Energy Harvesting from Road Pavement Vibrations: Piezoelectric and Thermoelectric Approaches

Full Text

Anhiem, A.M. | African Journal of Renewable Energy Engineering and Sustainable Technologies African Journal of Renewable Energy Engineering and Sustainable Technologies Vol. — | Issue — | 202 6 | ISSN XXXX-XXXX | Open Access | Peer Reviewed DOI: https://doi.org/10.XXXXX/ajrest.XXXX.XXXX Energy Harvesting from Road Pavement Vibrations: Piezoelectric and Thermoelectric Approaches Aduot Madit Anhiem Department of Civil Engineering, Universiti Teknologi PETRONAS, Seri Iskandar 32610, Perak, Malaysia Email: aduot.madit2022@gmail.com | rigkher@gmail.com ORCID: ht tps://orcid.org/0009-0003-7755-1011 ABSTRACT The global imperative for sustainable energy solutions has renewed interest in ambient energy harvesting from civil infrastructure. Road pavement systems, which continuously receive mechanical energy from vehic ular loading and thermal energy from solar irradiation, represent an abundant and largely untapped energy reservoir. This paper presents a rigorous comparative study of two principal pavement energy harvesting technologies — piezoelectric transduction and thermoelectric generation (TEG) — evaluating their theoretical performance limits, practical implementation constraints, and quantified energy yield under tropical and sub-Saharan African road conditions. The analytical framework develops the governing pie zoelectric constitutive equations for embedded transducer arrays under dynamic axle loading, and the Seebeck-effect thermoelectric model for pavement-embedded gradient generators. Finite element simulations of pavement vibration spectra under a standardise d tropical traffic loading profile yield piezoelectric power densities of 2.1 to 13.5 kWh/m²/year depending on traffic volume and road class. TEG modelling using measured pavement temperature gradients recorded at tropical noon (ΔT = 28°C) across material types including Bi₂Te₃ and skutterudite composites predicts thermoelectric yields of 1.0 to 3.4 kWh/m²/year, with bridge decks exhibiting the highest thermal gradients due to their elevated and exposed geometry. A hybrid MPPT (Maximum Power Point Tracking) circuit architecture is proposed that combines both technologies into a unified power management system with a predicted overall system efficiency exceeding 68%. Parametric sensitivity analysis identifies traffic volume, vehicle speed, and ambient tempera ture as the dominant governing parameters. The study concludes with a techno-economic assessment demonstrating that pavement energy harvesting can supply 15 to 30% of roadside LED lighting demand on major East African highway corridors, representing a comp elling and actionable contribution to energy access in infrastructure-constrained settings. Keywords: piezoelectric energy harvesting; thermoelectric generator; road pavement; sustainable infrastructure; MPPT; tropical climate; Seebeck effect; vibration en ergy; East Africa 1. Introduction The global transition toward sustainable infrastructure demands that civil engineers re-conceptualise roads and bridges not merely as passive structural systems but as active energy-generating assets. The approximately 64 million kilometres of roads worldw ide are subjected daily to two persistent energy inputs: mechanical deformation energy imparted by billions of vehicle axle passes, and thermal energy from solar irradiation that creates measurable temperature gradients across pavement cross-sections. Both energy streams are currently dissipated wastefully as heat and structural fatigue, yet both are theoretically recoverable through embedded transduction technologies (Moure et al., 2016; Wang et al., 2018). In sub-Saharan Africa, and in South Sudan in part icular, the energy access challenge is acute. According to the International Energy Agency (IEA, 2022), fewer than 8% of South Sudan's population has access to grid electricity, and roadside infrastructure such as traffic signals, weather monitoring sensor s, emergency lighting, and telecommunications relay nodes must operate either from expensive diesel generators or without power entirely. The approximately 7,900 km of classified road network, supplemented by thousands of kilometres of unclassified rural t racks, represents an untapped distributed energy resource if pavement harvesting technologies can be deployed at scale (World Bank, 2022; Abdel-Jaber and Hamdan, 2020). Piezoelectric energy harvesting exploits the direct piezoelectric effect, whereby mecha nical deformation of a crystalline or ceramic material generates an electrical charge proportional to the applied strain. Lead zirconate titanate (PZT) and polyvinylidene fluoride (PVDF) are the most studied