Reviving indigenous technology for promoting the concept of failsafe design in building climate-resilient weirs for irrigation

Climate change is a global phenomenon, and Nepal is not an exception. Furthermore, being located in the Hindu Kush-Himalayan region where the rate of warming has increased at a faster rate over the last five decades against the global average that varies between 0.1 and 0.3 °C (average 0.2 °C) per decade (IPCC 2013, IPCC 2018, WMO 2024), climate change–induced threats to Nepal are higher compared to the rest of the world (Mott MacDonald 2017, Wester et al. 2019, WB CCDR 2022). According to the Global Climate Risk Index 2020, Nepal is ranked among the 10 countries that were most affected by extreme weather events between 1999 and 2018 (Eckstein et al. 2019). The scenario of climate change and the associated threats, which will further increase in the coming years¹, are equally applicable to other mountainous countries in the region.

The climate change–induced threats to the rural communities, especially in maintaining their livelihood, are further intensified because of the natural resources–based subsistence farming system that is shaped by the general integration of the three natural resources, namely, land, water, and forest for crop production. Of these resources, water for agriculture is the most influenced by climate change. The increasing rate of intense and erratic rainfall leading to instantaneous peak floods, declining water availability, and increasing cases of drought and inundation are some of its key consequences influencing the subsistence farming system. As a result, local communities are experiencing multiple challenges in maintaining their livelihood (Climate Risk Country Profile: Nepal 2021, WB CCDR 2022, Baral et al. 2023). Of these consequences, this article is concerned about the threats due to the ever-increasing rate of instantaneous peak floods.

Despite these challenges, there is also a growing appreciation for the existence of indigenous water management technologies and practices and, subsequently, widespread recommendations to adopt their concepts that could provide innovative and applicable adaptation solutions to climate change (MoSTE 2015, IPCC 2018, Nakashima et al. 2018, GON 2019, Wester et al. 2019, WB CCDR 2022). This is specifically true in the case of mountainous environments where patchy climatic zones exist in areas due to their wide altitudinal variations within a short horizontal distance. Such a scenario thus demands a site-specific climate adaptation approach and technology to match the local endowments to natural resources, vulnerability patterns, and adaptive capacity.

This article thus focuses on one such technology of boulders and/or brushwood irrigation water diversion weir designed to fail to cope with unanticipated floods. While doing so, this article concentrates on the small and medium run-of-the-river irrigation systems located in river valleys and down the foothills.

In this context, this article first documents the features and layouts of these weirs and analyzes their configuration from the perspective of basic sciences that enable them to undertake their functions, especially the failsafe function. This article, through short case studies, then documents how engineered weirs fail due to floods to examine the significance of adopting the concept of failsafe design in building them. In this context, this article also presents a couple of conceptual frameworks for the failsafe designs of modern water diversion weirs.

The primary objective of this article is to generate new scientific knowledge on the design of climate-resilient modern irrigation infrastructure by blending indigenous knowledge with modern engineering science.

This article is based on several field-based research studies on various aspects of indigenous irrigation technologies and climate change conducted by the author in the past² followed by the two short specific case studies. These case studies focused on the engineered water diversion weirs of the Dunduwa and Singheghat Irrigation Systems in which quantitative and qualitative data were collected using various tools and techniques such as measurement, unstructured interviews, and observation. In addition, this article also draws on the rapid field visits made to the Julphe and the Rani Jamara Kulariya Irrigation systems. The field visit to the Julphe Irrigation System was made to reconfirm the existence of some of the indigenous irrigation technologies (like the boulder and brushwood diversion weir) that were studied in greater detail during 1997–1998 (Parajuli 1999).

In addition to the above, this article is also based on an in-depth review of the literature. Finally, and more importantly, this article draws heavily on over 40 years of experience of the author in managing and studying irrigation systems in Nepal and other Asian countries in the region.

An irrigation system refers to both the physical infrastructure of works and the social infrastructure of rules and procedures that enable it to undertake its functions through the integration of several activities for transforming irrigation water from one level of an irrigation system to another, and subsequently delivering it to the farmer's field for agriculture production.

