Infrastructural Underperformance and Spillway Geotechnical Failure: A Forensic Investigation of the Gerebsegen Multi-Outlet Reservoir, Ethiopia.

1. Introduction

The strategic deployment of large-scale water storage assets across East Africa particularly within the complex geomorphological terrain of the Ethiopian Highlands and the East African Rift System is a fundamental pillar supporting regional climate resilience, smallholder food security, and long-term municipal water sustainability [1,2,3]. However, this rapid expansion of structural assets has increasingly collided with deep-seated geomechanical, hydraulic, and institutional distress [4]. While modern hydraulic engineering frameworks focus heavily on preventing macro-hydrologic overtopping events, forensic data indicates that a substantial proportion of dam safety incidents stem from localized geomechanical adjustments and hydrogeological oversights localized within ancillary spillway structures, adjacent natural flanks, and abutment retaining walls [5,6,7]. This physical vulnerability is further exacerbated by systemic administrative inefficiencies, project management discontinuities, and governance bottlenecks that prevent water assets from achieving their intended design utility [8].

This intersection of physical and operational vulnerability is heavily documented across sub-Saharan Africa’s variable micro-climates. For example, regional experiences highlight that minor geomechanical discrepancies often propagate into major structural and socioeconomic failures [9,10]. In Kenya’s Rift Valley region, unmanaged hydraulic forces have frequently triggered downstream scour damage and localized embankment slope instabilities along seasonal river passages [11,12]. Similarly, historical failures of masonry structures in weathered Tanzanian basements demonstrate that rigid gravity segments are poorly suited to absorb lateral earth shifts without internal pressure relief networks [13]. These macro-regional failure modes show that complex interactions at the boundary where artificial masonry or concrete anchors into variable, fractured natural rock masses require a shift from deterministic design standards toward modern limit state analysis [14].

The Gerebsegen Dam, situated on the Gabbat River at a baseline elevation of 1980 meters above sea level (UTM, WGS 84: 542000 E 1481500 N) within the Tigray region of Northern Ethiopia, exemplifies these dual technical and administrative challenges [8]. Built with a gross reservoir storage capacity of 25 million cubic meters (Mm³), the multi-purpose asset was engineered with a projected 30-year operational design life to serve a critical regional demographic [5]. The project fulfills a dual mandate: bolstering rural agricultural yields through downstream smallholder irrigation networks across a designated 500-hectare command area, and securing the municipal water supply of Mekelle city [5]. Operationally, this layout relies on a localized distribution configuration utilizing three distinct structural outlet conduits: one dedicated specifically to irrigation routing and two separate pipelines designed for municipal water treatment supply lines.

Despite its vital socioeconomic role, a profound discrepancy exists between the project’s chronological timeline and its operational efficacy. Although 12 years have elapsed since the completion of its construction representing nearly half of its intended design life the reservoir currently utilizes a mere 20% of its total water volume [5]. This minimal yield is directed exclusively toward municipal water supply through just one of the two installed water supply lines, meaning the municipal water delivery system functions at a tight 50% operational capacity. Meanwhile, the second municipal line remains completely inactive, and the infrastructure required to initiate the planned 500-hectare irrigation scheme has failed to materialize, resulting in a staggering 0% service delivery for agricultural development. Given that the project has nearly reached the midpoint of its intended design period while delivering only a fraction of its planned baseline utility, this systemic underperformance stands as a stark indicator of institutional mismanagement, professional irresponsibility, leadership incompetency, or the restrictive influence of patronage networks.

Compounding these institutional failures, immediate structural distress occurred post-impoundment. Severe, longitudinal open cracks quickly propagated along the left abutment flank of the spillway, culminating in progressive overturning, lateral displacement, and the ultimate structural failure of the stone masonry gravity retaining structures. These critical failures are attributed to intense lateral earth pressures compounded by a total lack of internal weep-hole drainage networks and massive, unmanaged subsurface seepage bypassing the core. By connecting the localized forensic data collected from the Gerebsegen site with systemic design and governance limitations identified across other East African infrastructures, this study advances the technical and administrative standards required to enforce long-term structural reliability and safety in complex sub-Saharan geomechanical environments.

