Seepage Meter Design
Over recent decades, various modifications have been made to the basic seepage meter to address potential sources of measurement error and to handle operational issues
The inverted open drum is still the basis of the chamber (Figure 1). A wide range of options have been used by previous investigators including capped PVC casing (Schincariol and McNeil, 2002), plastic buckets (Cey et al, 1988; Alexander and Caissie, 2003), a purpose built rectangular stainless-steel funnel (Paulsen et al, 2001), fibreglass domes (Shinn et al, 2002) or a cut-down galvanised water tank (Rosenberry and Morin, 2004). A chamber with a relatively large radius is recommended, as laboratory tests indicate that variability in seepage measurements decreased with increasing diameter of the chamber (Isiorho and Meyer, 1999). There may be a requirement for the chamber to be robust and stable for use in dynamic flow conditions. In a particular seepage study of Lake Michigan, the chamber used was modified by adding a 50-70kg layer of concrete to the inside of the chamber, with the lower surface of the concrete conically-shaped to direct upwards flow of water and gas to the chamber outlets (Cherkauer and McBride, 1988). However, such chambers may be too heavy for use in soft sediments.
Figure 1: Basic design of a seepage meter with inverted open chamber (1) with flanges to assist in installation and recovery (2). The chamber has a sloping top with a gas venting tube (3) attached at the most elevated side. A 4-L wine bladder acts as a seepage collection bag (4) which is housed in an open protective housing (5). The connecting hose (6) has fittings (7) to enable quick release and a valve (8) near the bag. (Brodie et al, 2005)
Other modifications to the chamber include incorporating lugs onto the top (Figure 1). This allows a rod to be inserted across the chamber top, to facilitate rotation of the chamber during installation. The rod or a lightweight notched steel picket can be hooked onto the chamber (or alternatively ropes attached) to remove it from the sediment. A central fitting can also be incorporated to allow attachment of a rigid vertical pole to help position or remove the chamber. The top of the chamber can be made removable to minimise disturbance of the sediment bed during installation. The top of the chamber can also be painted white to help find and recover the meter, particularly in turbid water.
The tube for the collection bag can be placed to the side of the chamber and another tube at the chamber top is extended above the water surface and open to the atmosphere (Figure 1). This configuration is useful in shallow water to keep the bag submerged, while allowing venting of any gas. Alternatively, a small pipe with a ball valve can be added to the top of the chamber, and used to vent any trapped air when the chamber is initially placed into the water body (Cherkauer and McBride, 1988). After the air is released the valve is closed, so this approach does not allow release of gas accumulated during the actual operation of the meter.
A flexible bag is used rather than a rigid container for the water storage device as the water in the bag needs to be in hydraulic equilibrium with the chamber and surface water body (Figure 2). The principle is that any discharge of groundwater across the surface area of the bed should displace water trapped within the chamber into the bag. Likewise, any recharge of water to the aquifer would be reflected in loss of water from the bag. The selection of an appropriate bag is based on the objective of minimising the energy required to exchange water between the bag and the chamber. Hence, the bag should be robust but flexible, smooth, compliant and thin-walled to reduce head losses. In many studies, hospital dialysis or intravenous bags are used as they are relatively rugged and designed to be attached to tubing. Oven basting bags have also been used (Shinn et al, 2002). Small-volume elastic bags such as balloons or condoms have been trialled previously but are not recommended as the elastic stretch in the bag can create artificial pressure differences (Harvey and Lee, 2000; Schincariol and McNeil 2002). Bladders from wine casks have also been successfully used in Australian trials (Brodie et al, 2005). The chamber can also be connected to a small open container set in the bank of the stream or lake instead of using a submerged plastic bag as a collection device. The inlet to the container is positioned at the same level as that of the water level of the surface water body (Langhoff et al, 2001).
Figure 2: Basic components of the seepage meter including seepage chamber and collection bag (Brodie et al, 2005)
The bag can be housed in a protective cover such as an open length of PVC pipe or a perforated bucket (Figure 1). Field and laboratory studies have shown that surface water movement like waves, currents or streamflow can cause a venturi effect that reduces the hydraulic head in the collection bag and hence the chamber by a centimetre or more. This head loss is significant compared to the natural hydraulic gradient and can induce anomalous upwards groundwater seepage (Libelo et al 1994). Measurements from seepage meters are more reliable in slow-moving water with velocities less than 0.6m/s (ANCID, 2000).