In simple terms, technology refers to the capacity to transform goods into desired things (Vincent 1997). An irrigation system controls and transfers water from one level to another (intake, main canal, distribution canals, farm levels, etc.) through various types of infrastructural objects to support plant growth. The said transformation involves a complex process, which is caused by the interaction between hydraulic parameters – water level, velocity, and discharge – of infrastructure, social infrastructure of rules and procedures, and the people and includes multiple activities. Such an infrastructural object that is designed and built for multiple functions with an intended capacity to transfer irrigation waters from one level to another is termed here as irrigation technology. In this sense, technology is a system consisting of a material object, which can transform things into desired outputs through a process that is governed by science. Such objects when created using indigenous knowledge are termed indigenous technology.

The term ‘indigenous’ is sometimes confused with the term ‘traditional’, but they are not necessarily the same. Following Gill (1992), this article uses the following definition of indigenous and traditional:

Indigenous knowledge refers to a cumulative body of knowledge and know-how that have been accumulated across generations by adaptive processes and handed down through generations by cultural transmission (Nakashima et al. 2018). Indigenous knowledge does not have to be knowledge generated by indigenous people.

Climate is the statistics of weather variables over a long period. Climate change refers to the long-term shifts in such statistics, particularly of temperature. Climate change is predicted to affect irrigation systems in two ways: direct and indirect. The likely changes in evapotranspiration and biomass production as a result of increasing temperature are the direct impacts, whereas the increasing rate of unanticipated instantaneous floods and declining water availability are the indirect impacts.

Resilience is an evolving concept in irrigation: it is highly contextual and varies greatly from one location to another. The resilience of irrigation systems can be understood as the ability of the systems to adjust in response to changes.

Most of the run-of-the-river irrigation systems located in the river valleys and foothills of mountains acquire (or used to acquire) water through the respective boulder and/or brushwood water diversion weir built across their source rivers. Such weirs are termed hereunder as BB weirs. They are found in many countries and are known by several names³. By their nature, these weirs are temporary, and thus, they are sometimes known by the name of temporary diversion weirs. They are built following the concept of failsafe design explained below.

The concept of failsafe design (or design-to-fail) does not literally mean that they are designed to fail. What it means is that even if these weirs fail, their parts can easily be reassembled and rebuilt to enable them to function as per their original design with minimal damage to river morphology. This concept – failsafe design – not only reduces the costs of infrastructure but also ensures that the effects of failure and costs to the community are controlled.

In contrast to the situation depicted above, if a modern concrete weir fails, the effects of failure on river morphology and costs to rebuild such a structure will be much higher.

In areas where boulders are not easily available, especially in the wide river valleys and down the foothills of mountains, such weirs are built by layers of brushwood kept inclined over the horizontal wooden bars supported by a series of wooden porcupines⁴ (Figure 2(b)). The brush woods are laid with their leaves facing downward and the sticks upward (Figure 2(d)). The leaves are pressed with large pebbles and gravel so that they do not float. The porcupines increase the stability of the weir, while brushwood leaves first help obstruct the river flow and then divert it toward the intake. Subsequently, they help deposit sediment in their front making the weir more stable.

Unlike the conventional engineering design in which weirs are built perpendicular to the main direction of flow, BB weirs are built skewed to the river axis (Figure 2(a) and 2(c)). Its main reason is to increase its crest length across the direction of flow to accomplish two main functions. The first function is to increase its stability to a certain extent (Pandey 1995; Parajuli 1999; Parajuli et al. 2023). The second function⁵ is to control the entry of flows into the respective off-taking canal within a certain range of the incoming river flow, which if it increases beyond the said range, the weir fails. This function is explained below.

It is to be noted that these weirs supply water to respective canals with open intakes. The flow in such a canal is shaped by the water depth (h) upstream of the weir (Figure 2(a)), which in turn is shaped by the incoming river flow. The higher the incoming flow, the larger the water depth (h) above the weir, and the more the water flows through the open-intake canal and vice versa. However, a long-crested weir helps in stabilizing the water depth (h) above the weir within a certain range of incoming flows. This hydraulic phenomenon of a long-crested weir thus controls the entry of flow into the open-intake main canal up to the said range of incoming flows.

If the incoming flow in the river further increases beyond the said range, such a weir fails automatically due to its inbuilt failsafe design characteristic, and no flow enters the canal. This incident minimizes the risk of flood flow entering the canal and thereby protects the canal and associated structures. This phenomenon is also regarded as automation for safety in terms of preventing flood water from entering the canal (Jacob 1995; SPWP-ILO 2014). The said automation is critically important for a weir built in small and medium irrigation systems as managing floods by manipulating the flow regulating gates in such a weir is not practical due to an accessibility issue, especially during the pouring rains at night.

