using impellor flowmeter to record vertical water velocity in boreholes, (a) Impellor flowmeter log from two wells in fractured Chalk bedrock tested under unpumped and pumped conditions, Beckton, East London, United Kingdom. The change in velocity under pumped conditions indicates significant inflow at approximately 90 to 91 m depth, (b) Annotated impeller flowmeter log with assessed permeability based on velocity data, Storebaelt, Denmark. (From Roberts, T О L and Hartwell, D J, Chalk permeability, in Engineering in Chalk: Proceedings of the Chalk 2018 Conference (Lawrence, J A, Preene, M, Lawrence, U L and Buckley, R, eds)., ICE Publishing, London, pp. 425-430, 2018. With permission.)

test), the vertical flow rate profile can be used to allocate the transmissivity proportionally to different depth zones of the borehole. The thickness of each zone can then be used to derive a permeability for each zone of transmissivity.

Figure 12.21b shows an example of a flowmeter log assessed in this way. Roberts and Hartwell (2018) state that analysis of a pumping test gave a transmissivity that implied an average permeability of 10-5 m/s over the whole length of the borehole. The velocity recorded by a flowmeter log indicated a small number of distinct flow zones with little inflow in between. The velocity log allowed inflow to be allocated to different depth zones, and transmissivity to be proportioned and permeability assigned for each vertical section. The interpreted log in Figure 12.21b shows that assessed permeabilities are high (of the order of К)-4 m/s) in the flow zones and much lower (<10-7 m/s) in the other zones; there is no zone with a permeability of 10-5 m/s, as implied by the pumping test.

< Prev   CONTENTS   Source   Next >