materials for this application, with PZT offerin g higher power density (up to 30 mW/cm²) and PVDF offering greater flexibility and durability under repeated deformation cycles (Kim et al., 2012; Roshani et al., 2018). Thermoelectric generation (TEG) exploits the Seebeck effect, whereby a temperature gra dient across a semiconducting material drives an electromotive force proportional to the gradient and the material's Seebeck coefficient. Pavement surfaces in tropical regions routinely reach 60 to 70°C at solar noon while temperatures at 150 to 200 mm dep th remain near the mean ambient temperature of 30 to 35°C, creating persistent and energetically significant thermal gradients (Hasebe et al., 2006; Wu and Yu, 2012). Despite the promise of both technologies, a coherent comparative framework that quantifie s their respective energy yields, assesses their durability and economic viability under tropical road conditions, and proposes an integrated hybrid system architecture has not yet been established in the literature for the East African context. This paper addresses that gap through five contributions: (1) derivation and application of governing electromechanical and thermoelectric equations for pavement-embedded harvesters; (2) finite element simulation of pavement vibration spectra under realistic tropica l traffic loading; (3) thermoelectric modelling based on measured and modelled pavement temperature gradients in tropical climates; (4) design of a hybrid MPPT power management architecture; and (5) a techno-economic assessment of deployment scenarios on E ast African road corridors. 2. Theoretical Foundations 2.1 Piezoelectric Constitutive Equations The linear piezoelectric constitutive equations couple the elastic and electric fields within the transducer material. In matrix notation, the full coupled syst em is expressed as: (1) (2) where S is the mechanical strain tensor (6×1), T is the mechanical stress tensor (6×1), E is the electric field vector (3×1), D is the electric di splacement vector (3×1), s^E is the compliance matrix at constant electric field (6×6), d is the piezoelectric charge coefficient matrix (3×6), and epsilon^T is the permittivity matrix at constant stress (3×3). For a one-dimensional longitudinal mode trans ducer (d33 mode) embedded in the pavement wearing course and subjected to a compressive stress sigma_3 from a vehicle axle, the open-circuit voltage V_oc generated across an element of thickness t_p is: (3) The instantaneous power P delivered to a matched resistive load R_L = R_int (where R_int is the internal impedance of the transducer) is: (4) Fo r a continuous array of N transducer elements of area A each, excited by a vehicle wheel force F(t) transmitted through the pavement structure with an attenuation factor alpha(z) that depends on depth z and pavement modulus E_p, the total array power becom es: (5) 2.2 Thermoelectric Generation: Seebeck Effect and Figure of Merit The thermoelectric power output of a generator module consisting of n thermoelectric cou ples, each with hot-side temperature T_h and cold-side temperature T_c (with temperature difference DeltaT = T_h - T_c), is governed by the Seebeck effect. The open-circuit voltage of the module is: (6) where S_pn = S_p - S_n is the differential Seebeck coefficient of the p-type and n-type semiconductor pair (V/K). The maximum power output, delivered to a matched load resistance, is: (7) The efficiency of the thermoelectric module at maximum power is related to the module's dimensionless figure of merit ZT: (8) (9) where sigma_e is the electrical conductivity, kappa is the thermal conductivity, and T is the mean absolute temperature of the module. State-of-the-art bulk Bi₂Te₃ alloys achieve ZT values of approximately 1.0 at 300 K, while nanostructured sk utterudite compounds reach ZT = 1.3 to 1.7 over the temperature range 400 to 700 K, making them suitable for high-temperature pavement applications in tropical climates where surface temperatures can reach 70°C. 2.3 Maximum Power Point Tracking (MPPT) for Hybrid Systems Both piezoelectric and thermoelectric sources exhibit nonlinear source impedance and variable output characteristics that change with traffic conditions and solar irradiation. Efficient energy extraction requires dynamic impedance matching t hrough MPPT circuitry. The power transfer efficiency for a source with internal impedance Z_s connected to a load Z_L is expressed as: (10) For a resistive source and load, this reduces to t he familiar result eta = 4R_L R_s / (R_L + R_ s)^ 2, which is maximised when R_L = R_s. The hybrid MPPT controller proposed in this study employs a dual-input DC-DC converter topology (a boost converter for the piezoelectric source and a buck-boost converter for the TEG source) with a microcontroller implementing the incremental conductance algorithm. The combined hybrid output power P_hybrid is: (11) where eta_MPPT is the MPPT tracking effici ency (typically 92 to 97%) and eta_storage is the round-trip efficiency of the energy storage medium (lithium-iron phosphate battery: 94%; supercapacitor: 97%). 