By: Theresa Boateng
Ghana’s response to flooding has traditionally placed considerable emphasis on moving water away from vulnerable areas through drains, channels and other conveyance infrastructure. But a critical engineering question is: what happens when the volume of water entering the drainage system exceeds its safe capacity? One approach is to increase the size and capacity of drainage infrastructure. Another is to prevent the entire flood volume from reaching the downstream drainage network at the same time. This is the principle behind strategic stormwater storage: temporarily capture excess runoff upstream, regulate its release, and allow the downstream drainage system to handle a smaller and slower flood wave.
A study of the Dzorwulu Basin in Accra modelled the use of an existing reservoir for urban flood management. For the 3–4 June 1995 rainfall event, the simulated storm generated approximately 13.09 million m³ of floodwater. The reservoir was estimated to store approximately 34.5% of the flood volume, producing about a 45% reduction in peak discharge and a 38.5% reduction in downstream inundation depth. The peak also occurred approximately 9.3 hours later downstream, and the inundated area was reduced by about 16% [1]. The study concluded that reservoir storage could be particularly suitable where highly urbanised communities downstream have limited space for expanding drainage infrastructure, but suitable storage opportunities remain upstream.
The significance of this approach is that it shifts the engineering objective from simply “move the water faster” to “control when and how much water reaches the vulnerable area.” Flooding can occur when water arriving at a location exceeds the capacity of the surrounding drainage and landscape to accommodate it. If a catchment generates a large volume of runoff over a short period, even a well-designed downstream drain can become overwhelmed. In some cases, increasing water channel capacity can address part of the problem, but in densely developed urban areas there may be physical limits to how much a drain can be widened or deepened. Strategic water storage provides an alternative: modifying the flood hydrograph before it reaches the constrained section of drainage networks. The underlying principle is that rainfall generates runoff across the catchment, which is collected by smaller drainage channels and eventually concentrated into larger waterways. Without sufficient storage, this produces a rapid increase in downstream discharge and a high flood peak. A strategically located detention reservoir, retention basin or flood-storage area intercepts part of this flow. Instead of allowing the entire inflow to pass immediately downstream, the structure temporarily stores water and releases it through a controlled outlet, thereby lowering the downstream hydrograph.
In hydrological terms, the intervention does not necessarily eliminate the flood volume; rather, it attenuates the flood peak by redistributing the volume over time. If the incoming flood has a peak discharge Q_in, the objective is to produce an outflow Q_out that is lower than the inflow peak while remaining within the safe capacity of downstream infrastructure. The relationship can be conceptualised through the reservoir water balance:
Storage change = inflow − outflow.
During the rising portion of a storm, inflow exceeds controlled outflow and the reservoir fills. As rainfall decreases and inflow falls, stored water can subsequently be released. The effectiveness of the system therefore depends on storage volume, the geometry of the reservoir, the characteristics of its outlet, the timing and magnitude of rainfall, and the capacity of the downstream drainage network.
The significance of the Dzorwulu study is not to establish its practicability across a general scope as it was conducted for a particular catchment, reservoir configuration and historical storm. The value of the research lies in demonstrating the physical mechanism and showing that, under appropriate conditions, upstream storage can materially alter downstream flood behaviour.
Stormwater storage is particularly relevant to rapidly urbanising cities in Ghana. Studies have shown that urban development increases impervious surfaces, reducing the opportunity for rainfall to infiltrate and increasing the speed and volume of runoff reaching drainage systems. In Kumasi, for example, Abass et al. (2020) found that impermeable areas expanded by approximately 54% between 1986 and 2016, while permeable space declined correspondingly, contributing to increased flood incidence alongside inadequate drainage and other urban pressures [2]. Another recent assessment of Kumasi similarly identifies land-cover change, drainage limitations and other urban dynamics as important contributors to flood susceptibility [3]. The implication is that drainage capacity cannot be considered independently from the amount and timing of runoff entering the system. If urbanisation continually increases runoff while downstream channels remain spatially constrained, there is a limit to what conventional channel expansion can achieve.