East African Regional Context and Dam Safety Paradigms

Dam safety governance and technical execution across East Africa are undergoing a critical paradigm shift as regional water storage assets age, climate variability intensifies, and institutional delivery frameworks face intense scrutiny [15,16]. Historically, unreinforced stone masonry and gravity concrete configurations have been widely utilized for low-to-medium head water retention and spillway training structures across Ethiopia, Kenya, Uganda, and Tanzania due to the immediate socioeconomic benefits of using local labor and readily available regional materials [17]. However, modern forensic monitoring reveals that these rigid structures lack the structural flexibility and built-in internal drainage systems required to withstand the high hydrostatic forces and rapid pore-water pressure spikes characteristic of the region’s highly weathered, fractured basaltic and dolerite rock formations [18,19].

This structural vulnerability is heavily highlighted by engineering failures across the East African Rift System (EARS), where complex regional tectonic histories create localized geomechanical anomalies [20,21]. For instance, forensic reviews of auxiliary structures in the weathered metamorphic baselines of Kenya’s centralized rift basins show a repeated pattern of localized failure [22]. This occurs because rigid masonry components cannot easily adapt to the differential settlement or minor lateral shifts of natural slopes under saturated conditions [23].

Similarly, deep seepage anomalies and piping failures observed in unlined spillway channels within Ugandan basement complexes demonstrate that standard, deterministic geological assumptions often fail to identify hidden sub-surface micro-fractures and active hydraulic pathways [24]. Furthermore, during excavation stages for major water infrastructure projects in Tanzania such as the auxiliary works along the Rufiji River basin the sudden unloading of highly jointed rock masses frequently triggered retrogressive block movements, demonstrating how vulnerable steep construction cuts are when rock mass quality is poor [25].

These regional experiences show that the structural and operational distress at the multi-purpose Gerebsegen Dam is not an isolated event. Instead, it reflects a widespread regional issue where structural damage is concentrated directly at the interface where rigid, artificial masonry or concrete elements anchor into highly variable, weathered natural abutments, combined with a lack of institutional accountability during the operational phase of the asset.

When these structural interfaces are exposed to excessive subsurface flows such as the unallowable 0.568 m³/s seepage rate observed at Gerebsegen they quickly experience rapid pore-pressure build-up, accelerated shear strength degradation, and eventual structural collapse, while the administrative leadership fails to execute downstream distribution assets like the irrigation networks.

Addressing these systemic vulnerabilities requires updating regional design and governance standards. Workflows must shift away from simplified empirical rules and fragmented project execution. Instead, regional development bodies must integrate limit state structural checks, advanced limit equilibrium numerical simulations, mandatory subsurface seepage barriers like pressure-grouting curtains, and rigid, transparent lifecycle management protocols into standard engineering practice across Sub-Saharan Africa.

2. Materials and Methods

The forensic investigation and subsequent structural stabilization design for the Gerebsegen Dam spillway were executed via four deeply integrated technical phases:

• Topographic Surveys and Failure Mapping: High-precision total station cross-sectional profiling was completed along a 250-meter critical zone covering the spillway approach, natural flank slopes, and displaced structural segments to define exact lateral and rotational failure kinematics.
• Geotechnical Core Parameterization: Geological structural logging of core drillings was cross-referenced with surface engineering-geological mapping. Joint frequencies, aperture sizes, and weathering profiles were assessed to index the rock mass using the Rock Mass Rating (RMR) methodology, defining the structural properties at the weathered dolerite-shale boundary.
• Hydrogeological Seepage Diagnostics: Reservoir water balance computations were coupled with field floating-velocity array measurements along downstream discharge zones to map and quantify subsurface bypass vectors.
• Code-Compliant Engineering Redesign: Structural capacity evaluations were conducted to replace the failed gravity masonry walls with optimized cantilever reinforced concrete (RCC) configurations. Calculations were strictly based on Eurocode 2 (EN 1992-1-1), Eurocode 7 (Geotechnical Design), and the Ethiopian Building Code of Standards (EBCS-7: Foundations) to verify stability against sliding, overturning, and structural base bearing failure.

Figure 1. Workflow methods.

Figure 1. Workflow methods.