The state of the art depicted above demonstrated how a locally designed and built infrastructural configuration of BB weir could accomplish its functions by applying the basic knowledge of flow hydraulics. Such an infrastructural configuration of BB weir, which may not be that efficient compared to the modern engineering infrastructure, is certainly a wonderful invention that took into account the basic principles of science. Although many of these weirs have already been replaced or are in the process of being replaced by modern infrastructure, the concept of failsafe design and the other underlying scientific principles adopted in building these weirs are still invaluable.

The section below, through short case studies, presents the design and functions of the engineered weir and examines how they fail due to unanticipated floods. Learning from these failures, the subsequent section then presents likely approaches to building climate-resilient water diversion weirs.

One of the main functions of a water diversion weir in a river is to raise its water level to the required extent for diverting irrigation water to the corresponding canal. This function thus demands that the weir be built above the riverbed up to a certain height. However, the other important function of a weir is to protect the upstream area from the inundation as a result of raised water level at the point of diversion. Thus, other than the requirement to raise the river water level at the point of diversion, the height of the weir that can be built above the riverbed is also shaped by the allowable afflux⁶, which in turn is shaped by the likely flood in the river. The taller the water diversion weir above the river bed, the larger would be the afflux during the flood, and the higher would be its impacts in its upstream area due to the food inundation⁷. In addition to this consequence, the increased flood water depth upstream beyond the designed flood depth⁸ is likely to cause outflanking of the river as a result of its increased kinetic energy and create heavier impacts on the river floor downstream due to the increased discharge intensity. Once a river outflanks, a weir fails to fulfill its function. Likewise, heavier impacts on the river floor downstream threaten the stability of the weir as a whole.

The situation depicted above suggests that the nexus between weir height and river flood is one of the important parameters to be decided in designing a water diversion weir above the riverbed. However, in the prevailing scenario of climate change resulting in unanticipated peak floods with higher levels of uncertainty, it is becoming difficult to define this nexus. The short case studies presented below first highlight how significant is the nexus between the weir height and river flood. The other case study then presents the operational and stability consequences of a weir as a result of unanticipated floods.

The Dunduwa Weir, built across the Dunduwa River during the 1960s to divert the river flow to its main canal, was about 2 m high above the river bed. In addition, the weir was designed with 4-ft⁹ tall wooden falling shutters to further raise the river water level upstream mainly during the period of low flow.

As a result of the above actions, the 2019 flood badly smashed the weir in two ways. First, the settlement and agricultural lands upstream were inundated, and second, the river outflanked the weir (Figure 3(c)).

In 2021, the 1.5 m tall concrete wall built over the weir was dismantled to lower the weir height, and the outflanked river was brought to its normal course by constructing a more robust concrete abutment wall (Figure 3(d)) in its left bank against the beautifully built ancient masonry abutments. To facilitate this construction, the weir was cut in its middle portion to drain the upstream water (Figure 3(d)). Even with this development, the weir could not be used partly due to a fear of afflux upstream. Presently, the weir is abandoned, and its main canal receives water from some other sources.

The Singheghat weir with a gated under-sluice was built in the Banganga River to divert water to its main canal through a gated head regulator to irrigate 2,168 ha of cultivated lands in the Kapilbastu District.

The first case study demonstrated that the height of the weir above the riverbed is one of the critical parameters causing inundation in its upstream and outflanking of the river due to unanticipated floods. The second case study highlighted how an unanticipated flood can erode a weir, especially in its downstream, threatening its stability as a whole. In addition, it also raised an important aspect regarding the operability of gated structures (under-sluice and canal head regulators) built along the weir, especially during unanticipated flooding. It is to be noted that despite the provisions of the gated under-sluice and canal head regulators, they were not operable for several reasons.

There exists a consensus among the researchers and development workers that the irrigation infrastructures need to be designed and built as climate resilient. A water diversion weir is not an exception, which should also be resilient against unanticipated river floods. This is specifically true in the case of small and medium irrigation systems that acquire water from smaller basins. This is because small basins are more sensitive to climate change compared to larger ones (Mott MacDonald 2017). The following two approaches are likely in building climate-resilient water diversion weir.