The tubing used to connect the bag to the chamber should be sufficiently rigid to avoid kinking or flexing. Again, the objective is to minimise head losses by using relatively large diameter tubing and avoiding the use of small-diameter fittings that constrict water flow. This is because frictional head loss is inversely proportional to the diameter of the flow conduit. Laboratory tests recommend that tubing diameter should exceed 7.9mm to reduce the hydraulic resistance that can cause measurement error (Fellows and Brezonik, 1980; Rosenberry and Morin, 2004).
A valve can be been incorporated between the chamber and the collection bag, located as close as practical to the bag (Figure 1). The valve can be opened to commence the test and closed to finish the test. A two-way valve can also be used, with one tubing connected to the bag, and the other being a short length of tubing open directly to the surface water body. The valve can be manually operated, but remotely operated versions, using a solenoid-controlled switch have been applied (Cherkauer and McBride, 1988). Initially the valve directs flow to the short open tube to allow equilibration of water pressure between the inside and outside of the chamber. After a period of stabilisation, the valve is switched to allow connection between the chamber and the bag.
The basic design can be modified into a constant-head seepage (Idaho) meter. This incorporates a removable lid to the chamber to minimise disturbance of the sediment bed during installation, an inverted U-tube manometer to monitor when water pressure inside and outside the chamber have equilibrated, a rigid vertical handle to faciliate positioning of the meter, and a siphon reservoir rather than a plastic bag (Worstell and Carpenter, 1969). These meters have been developed to measure negative leakage from irrigation channels into aquifers and trialled under Australian conditions (Byrnes and Webster, 1981). Redesigning into a variable head seepage meter can overcome the issue of maintaining hydraulic equilibrium between the inside and outside of the chamber. A head difference is deliberately induced and the flow into or out of the chamber measured (Bouwer, 1962).
References
Alexander MD, Caissie D, 2003. Variability and comparison of hyporheic water temperatures and seepage fluxes in a small atlantic salmon stream. Ground Water 41: 72-82.
ANCID, 2000. Open channel seepage and control. Vol 1.4 Best practice guidelines for channel seepage identification and measurement. Australian National Committee on Irrigation and Drainage.
Brodie RS, Baskaran S, Ransley T, Spring J, 2005. The seepage meter: Progressing a simple method of directly measuring water flow between surface water and groundwater systems. International Association of Hydrogeologists Conference, Auckland. NZ
Bouwer H, 1962. Variable head technique for seepage meters. Transactions ASCE Vol 127 Part III, 434
Byrnes RP, Webster A, 1981. Direct measurement of seepage from earthen channels. Technical Paper No 64. Australian Water Resources Council, Canberra.
Cey EE, Rudolph DL, Parkin GW, Aravena R, 1998. Quantifying groundwater discharge to a small perennial stream in southern Ontario, Canada. Journal of Hydrology 210:21-37.
Cherkauer DS, McBride JM, 1998. A remotely operated seepage meter for use in large lakes and rivers. Ground Water 26:165-171.
Fellows CR, Brezonik PL, 1980. Seepage flow into Florida lakes. Water Resources Bulletin 16:635-641
Harvey FE, Lee DL, 2000. Discussion on "The effects of bag type and meter size on seepage meter measurements" by SA Isiorho and JH Meyer, May-June 1999 issue, v 37, no 3:411-413. Ground Water 38(3):326-327.
Isiorho S, Meyer JH, 1999. The effects of bag type and meter size on seepage meter measurements. Ground Water 37:411-413.
Langhoff JH, Christensen S, Rasmussen KR, 2001. Scale dependent hydraulic variability of a stream bed on an outwash plain. IAHS-AISH Publication 269, 205-212.
Libelo EL, McIntyre WG, 1994. Effects of surface-water movement on seepage meter measurements of flow through the sediment-water interface. Hydrogeology Journal 2:49-54.
Paulsen RJ, Smith CF, O'Rourke D, Wong TF, 2001. Development and evaluation of an ultrasonic ground water seepage meter. Ground Water 39:904-911.
Rosenberry DO, Morin RH, 2004. Use of an electromagnetic seepage meter to investigate temporal variability in lake seepage. Ground Water 42(1):68-77.
Schincariol RA, McNeil JD, 2002. Errors with small volume elastic seepage meter bags. Ground Water 40(6):649-651.
Shinn EA, Reich CD, Hickey TD, 2002. Seepage meters and Bernoulli's revenge. Estuaries 25:126-132.
Worstell RV, Carpenter CD, 1969. Improved seepage meter operation for locating areas of high loss in canals and ponds. Oregon Reclamation Congree, Klamath Falls, Oregon, USA, November, 1969.