3. Results and Discussion

3.1. Geographic Framework and Field Survey Mapping

To capture the spatial and geometric nuances of the failure plane, the forensic team established a baseline coordinate matrix across the reservoir boundary, leveraging high-resolution satellite background datasets to understand macro-structural trends. The comprehensive spatial framework, mapping operations, and geomorphological terrain attributes generated during this multi-disciplinary investigation are consolidated in Figure 2. Within this framework, Figure 2a establishes the macro-geographic setting of the project, locating the multi-purpose Gerebsegen Dam on the Gabbat River within the complex geomorphological and tectonically active terrain of the Tigray Highlands in northern Ethiopia, detailing the primary reservoir water body layout, dam embankment fill, local access routes, and explicitly anchoring the project within the WGS 84 UTM zone grid coordinates. Building upon this regional context, Figure 2b delivers a micro-spatial diagnostic breakdown of the localized failure zone bounding the left shoulder of the spillway structure, marking the exact trace of the primary structural crack line (represented by the prominent green trace line) running parallel to the spillway masonry wingwall and highlighting the critical slope stabilization boundary (demarcated as a yellow hatched polygon) where a 1,700.50 m³ upper-slope mass offloading excavation is designated to reduce the active driving force down to a stable 45° profile. Complementing these spatial maps, Figure 2c documents the core forensic photogrammetry and field operations, capturing the field survey team utilizing high-precision total station instruments directly on-site to establish exact lateral, cross-sectional, and rotational kinematic profiles along the highly distressed, overturning masonry gravity retaining structures flanking the spillway approach. Finally, Figure 2d outlines the detailed catchment and reservoir perimeter topography, using tight 0.5-meter minor contour lines and 1.0-meter primary contour spacing to delineate the natural gradients, slope faces, and precise elevations across the critical natural flanks and abutment boundaries, while explicitly outlining the structural survey points and core test boring or test pit coordinate locations used to index the sub-surface rock mass quality.

3.2. Geomechanical Modelling, Slope Stability Parameters, and Offloading Envelopes

To establish a rigorous analytical foundation for the left abutment stabilization, forensic field core drillings and surface engineering-geological mappings were synthesized to index the site-specific lithological properties. As consolidated in the engineering-geological parameter index of Figure 3, the stratigraphic profile is characterized by a weak, highly fractured upper stratum composed of weathered dolerite rock mass, underlain by a more competent, less deformable shale-tiny limestone interbed foundation layer. This upper dolerite slope-forming mass exhibits a poor geomechanical rock quality, indexed at a weak Rock Mass Rating of RMR = 40, and is parameterized by a rock bulk unit weight (γ) of 21.5 kN/m³, an effective cohesion (C) of 150 kPa, and a remarkably low internal rock friction angle (φ) of 17.5°. Due to this “Poor Rock” classification, the mechanics governing slope instability are predominantly driven by continuous mass yielding and block-raveling shifts along unmitigated cuts, rather than purely structurally controlled single-discontinuity failures.

To mathematically model and quantify the structural stability of this weak dolerite slope unit, a deterministic Limit Equilibrium Planar Failure Wedge Model was deployed to evaluate the Factor of Safety (FS) across varying geometries. Under dry, idealized ambient conditions where surface runoff is strictly controlled and tension cracks remain un-hydrostatically pressurized the standard multi-variable equilibrium formulation is derived as follows:

$$\mathbf{F}\mathbf{S}={\displaystyle \frac{2\mathbf{C}}{\mathsf{\gamma }\mathbf{H}\mathbf{s}\mathbf{i}\mathbf{n}²\mathsf{\psi }(\mathbf{c}\mathbf{o}\mathbf{t}\mathsf{\psi }-\mathbf{c}\mathbf{o}\mathbf{t}\mathsf{\beta })}}+{\displaystyle \frac{\mathbf{t}\mathbf{a}\mathbf{n}\mathsf{\phi }}{\mathbf{t}\mathbf{a}\mathbf{n}\mathsf{\psi }}}$$

(1)

Within this governing kinematics framework, stability is fundamentally constrained by the requirement that the theoretical critical failure plane dipping angle bounding the unstable wedge (ψ) must satisfy the geometric boundary criteria (φ < ψ < β), where β represents the programmed face slope excavation angle. The parametric relationship between the computed Factor of Safety (FS), the slope cut angle (β), and the vertical slope height (H) is systematically mapped in the analytical curves of Figure 3. These curves demonstrate that as the vertical cut height (H) escalates from a baseline of 15 meters up to a critical height of 40 meters, the unreinforced rock mass approaches an ultimate limit state of collapse ( $FS\approx 1.08$ ), rendering the original steep cuts highly unstable and necessitating an immediate slope reduction campaign.