A more efficient engineering approach would therefore combine conveyance with temporary storage. This could involve a network rather than a single large reservoir. Smaller detention basins could capture runoff from particular sub-catchments before it enters major drainage corridors. Larger upstream reservoirs could provide greater attenuation where suitable topography and available land exist. Existing reservoirs or water bodies could potentially be assessed for dual flood-management functions where their primary purposes and operating requirements permit it. In highly constrained urban areas, storage could potentially be incorporated into parks, road infrastructure or other engineered spaces, provided that the design deliberately accommodates temporary inundation without creating additional risk. The key is that storage should be designed according to the hydrology of the catchment rather than simply the availability of land. Engineers would need to determine several critical parameters: the contributing catchment area, design rainfall, expected runoff volume, peak inflow, required storage volume, outlet capacity and downstream drainage capacity. Digital elevation models, rainfall records, land-use data and hydrological-hydraulic models can be used to simulate these relationships. The Dzorwulu research demonstrates precisely this type of integrated Geographical Information System (GIS), rainfall-runoff and hydraulic modelling approach [1].
The design objective should also extend beyond reducing peak discharge. A successful storage system should reduce downstream flood depth and extent while ensuring that stored water can be safely released before the next major rainfall event. This makes drawdown time important. A reservoir that remains full after one storm may have little capacity to attenuate another storm shortly afterwards. Storage therefore has to be managed as a dynamic component of the drainage system rather than simply as a static basin.
Additionally, combining stormwater storage with improvements to the existing drainage network helps mitigate the challenges posed by inadequate drainage capacity and significant siltation, which reduce the effective capacity of drainage systems to handle increased stormwater runoff [4]. The Ghana Hydrological Authority’s Drainage Unit likewise identifies the planning, design, supervision and maintenance of stormwater drainage systems as central to national flood management [5]. This emphasises the need for a layered engineering system. Stormwater storage reduces the flood peak; drainage conveys the remaining flow; hydraulic structures regulate water movement; and monitoring and forecasting help operators respond to changing conditions. Increasing drain capacity alone may move more water downstream, but if the downstream system remains constrained, this merely transfers the problem rather than solving the flood risk. Conversely, storage without adequate drainage or controlled outlets can create another point of failure.
An important extension of this concept for Ghana is the water–energy nexus. Some reservoirs already perform functions beyond flood management, including water supply, irrigation or hydropower. This creates an opportunity but also a technical trade-off. Reservoir levels cannot be managed solely for flood protection if the same storage is required for electricity generation or other water uses. Research examining Ghana’s White Volta Basin has shown that hydropower-related dam operations can interact with downstream flooding; modelling cited in the study found that Bagre Dam water spills can increase flood levels downstream, with the strongest effects occurring around 100–150 km downstream at Pwalugu [6]. The lesson is not that dams are inherently harmful for flood management. Rather, reservoir operation matters. A multipurpose reservoir can potentially contribute to flood mitigation if sufficient storage is deliberately maintained before periods of high inflow and releases are managed to avoid compounding downstream flood peaks. Achieving this requires reliable rainfall and inflow forecasting, reservoir rule curves, downstream flood information and coordination between water and energy objectives. It also requires recognising that a reservoir designed primarily for hydropower may have different operating priorities from a reservoir designed specifically for flood attenuation.
