5. Groundwater Irrigation Potential: Groundwater Pumps

My last post outlined that irrigation significantly increases agricultural productivity and reliability. Globally, 40% of irrigation water is derived from groundwater (Siebert et al., 2010) and increased groundwater extraction has substantially contributed to expanding the global food supply (Giordano & Villholth, 2009; UNESCO, 2012). Yet despite the presence of vast groundwater reserves, only 5% of arable land in SSA is irrigated (Siebert et al., 2010), primarily due to a lack of extraction infrastructure. This post investigates how we could improve groundwater access via cheaper, smaller-scale and more easily maintainable groundwater pumps. 

What needs to be replaced?

… diesel pumps. The smallest diesel pumps available are ~5 horsepower (~3750 watts). They are expensive (around US$500) (Polak, 2000) and ~10 times more powerful than small-scale farmers (with farms of 1-5ha) need (Shah et al., 2000). In fact, overuse of large pumps on small areas of land can lead to over-irrigation. Furthermore, diesel is expensive and often hard to access and the pumps tend to break down regularly (Odeh et al., 2006).

Fig. 1 Installing a diesel groundwater pump (Source: International Water Management Institute; Photographer: Joe Ronzio) 

Suction Treadle Pumps (STPs):

STPs were first developed in Bangladesh in the 1980s (Polak, 2000). They are similar to traditional hand pumps, except the driving force is changed from the arms to the legs. This enables more water to be pumped, at a greater rate. There are two ‘treadles’ connected to two pistons, which are contained within two cylinders. The operator stands on the treadles and presses them up and down in a sort of walking/pedalling motion. The movement of the pistons creates suction in the cylinders, which draws water up and out of a spout into a canal for gravity irrigation. The figures below show a) the basic STP design and b) a bamboo STP in use in India. STPs are cheap, small-scale, do not require fuel, and are less likely to breakdown than diesel pumps. Therefore, they can be manufactured, operated and maintained at low cost in developing countries (Kay & Brabben, 2000)

Fig. 2 Diagram of the basic treadle pump design (Source: Practical Action Publishing)

Fig. 3 Bamboo treadle pump in use in India. (Source: International Development Enterprises India (IDEI))

The average STP in Asia costs ~US$20, whereas in Africa they tend to cost ~US$60 due to more expensive materials (Polak, 2000, p.27). Either way this is approximately a tenth the cost of a diesel engine. Under the conditions outlined in the table below (deemed as average for rural farmers in most of SSA) each STP irrigates an area of ~0.4ha, whereas using watering cans under similar conditions could only irrigate 0.03 ha (Kay & Brabben, 2000).

Fig. 4 Variables used to calculate average area of land that can be irrigated using one STP (Source: Kay & Brabben, 2000)

Kay & Brabben (2000) describe the design and dissemination of STPs as leading an irrigation ‘revolution’ across Asia, with over 2 million sold to rural small-scale farmers in Bangladesh and India (Kay & Brabben, 2000). Owners of STPs in these countries now earn around US$100 more each year than they did before (Polak, 2000). However, unfortunately this revolution has not yet reached Africa.

Why are STPs not widely used in Africa?
(1) Groundwater in Africa is often deeper than in much of Bangladesh and India. Suction uplift is limited to 1 atmosphere of pressure, which is equivalent to a 10m water head. Therefore, the maximum extraction depth for a 100% STP would be 10m. However, in practice STPs are not 100% efficient so their maximum extraction depth is only ~6m and they actually work best at around 2-3m (Polak, 2000; Kay & Brabben, 2000). Yet, much of African groundwater is >10m deep, as shown in the diagram below (MacDonald et al., 2012) and therefore cannot be extracted via suction.
(2) The land in Africa is often more hilly than in much of Bangladesh and India and gravity-driven canal irrigation does not work well on undulating land, unless it is terraced. As a result, the original STP design is only really appropriate for African floodplains, where groundwater is shallower and land is flatter (Kay & Brabben, 2000). 

Fig. 5 The depth of groundwater storage in Africa (Source: MacDonald et al., 2012)

Pressure Treadle Pumps (PTPs):

PTPs were designed to try to accommodate the needs of African farmers (Kay & Brabben, 2000). They are marginally more expensive than STPs (in the range of US$60-120, as opposed to US$ 50-70) but not excessively so (Kay & Brabben, 2000). In a PTP the piston ‘pushes’, rather than sucks, the water up, as demonstrated in my diagram below. (Adeoti et al., 2007). This allows water to be extracted from greater depths. Water is then be fed under pressure into sprinklers or hoses, which can be used to irrigate undulating land. 

Fig. 6 Simplified, hand drawn diagram demonstrating the difference between a) a suction pump and b) a pressure pump.

However, the depth of PTP extraction is still limited! It is constrained by how much power a person can generate by pedalling. For any given flow rate, the deeper the water… the more power is needed to extract it. The energy needed to lift 1L of water by 1m is ~ 10 joules (J). Therefore, assuming 100% pump efficiency (which is impossible) lifting 1L by 1m per second requires 10J/s i.e. 10 watts. So if you are trying to extract water that is 10m deep, assuming the pump has ~50% efficiency (which is not unreasonable) the person operating the pump would need to be generating 200 watts, which is equivalent to what a fit cyclist would generate on a bike. Any greater rate or depth of water abstraction is beyond human capability. (NB these are all my own calculations and are only intended to be rough estimates). 

Solar pumps:

Replacing manpower with solar power would allow water to be extracted from deeper sources, whilst freeing up labour (Blagbrough, 2001). Solar panels rarely breakdown, can be made to any scale, do not have any operating costs and don't need to be connected to the grid or a fuel supply (Kolhe et al., 2002; Odeh et al., 2006). Unfortunately, they do still have high capital costs, only slightly less than diesel pumps, but these are falling rapidly. 

Fig. 7 Solar powered pressure pump (Source: Greenmax Technology)

Summary:
Treadle pumps have been successful in India and Bangladesh but are inappropriate for much of Africa’s arable land because the groundwater is too deep (often >10m) and the land too hilly. Although PTPs can extract water from greater depths than STPs, this is still limited to around 10m due to manpower constraints. Replacing manpower with solar power could hugely increase extraction depths whilst freeing up labour.