This geomechanical vulnerability is evaluated via advanced limit equilibrium numerical simulations performed in SLOPE/W, as documented in Figure 4. While the deeper, global non-circular slip surface simulation yields a nominal static Factor of Safety of 10.72 indicating that the deep rock mass is globally secure the detailed forensic mapping in Figure 4a establishes that localized, shallow geomechanical distress actively propagates parallel to the spillway chute. Severe subsurface micro-discontinuities and an active crown crack line (delineated by the prominent crack trace) triggered secondary failure surfaces with localized wedge slip profiles (FS = 11.2), driving the physical degradation of the slope flanking the left shoulder of the spillway structure. To eliminate this active driving force, the numerical model dictates an aggressive mass reduction campaign to flatten the slope profile above the primary crack boundary down to a stable, optimized angle of 45°. As detailed in the topographic stabilization layout of Figure 4b, achieving

this mitigated slope geometry requires the targeted excavation and disposal of 1,700.50 m³ of rock mass, precisely split between 921.50 m³ of hard rock and 779.00 m³ of moderate-to-soft rock materials.

This geomechanical vulnerability is evaluated via advanced limit equilibrium numerical simulations performed in SLOPE/W, as documented in Figure 4. While the deeper, global non-circular slip surface simulation yields a nominal static Factor of Safety of 10.72 indicating that the deep rock mass is globally secure the detailed forensic mapping in Figure 4a establishes that localized, shallow geomechanical distress actively propagates parallel to the spillway chute. Severe subsurface micro-discontinuities and an active crown crack line (delineated by the prominent crack trace) triggered secondary failure surfaces with localized wedge slip profiles (FS = 11.2), driving the physical degradation of the slope flanking the left shoulder of the spillway structure. To eliminate this active driving force, the numerical model dictates an aggressive mass reduction campaign to flatten the slope profile above the primary crack boundary down to a stable, optimized angle of 45°. As detailed in the topographic stabilization layout of Figure 4b, achieving

this mitigated slope geometry requires the targeted excavation and disposal of 1,700.50 m³ of rock mass, precisely split between 921.50 m³ of hard rock and 779.00 m³ of moderate-to-soft rock materials.

To guarantee global slope stability and safeguard construction personnel during the rehabilitation lifecycle, execution protocols must strictly adhere to operational constraints derived from these minimum analytical envelopes. Ground-breaking and rock mass removal must execute systematically in a top-down sequence, progressing from the crest downward to the toe, to prevent the initialization of dangerous, retrogressive slope failure profiles beneath un-cleared overhead catchments. Concurrently, surface grading and stabilization leveling at the programmed 45° profile must cleanly bisect and eliminate the active crown crack line, structurally removing the tensile failure root zones driving secondary slope degradation. Finally, based on the critical limit state thresholds identified when vertical cut heights (H) approach 40m, the engineering specifications dictate that for vertical slope intervals exceeding an elevation threshold of 25m, structural catch benches featuring a minimum horizontal width of 4.0m must be cleanly integrated into the profile, or alternatively, active rock bolts and tensioned anchors must be deployed to provide permanent lateral restraint against wedge initialization.

3.3. Reservoir Volumetric Losses and Seepage Analysis

To contextualize the physical and systemic vulnerabilities of the water asset, the forensic investigation coupled direct hydrogeological field metrics with a rigorous analysis of the reservoir’s capacity data and subsequent remedial design requirements. As illustrated by the field documentation in Figure 5, direct visual evidence captures a severe, almost flooding hydraulic short-circuiting where heavy, unmitigated seepage discharge continuously emerges from the highly jointed and fractured left abutment dolerite rock strata, bypassing the dam core entirely to drain directly into the downstream river channel. Quantified via the calibrated floating velocity array method (Figure 6a), these hydrogeological assessments captured a catastrophic average daily seepage discharge rate of 568 lit/sec (0.568 m³/s). This measured flow stands in stark, unacceptable contrast to the dam’s original safe engineering design baseline, which was modeled at a tight 0.61 lit/sec (0.00061 m³/s). This massive hydraulic short-circuiting indicates that major macro-porosity bypass vectors are actively cutting through the jointed dolerite channels directly adjacent to the spillway chute. Beyond introducing extreme hydrostatic backpressure that led to the tensile separation and progressive physical collapse of the rigid stone masonry gravity retaining structures flanking the spillway, this unmanaged subsurface flow inflicts an aggressive drain on the reservoir’s primary asset: its stored water volume.