Mott MacDonald (2017) recommends adopting a risk-based approach in designing climate-resilient water infrastructure that follows the conventional engineering method but with improved design basis and procedures. The improved design basis may call for revisiting the hydrological design parameters such as allowable flood return period, waterway, and effective rainfall. In addition, it may also seek to use the forecasted values of flood flows rather than their recorded values in designing water infrastructures.

A risk-based approach to design¹⁰ includes identification, assessment, understanding of risks in designing water infrastructure, and its likely consequences. This approach provides higher priority to safety.

The risk-based approach to design may, however, lead to a robust weir coupled with several engineering measures to protect it from unanticipated river floods. As a result, such an option may not be financially viable, especially for small and medium irrigation systems. Furthermore, if these weirs are equipped with gated structures for managing peak floods, it may not always be possible to operate them when required for reasons already explained above.

In a mountainous environment, especially for small and medium irrigation systems, it may not always be possible to design and build water diversion weirs without risk of damage. This scenario therefore calls for adopting the concept of failsafe design. As this concept is not commonly practiced in the present-day engineering design of rural infrastructure like water diversion weirs, its design concepts need to be developed by incorporating the elements of indigenous technologies, knowledge, and practices.

Under this concept, a water diversion weir or its parts need to be designed in a disaggregated manner to enable them to function intact during normal circumstances but fail in part automatically under a situation of unanticipated floods for their safe release downstream to avoid inundation upstream, outflanking the river away from the weir, and excessive erosion downstream.

The forthcoming section presents some of the conceptual design framework of water diversion weirs with a concept of failsafe design.

This section presents a couple of failsafe design frameworks for water diversion weirs at a conceptual level. These examples are not novel. They are (or were) already practiced under different circumstances. There can be many other such design frameworks, which need further studies in the coming days.

A weir with falling shutters is not a novel technological option. Several engineering textbooks have recommended this option for diverting water from a source river to the corresponding canal under different conditions (Singh 1972; Sharma 1984). It simply consists of a low-height weir with a series of falling shutters fixed at its crest with some kind of hinging arrangement. When these shutters are raised to their vertical position, they help raise the water level upstream thereby maintaining the designed pond level. These shutters, which are annexed to one another, are supported by inclined rods for counterbalancing the water pressure¹¹.

The textbook recommended heights of these shutters usually remain below 1.5–2 m. With the increase in water level upstream as a result of unanticipated floods, these shutters fall automatically. For this reason, they are termed falling shutters.

The option of falling shutters is not common in present-day engineering design. This is primarily because of its cumbersome tasks of raising the falling shutters repeatedly after each fall due to a flood. Engineers have therefore opted to replace this option with barrages despite their higher costs of construction. As a result, the concept of failsafe design has disappeared in engineering science.

Although a barrage is still a preferable option for a relatively large-scale irrigation system, this option, as noted above, is not that viable for small and medium irrigation systems primarily because of their accessibility issues for operating their gates during periods of floods. In such a situation, the technological option of ‘low-height weirs with falling shutters’ with some innovative arrangements for raising the shutters when required is one of the considerable options for a failsafe design.

The Hydro-Tech (or HTE) Engineering claims that an inflated rubber dam allows considerable savings on the cost of materials, construction, and installation is much quicker and simpler compared to conventional gate structures, and low comparative operation and maintenance costs. Chanson (1998) however notes that its overflow characteristics need due consideration.

The height of a rubber dam can range from 0.30 to 5 m¹³, which is usually built in a section that may vary between 30 and 100 m or even more (Figure 5(b)). When anchored at the elevated weir, the impact due to sediment is low. Furthermore, when a weir is built with an under-sluice, it can easily manage the incoming sediment.

During the irrigation season, the gaps between the two consecutive concrete walls are blocked by temporary check structures like sandbags, boulders, and earth-bank to raise the river water level upstream and to divert it to the concerned canal. During the flood, such temporary check structures get washed away, which can be easily replaced with a little effort. Box 1 presents an example of the replication of this technology in the Amu River in Afghanistan.

The crest level of the earthen embankment should match the designed flood level. When the incoming flood rises above the designed flood level, water spills over the earthen embankment. Because it is made up of earth, the spilling water erodes it quickly and subsequently the embankment collapses in no time resembling a fuse plug. The bypass channel, which connects the earthen embankment (fuse plug) with the river downstream, discharges the river flood back to the river making the weir safe. Once the flood is over, the earthen embankment can be rebuilt quickly.

