Project Overview: Plovdivtsi Dam
Plovdivtsi dam is located in southern Bulgaria near the Greek border, at approximately 900 m above sea level, forming part of a potable water supply system. Designed in the 1980s and constructed through a staged process, the dam stands approximately 50 m high at its deepest point, with a crest length of 225 m and a total reservoir capacity of 2.3 million m³.
The embankment consists of upstream and downstream rockfill shells, transition zones, filters, and a central asphalt-concrete diaphragm (ACD) founded on a reinforced concrete slab. The ACD functions as the primary impermeable barrier and is connected to the surrounding embankment through carefully graded filter and transition zones.
For the numerical investigation, these zoned material interfaces are particularly important because they directly influence permeability contrasts, stress redistribution, and transient pore pressure development during reservoir filling.
First Impoundment Loading Conditions
The initial filling of the Plovdivtsi reservoir was completed in just 18 days, with a significant portion of inflow occurring during a short-duration rainfall event.
This created a transient loading condition in which reservoir filling progressed faster than pore pressure equalization within parts of the embankment. For asphalt-core rockfill dams, this behavior is particularly important because low-permeability zones adjacent to the ACD can delay hydraulic response and influence the timing of stress redistribution and deformation development.
Following first filling, the reservoir underwent seasonal operational fluctuations, introducing repeated loading and unloading cycles that continued to influence long-term embankment response.
Understanding how these effects evolved over time formed one of the central objectives of the numerical investigation.
Numerical Modeling Approach
Methodology
A two-dimensional plane-strain finite element model of the dam–foundation–reservoir system was developed in DIANA using the highest instrumented cross-section of the dam.
The analysis explicitly simulated construction staging in 10 sequential layers, initial reservoir filling, reservoir drawdown, and subsequent operational loading cycles. This staged workflow was critical because the mechanical response of compacted rockfill evolves continuously during loading, unloading, and reloading.
The authors also note that some differences between numerical predictions and field measurements may partially originate from the plane-strain assumption, since the actual structure behaves three-dimensionally.
Constitutive Models
Different constitutive models were adopted to represent the embankment zones and foundation materials:
- Rockfill, filters, and transition zones: Extended Hardening Soil Model.
- Asphalt-concrete core: Elastoplastic Mohr–Coulomb model with hardening.
- Foundation and plinth: Linear elastic behavior.
The Hardening Soil Model was selected because it can represent stress-dependent stiffness, plastic shear hardening, dilatancy effects, and irreversible volumetric pre-consolidation. These mechanisms are particularly important for embankment dams because stiffness and deformation behavior evolve significantly as loading history develops.
To represent compaction effects during construction, a pre-consolidation pressure of 100 kPa was introduced for all embankment materials.
Hydraulic and Structural Analysis
The analysis was performed in two sequential stages:
- Transient groundwater flow analysis to determine pore pressure evolution.
- Nonlinear structural analysis using the computed pore pressure fields.
This workflow allowed transient hydraulic effects within low-permeability embankment zones to directly influence the calculated stress and displacement response during rapid reservoir filling.
Mesh and Boundary Conditions
The numerical domain extended approximately one dam height upstream, downstream, and below the structure to minimize boundary effects.
The model consisted of 2112 plane-strain quadrilateral and triangular finite elements with refined discretization in critical regions of the embankment.
Results and Interpretation
End of Construction (EoC)
At the completion of construction, the model predicted:
- Maximum horizontal displacement of approximately 13 mm.
- Maximum settlement of approximately 122 mm.
- Vertical stresses reaching approximately 900 kPa in the lower central embankment.
Stress contours generally followed the geometry of the embankment, indicating consistent load transfer through the dam body.
End of Impoundment (EoI) and Early Operation
The analysis revealed a more complex response during first impoundment.
Maximum horizontal displacement increased to approximately 40 mm on 5 December 2017, even though the reservoir had already passed its peak water level. The delayed response was associated with transient pore pressure redistribution within low-permeability embankment zones adjacent to the asphalt-concrete core.
This observation is one of the study’s most important findings because it demonstrates that maximum structural demand may not coincide with peak hydraulic loading conditions.
During subsequent operation:
- Maximum horizontal displacement increased only slightly, reaching approximately 43 mm.
- Additional settlement developed gradually during reservoir cycling.
- The overall response progressively approached a stable long-term trend associated with material hardening.
The results indicate that most deformation occurred during initial loading, while later operational cycles produced comparatively limited additional displacement.
Stress and Strain Behavior
The analysis showed that effective stresses evolved differently across the embankment zones during impoundment:
- Effective stresses decreased in the upstream shell.
- Horizontal stresses increased in the downstream shell.
- Vertical stresses remained comparatively stable.
Most importantly, the asphalt-concrete core remained well within acceptable stress and strain limits throughout all investigated loading stages.
The analysis showed:
- Compressive stresses along the ACD surfaces.
- Maximum strains of approximately 0.8%.
- No indication of nonlinear distress within the core material.
These results confirm satisfactory structural performance of the impermeable core despite the rapid filling conditions.
Conclusion
The analysis demonstrates that Plovdivtsi Dam performs within safe limits during both initial impoundment and early operation.
Three findings stand out from the study:
- Peak structural response did not coincide with maximum reservoir level due to delayed hydraulic equilibration within low-permeability zones
- Long-term displacement evolution gradually stabilized as a result of material hardening behavior captured by the constitutive models
- The asphalt-concrete core remained within its elastic range throughout construction, impoundment, and operational loading stages
The study also highlights the importance of staged nonlinear analysis for evaluating embankment dam behavior under transient loading conditions. Capturing the interaction between loading history, permeability contrasts, and constitutive hardening behavior was essential for reproducing the observed structural response during rapid first filling.