This continuous drainage at a rate of 568 lit/sec generates devastating, compounding volumetric losses over time. On a daily operational basis, the dam wastes approximately 49,075.2 cubic meters (CUM) of water. Extended across key chronological milestones, this high-velocity drainage leads to a cumulative volumetric loss of 687,052.8 CUM within just 14 days of sustained flow. Over an extended two-month operational window of 62 days, the total volume of lost water escalates to a massive 3,042,662.4 CUM. When projected across a full calendar year (365 days), the cumulative volume lost through the left abutment fractures reaches a staggering 17,912,448 CUM, which equates to losing over 71.6% of the reservoir’s total gross design capacity (25 Mm³) every single year solely to subsurface seepage. Compounding these numbers across the 12 years that have elapsed since the dam’s construction completion reveals a staggering historical deficit: a total cumulative volume of 214,949,376 CUM (approximately 215 Mm³) of water has been lost to the subsurface environment. This astronomical waste represents more than 8.5 times the entire gross storage capacity of the reservoir completely drained away.

For a marginalized, socio-economically vulnerable, and impoverished society facing chronic water scarcity, recurrent droughts, and severe food insecurity, this lost volume represents a profound human tragedy. Had this water been successfully retained, it could have completely transformed local livelihoods by guaranteeing year-round domestic water security, supplying clean municipal water to thousands of citizens, and consistently driving multi-season agricultural production.

Consequently, this physical failure has directly cascaded into severe, long-term quantitative losses in reservoir storage and systemic infrastructural underperformance, a crisis explicitly illustrated by the cumulative operational database curves in Figure 6b. Currently, the reservoir utilizes a historical maximum of a mere 20% of its total design storage volume. Crucially, while initiatives are underway to finally construct and complete the downstream irrigation network and the second municipal outlet pipeline, this investigation delivers a stark warning: unless these subsurface seepage losses are aggressively managed and sealed first, the infrastructure currently under construction will completely fail to function. Building distribution canals and laying pipelines to carry water that does not exist in the reservoir will inevitably result in another high-profile engineering and financial mess, compounding regional asset underperformance. Therefore, the primary engineering defense must involve the immediate execution of a deep, multi-stage subsurface pressure-grouting curtain extended into the left abutment rock mass alongside an upstream clay blanket to intercept and seal the open joint networks.

Hydrogeologically, this integrated barrier is designed to force the phreatic line downward, driving the excessive 568 lit/sec seepage velocity back toward safe baseline limits. By cutting off this subsurface flow, the design drastically lowers back-of-wall hydrostatic pore pressures, creating a stable environment essential for replacing the failed masonry walls with optimized cantilever reinforced concrete retaining structures, thereby securing the hydraulic volume necessary to make the expanding downstream infrastructure viable.

Hydrogeological Mass-Balance and Baseflow Quantification

Crucially, a comparative evaluation between the raw cumulative seepage loss rate and the empirical reservoir drawdown records plotted in Figure 6a,b reveals an apparent volumetric discrepancy. The net reduction in reservoir storage volume (ΔV_observed) over any elapsed chronological window (Δt) does not match the total theoretical volume computed exclusively from the baseline subsurface seepage discharge (Q_seepage = 0.568 m³/s). This systematic deviation is driven by the continuous, stabilizing contribution of the upstream perennial baseflow (Q_baseflow) entering the reservoir body along the Gabbat River channel, which partially dampens the net drawdown rate.