Climate is changing and this will get worse. One of the most obvious physical impacts of climate change is likely to be more intense and erratic rainfall, leading to unanticipated instantaneous floods causing physical damage to irrigation systems particularly to the water diversion weirs built across the river¹⁴. Some of the likely damages to the engineered water diversion weirs are (a) inundation to the settlement and agricultural land upstream of the weir, (b) outflanking the weir by the concerned river, and (c) erosion downstream of the weir. The height of the weir above the river bed is one of the critical physical parameters exaggerating these damages.

There is a consensus that the modern water diversion weirs across rivers need to be designed as climate resilient, especially from the perspective of unanticipated river floods. This means that such a weir should have the ability to cope with unanticipated floods. This ability is required specifically for weirs built in small and medium irrigation systems, as such systems acquire water from smaller basins that are more sensitive to climate change.

Resiliency can be achieved by taking action on adaptation. In irrigation, a water diversion weir can be built to be climate resilient by adopting an appropriate approach in its design.

One such approach is to adopt a risk-based design with improved design basis and procedures for its engineering design (Mott MacDonald 2017). However, the design process for such an approach is not yet fully developed, especially in developing countries. Furthermore, this approach to design is likely to lead to a more robust infrastructure coupled with several engineering measures including the gated water control structures (under-sluice, escape, overflow, etc.) for regulating and managing the unanticipated flood. However, despite adopting a risk-based approach to design, such a weir may still not be climate resilient unless the community poses its operating capacity, which is one of the important indicators of resiliency. The case study described earlier has also documented the lack of such capacity in operating the regulating structures of the modern water diversion weir, especially during periods of peak floods. This aspect thus suggests that large-scale and costly adaptation projects may not always be climate resilient unless the local community and the society pose the required adaptive capacity.

The other approach to designing a climate-resilient modern water diversion weir is to adopt the concept of failsafe design described earlier in this article. This approach envisions that the effect of failure of an infrastructural object is controlled, though it is not designed to fail. In addition, this approach enables the concept of automation for safety, which is the most preferred concept at the local level because it is virtually impossible to control or regulate flood flows in small and medium irrigation systems manually, especially during pouring rains at night. Furthermore, it is also not likely that small and medium irrigation systems in a mountainous environment can be designed and built without a risk of damage.

The situation depicted above thus suggests that the design approach that minimizes (or controls) the risk of damage to an infrastructural object is a viable design option that may be adopted in building a climate-resilient irrigation infrastructure rather than adopting a risk-based approach to a design that leads to a more robust and expensive risk-free infrastructural object, which however may not be familiar to (or adaptable by) the community in the given context.

Physical damage to the irrigation infrastructure, especially water diversion weirs built across the river, is one of the obvious consequences of climate change–induced unanticipated floods.

Although climate change and its consequences have become one of the global agendas in the recent past, there is also a great deal of appreciation for the existence of the indigenous technologies developed by local communities that could provide innovative and applicable solutions to climate change. Recognizing the significance of such technologies, this article examined the configurations and functions of one such indigenous technology of boulder and brushwood water diversion weirs.

Unlike the conventional engineering design in which weirs are built perpendicular to the main direction of flow, these weirs are built skewed to the river axis. One of the main hydraulic reasons behind this skewing is to regulate the entry of flow through the ungated off-taking canal by designing the weir with a failsafe function. Such a design is known as a failsafe design. The concept of a failsafe design does not literally mean that they are designed to fail. What it meant is that even if these weirs fail, the effects of failure are controlled. In addition, it enables the concept of automation for safety. Though the configuration and functional design of this indigenous technology (the boulder and brushwood water diversion weir) may not be highly efficient compared to the modern-day engineering design, its concept of ‘failsafe design’ that follows the basic scientific principles is still a wonderful invention that can be invaluable in building climate-resilient modern water diversion weirs.

With this recognition, this article argues that the failsafe design is an important concept that needs to be adopted and institutionalized in designing climate-resilient modern water diversion weirs across the river. In this context, this article also presented a couple of conceptual design frameworks for such weirs with a concept of failsafe design, especially for the medium and small irrigation systems located in wide river valleys or at the foothills of mountains. The conceptual design frameworks presented here are not novel. They are (or were) already practiced under different circumstances, and several such likely options need further studies. As the design frameworks presented here are at the concept level, this article further recommends establishing their detailed design parameters through action research.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.