Stormwater storage however has some documented limitations; Suitable upstream land may not exist in every urban catchment. Reservoirs and detention structures require substantial initial investment, engineering assessment and long-term maintenance. Sedimentation progressively reduces storage capacity. Poorly designed structures can themselves become flood hazards if spillways or outlets are undersized. Storage also cannot eliminate extreme flood risk: if a storm exceeds the design event such as a 1-in-100-year rainfall versus the 1995 event modelled at Dzorwulu—the excess water must still pass downstream. Climate uncertainty further complicates design because infrastructure based solely on historical rainfall may be inadequate if future rainfall intensity changes. Therefore, strategic stormwater storage should not be considered as Ghana’s replacement for natural flood management, drainage improvement or other flood-risk measures. Rather, it provides a different function. Whereas restoring wetlands and floodplains seeks to recover natural storage and attenuation, engineered detention and reservoir systems provide deliberately designed and controllable storage. The two approaches can potentially work together: natural systems provide distributed, resilient storage across the catchment, while engineered reservoirs and detention structures provide larger, controlled storage where hydrological modelling identifies significant downstream benefit.
Evidence from documented stormwater storage studies points to an important engineering principle: do not force the entire flood wave through the most vulnerable part of the drainage system at once [7, 8, 9]. Where downstream urban areas are already highly developed and drainage expansion is physically constrained, strategically located upstream storage can intercept part of the flood volume, reduce peak discharge and delay the arrival of water. The practical solution is not simply to build bigger drains—it is to design a catchment-scale system that combines conveyance with controlled storage. As Ghana’s cities continue to urbanise, deliberately making space for floodwater upstream offers an important engineering complement to mitigate flooding downstream.
References
[1] Owusu, P.A., Odai, S.N., Annor, F.O. & Adjei, K.A. (2013). Reservoir storage for managing floods in urban areas: a case study of Dzorwulu basin in Accra. Hydrological Processes, 27, 1615–1625. https://doi.org/10.1002/hyp.9286
[2] Abass, K., Buor, D., Afriyie, K., Dumedah, G., Segbefi, A.Y., Guber, I., Yankey, R. & Oduro-Ofori, E. (2020). Urban sprawl and green space depletion: Implications for flood incidence in Kumasi, Ghana. International Journal of Disaster Risk Reduction, 51, 101915. https://doi.org/10.1016/j.ijdrr.2020.101915
[3] Osei, M.A., Asare, M.Y., Asante, W.A. & Adjei, P.O.-W. (2025). Urban growth or urban risk? Unraveling the flood paradox in Kumasi (Ghana) through the lens of natural factors and spatial decision modeling. Journal of African Earth Sciences, 232, 105814. https://doi.org/10.1016/j.jafrearsci.2025.105814
[4] Danquah, I.O. (2022). Flood risk management and runoffs into drainage systems assessment in Kofrom – Kumasi, Ashanti Region of Ghana. International Journal of Research, 8(1), 1–16. https://zenodo.org/records/6155610
[5] Ghana Hydrological Authority. (n.d.). Drainage Unit. Retrieved August 2026, from https://hydro.gov.gh/drainage-unit/
[6] Li, C., Yu, W., Dzodzomenyo, M., Asamoah, M., Kerapetse, C.T., Kandel, M. & Wright, J. (2021). Growing spatial overlap between dam-related flooding, cropland and domestic water points: A water–energy–food nexus management challenge in Malawi and Ghana. Frontiers in Water, 3, 730370. https://doi.org/10.3389/frwa.2021.730370
[7] Acheampong, J.N., Gyamfi, C. & Arthur, E. (2023). Impacts of retention basins on downstream flood peak attenuation in the Odaw river basin, Ghana. Journal of Hydrology: Regional Studies, 46, 101336. https://doi.org/10.1016/j.ejrh.2023.101336
[8] Mguni, P., Herslund, L. & Jensen, M.B. (2016). Sustainable urban drainage systems: examining the potential for green infrastructure-based stormwater management for Sub-Saharan cities. Natural Hazards, 82(S2), 241–257. https://doi.org/10.1007/s11069-016-2309-x
[9] Amoako, C. & Boamah, E.F. (2015). The three-dimensional causes of flooding in Accra, Ghana. International Journal of Urban Sustainable Development, 7(1), 109–129. https://doi.org/10.1080/19463138.2014.984720