To formalize this hydrogeological relationship and isolate the baseflow contribution, the reservoir’s volumetric mass-balance equation during non-rainy, stable monitoring intervals where active spillway overtopping is absent and controlled irrigation delivery is at 0% is defined as follows:

$${\displaystyle \frac{{\mathsf{\Delta }\mathbf{V}}_{\mathit{O}\mathit{b}\mathit{s}\mathit{e}\mathit{r}\mathit{v}\mathit{e}\mathit{d}}}{\mathsf{\Delta }\mathbf{t}}}=\left({\mathit{Q}}_{\mathbf{s}\mathbf{e}\mathbf{e}\mathbf{p}\mathbf{a}\mathbf{g}\mathbf{e}+}{\mathit{Q}}_{\mathbf{m}\mathbf{u}\mathbf{n}\mathbf{i}\mathbf{c}\mathbf{i}\mathbf{p}\mathbf{a}\mathbf{l}+}{\mathit{Q}}_{\mathit{e}\mathit{v}\mathit{a}\mathit{p}}\right)-{\mathit{Q}}_{\mathbf{b}\mathbf{a}\mathbf{s}\mathbf{e}\mathbf{f}\mathbf{l}\mathbf{o}\mathbf{w}}$$

(2)

where:

• ${\mathsf{\Delta }\mathbf{V}}_{Observed}$: is the net change in reservoir storage volume extracted from the operational curves over an elapsed duration Δt (m³).

• $Q_{\mathbf{b}\mathbf{a}\mathbf{s}\mathbf{e}\mathbf{f}\mathbf{l}\mathbf{o}\mathbf{w}}$: is the incoming perennial baseflow from the upstream catchment (m³/s).

• $Q_{\mathbf{s}\mathbf{e}\mathbf{e}\mathbf{p}\mathbf{a}\mathbf{g}\mathbf{e}}$: is the forensic field-verified subsurface bypass rate through the fractured foundation (0.568 m³/s or 49,075.2 m³/day).

• $Q_{\mathbf{m}\mathbf{u}\mathbf{n}\mathbf{i}\mathbf{c}\mathbf{i}\mathbf{p}\mathbf{a}\mathbf{l}}$: is the baseline municipal water supply line extraction rate delivered to Mekelle city (m³/s).
• Q_evap: is the micro-climate volumetric evaporation loss rate across the reservoir surface area (m³/s).

Rearranging the governing mass-balance equation allows for the direct, empirical calculation of the upstream baseflow contribution (Q_baseflow):

$$Q_{\mathbf{b}\mathbf{a}\mathbf{s}\mathbf{e}\mathbf{f}\mathbf{l}\mathbf{o}\mathbf{w}}=\left(Q_{\mathbf{s}\mathbf{e}\mathbf{e}\mathbf{p}\mathbf{a}\mathbf{g}\mathbf{e}+}{Q}_{\mathbf{m}\mathbf{u}\mathbf{n}\mathbf{i}\mathbf{c}\mathbf{i}\mathbf{p}\mathbf{a}\mathbf{l}+}{Q}_{evap}\right)-{\displaystyle \frac{{\mathsf{\Delta }\mathbf{V}}_{Observed}}{\mathsf{\Delta }\mathbf{t}}}$$

To illustrate this balancing mechanism using representative seasonal monitoring data where the reservoir exhibits an empirical net volume drawdown (${\mathsf{\Delta }\mathbf{V}}_{Observed}$) of −1,935,000 m³ over a stable 2-month elapsed operational window (Δt = 60 days or 5,184,000 seconds), while the active single municipal pipeline pumps designed at a baseline rate of $Q_{\mathbf{m}\mathbf{u}\mathbf{n}\mathbf{i}\mathbf{c}\mathbf{i}\mathbf{p}\mathbf{a}\mathbf{l}}$= 0.250 m³/s. However, during inspection the pumps were working at a rate of 150 m³/s and mean evaporation accounts for Q_evap = 0.042 m³/s:

• Compute the Net Observed Volumetric Drawdown Rate:

$${\displaystyle \frac{{\mathsf{\Delta }\mathbf{V}}_{Observed}}{\mathsf{\Delta }\mathbf{t}}}=-\mathrm{1,935,000}{\mathrm{m}}^3/\mathrm{5,184,000}\mathrm{s}\approx -0.373{\mathbf{m}}^3/\mathbf{s}$$

2.

Isolate and Compute the Upstream Baseflow ($Q_{\mathbf{b}\mathbf{a}\mathbf{s}\mathbf{e}\mathbf{f}\mathbf{l}\mathbf{o}\mathbf{w}}$):

Q_baseflow = (0.568 m³/s + 0.150 m³/s + 0.042 m³/s) − 0.373 m³/s

Q_baseflow = 0.760 m³/s − 0.373 m³/s = 0.372 m³/s (or 372.24 L/s)

This hydrogeological computation demonstrates that an upstream perennial baseflow influx of approximately 0.372 m³/s (372.24 L/s) continuously recharges the reservoir basin. This incoming flow actively offsets a major portion of the gross 0.568 m³/s subsurface seepage loss. This explains why the net empirical storage reduction slope in Figure 6b appears less aggressive than a curve projected purely from unmitigated subsurface piping. Explicitly separating the upstream baseflow from the subsurface short-circuiting vectors eliminates the data ambiguity between Figure 6a and Figure 6b, confirming that the reservoir is trapped in a critical state where seepage exceeds the natural baseflow buffering capacity, resulting in long-term capacity underutilization.

3.4. Structural Sizing, Sizing Code Optimization, and Limit States

Forensic structural and geomechanical evaluations confirm that the localized collapse flanking the left spillway abutment was driven by structural design deficiencies under extreme boundary conditions and the original stone masonry wall failed primarily due to construction drawbacks of insufficient base dimensioning and a lack of functional weep-hole drainage networks [5]. This structural arrangement allowed high hydrostatic pressure to develop behind the wall stem, as systematically documented across the structural schematics and field records in Figure 7. As detailed in the architectural plan-view maps and cross-sectional elevations (Figure 7a), the original unreinforced stone masonry gravity wingwall was anchored directly onto a steep, weathered rock cut. Field forensic tracking (Figure 7b) reveals that the wall experienced progressive downstream leaning and multi-axial sliding displacement along its base interface. This kinematic displacement culminated in the catastrophic scenario captured in Figure 7c, where the masonry gravity structure suffered complete tensile separation, diagonal shear cracking, and total structural failure under excessive overturning forces.

Technically, the forensic diagnostic indicates that this structural failure was driven by a fatal combination of insufficient structural base dimensioning, lack of internal steel reinforcement, and a complete absence of a functional weep-hole drainage network. This design omission allowed the unmitigated 568 lit/sec left-abutment subsurface seepage to pool directly against the back face of the wall stem, driving up back-of-wall pore-water pressures. The accumulation of this massive hydrostatic load, combined with active lateral earth pressures from the poor-rock dolerite slope mass (RMR = 40), pushed the sliding and overturning failure planes beyond ultimate limit states. Because the masonry gravity wall relied solely on its self-weight and mortar shear strength to maintain structural equilibrium, it possessed zero capacity to resist the sudden, localized tensile stress concentrations, causing rapid, brittle structural collapse.

To rectify these vulnerabilities and guarantee multi-decade structural reliability, the failed gravity sections have been completely demolished and replaced with a high-durability, rigid cantilever reinforced concrete (RCC) structure optimized to Eurocode 2 (EN 1992-1-1), Eurocode 7, and EBCS-7 limit state standards [26]. This updated engineering design, synthesized in Figure 8, incorporates a tapered structural stem profile configured to adapt smoothly to the local natural terrain slope baseline at 1861.20 m, with vertical wall heights varying from a baseline of 4.0 meters up to a peak structural height of 7.20 meters at the deepest spillway interface. Structurally, the cantilever configuration optimizes stability by using the weight of the backfill material resting on its extended heel slab to generate a powerful vertical restoring moment, directly counteracting overturning and sliding forces. Furthermore, to eliminate the root cause of the initial failure, the new structural layout integrates a systematic back-of-wall gravel filter medium coupled with a staggered grid of horizontal weep-holes to freely drain any residual subsurface flows, thereby preventing the re-development of destructive hydrostatic backpressures and ensuring that the newly completed distribution infrastructure can operate safely without further structural risks.

4. Conclusions and Recommendations

4.1. Conclusions

This research executed a comprehensive forensic engineering and structural lifecycle diagnosis of the multi-purpose Gerebsegen Dam in northern Ethiopia, uncovering a critical intersection of geomechanical, hydrogeological, and institutional failure mechanisms. A profound divergence is established between the project’s chronological maturity and its socioeconomic asset utility. While the dam was explicitly engineered for a 30-year operational design life, 12 years have elapsed since construction completion with the reservoir delivering a mere 20% of its intended volumetric utility. This minimal output serves exclusively for municipal water supply through just one active pipeline representing a restricted 50% delivery rate for the municipal sector while the secondary pipeline remains completely dormant. Crucially, the planned 500-hectare smallholder agricultural irrigation scheme remains completely unexecuted, resulting in a 0% service delivery that points directly to deep-seated institutional mismanagement, administrative irresponsibility, leadership incompetency, or the restrictive outputs of localized patronage networks.

Compounding these operational and governing failures, the physical collapse and deep structural cracking observed along the left spillway abutment gravity retaining walls were driven by intense lateral earth pressures and unmanaged back-of-wall hydrostatic loads. Geomechanical kinematics reveal that the underlying slope-forming Dolerite unit possesses a “Poor Rock” classification (Rock Mass Rating, RMR = 40) with a low internal friction angle (φ = 17.5°), making steep vertical cuts highly dependent on its internal cohesion (C = 150 kPa). Without internal weep-hole relief networks, this configuration triggered progressive overturning and sliding along the structural interface. This geomechanical vulnerability was critically accelerated by hydrogeological short circuiting, where field diagnostics identified an unallowable, massive subsurface seepage rate of 0.568 m³/s bypassing the core via fractured dolerite channels. This preferential flow pathway caused both massive hydraulic volume loss and rapid pore-water pressure spikes behind the rigid stone masonry walls, leading to rapid shear strength degradation at the rock-structure boundary.

4.2. Recommendations

To mitigate these active physical failures, restore long-term structural safety, and reform project delivery frameworks across East African water infrastructure assets, several integrated engineering and administrative interventions must be systematically implemented. First, the vulnerable masonry structures must be completely replaced with a rigid, tapered cantilever reinforced concrete structure varying in height up to 7.20 meters, systematically designed to Eurocode 2 and EBCS-7 limit state standards. To eliminate driving forces above the main tension crack line, a 1,700.50 m³ upper-slope offloading excavation must be executed to achieve a stable 45° profile. Mechanically, a systematic subsurface pressure-grouting curtain alongside an upstream clay blanket must be deployed through the jointed dolerite formation. This barrier is essential to cut off preferential flow paths, reduce the excessive 0.568 m³/s seepage rate, and relieve destructive structural backpressure behind the newly constructed abutment wall.

On an administrative and operational level, regional authorities must immediately shift from a fragmented project execution model to an integrated lifecycle asset management protocol to rectify the severe capacity underutilization observed over the last 12 years. Leadership bodies must prioritize the immediate technical installation, testing, and commissioning of the dormant second municipal water transmission pipeline, elevating the municipal water supply delivery from its restricted 50% capacity to its full 100% design rate for Mekelle city. Concurrently, the regional governance framework must fast-track the funding and construction of the primary, secondary, and tertiary downstream distribution canals necessary to activate the completely unexecuted 500-hectare agricultural command area, thereby elevating the irrigation sector from 0% utility to full development.

Finally, to safeguard structural modifications and provide a data-driven early warning baseline against future failures, a comprehensive instrumentation array must be installed immediately across the rehabilitated spillway flanks. This active network must include multi-stage reservoir level staff gauges spanning elevations from 1835 m to 1864 m to correlate hydraulic head with structural stress, double flow monitoring V-notch arrays featuring distinct 50 L/s and 200 L/s discharge capacities to precisely track volumetric seepage trends, and a gridded network of vibrating-wire piezometers. The real-time integration of these devices is mandatory to enable early detection of pore-water anomalies and phreatic surface variations within the poor-rock dolerite slope units before they propagate into unmanageable structural failure cascades. Future multi-purpose water resource assets within the East African Rift System must standardize these limit state simulations, sub-surface seepage barriers, and instrumentation frameworks during primary design phases to ensure multi-outlet infrastructures deliver their planned socioeconomic utility across their designated lifecycle.

Ethics Statement: The authors confirm that this study was conducted in strict accordance with the academic, scientific, and professional ethical standards mandated by the publisher. This research is based entirely on non-experimental field assessments, numerical simulation modeling, and institutional engineering data records obtained from public water agencies. It does not involve human participants, animal testing, or clinical trials; consequently, specific institutional review board (IRB) approval was not required.

Declaration of Competing Interest: The authors declare that they have no known competing financial interests, professional conflicts, or personal relationships that could have influenced the work reported in this paper.