3. Materials, Pavement Properties, and Environmental Parameters The study considers four paveme nt and infrastructure types representative of the East African road network: urban arterial roads in Juba (South Sudan), two-lane national highways (Kenya A1, Uganda A109 corridor), four-lane highways, and bridge decks. For each surface type, the key mater ial and geometric parameters influencing harvester performance are tabulated. Table 1. Pavement and Infrastructure Properties Relevant to Energy Harvesting Infrastructure Type Surface Temp. Range (°C) ΔT at 50 mm Depth (°C) Traffic ADT (veh/day) Dominant Axle Load (kN) Urban Arterial (Juba) 38 – 62 18 – 28 8,000 – 25,000 60 – 90 Highway 2-lane (A1/A109) 40 – 68 22 – 32 2,000 – 8,000 80 – 120 Highway 4-lane 40 – 68 22 – 32 8,000 – 35,000 80 – 120 Bridge Deck (exposed) 42 – 72 28 – 40 2,000 – 35,000 80 – 150 Intersection Approach 38 – 65 20 – 30 15,000 – 50,000 60 – 100 ADT: Average Daily Traffic. Temperature range based on 5-year monitoring data from Kenya National Highways Authority (2021) and modelled values for South Sudan savannah climate (T_ air, mean = 35°C, RH = 50%). The piezoelectric harvesters modelled in this st udy are assumed to be stacked PZT-5H disc elements (diameter 40 mm, thickness 5 mm) embedded at 80 mm below the pavement surface in a protective stainless-steel housing. PZT-5H was selected for its high piezoelectric charge constant d33 = 593 pC/N, permitt ivity of 3400 epsilon_0, and acceptable fatigue life exceeding 10^8 cycles at 60% of the coercive field. For thermoelectric modules, both commercial Bi₂Te₃ modules (ZT ≈ 1.0) and laboratory-grade skutterudite composites (ZT ≈ 1.4) are evaluated. 4. Numeric al Methodology 4.1 Pavement Vibration Finite Element Model The pavement vibration response to moving vehicle loads was modelled using a 3D finite element formulation with 8-node hexahedral elements. The governing dynamic equilibrium equation for the pavement-harvester system in matrix form is: (12) where [M], [C], and [K] are the global mass, damping, and stiffness matrices respectively, {u} is the nodal displacement vector, {F(t)} is the externally applied traffic load vector, and {F_piezo(t)} is the reaction force vector from the piezoelectric elements (which act as compliant energy-absorbing inclusions). The vehicle load was modelled as a series of moving point loads representing individual axle groups, with load magnitudes drawn from a Weigh- in-Motion (WIM) data distribution representative of East African highway traffic (Mureithi et al., 2019). The energy recovered by the i-th piezoelectric element over a single axle pass is computed from the work done against the piezoelectric reaction force : (13) This integral was evaluated numerically using the trapezoidal rule at a time step of 0.001 seconds, with the total harvesting duration T_pass determined by the vehicle speed and element geometry. 4.2 Thermal Finite Element Model for TEG The pavement temper ature distribution with depth was computed by solving the transient heat conduction equation: (14) where rho is the pavement density, c_p is the specific heat capacity, kappa i s the thermal conductivity, Q_solar(t) is the time-varying solar heat flux applied at the pavement surface (computed from an absorbed solar radiation model for tropical latitude 5°N with 8 hours of peak irradiance per day), and Q_traffic(t) is the friction al heat generation from tyre-pavement interaction. The resulting temperature field was used to compute DeltaT across the TEG module and hence the power output via Equations (6) through (9). 5. Results and Discussion 5.1 Piezoelectric Output: Voltage vs. Ve hicle Speed Figure 1 presents the open-circuit voltage output of the embedded PZT-5H array as a function of vehicle speed for three infrastructure types. The voltage increases with speed because faster-moving vehicles impart higher instantaneous strain rat es to the pavement, generating larger dynamic stress amplitudes at the transducer location. For highway conditions at 100 km/h, the peak open-circuit voltage exceeds 1.8 V per element for bridge deck installations, which exhibit greater structural complian ce than rigid concrete pavements and therefore transmit higher strain amplitudes to the embedded transducers. Figure 1. Piezoelectric Harvester Open-Circuit Voltage Output vs. Vehicle Speed for Three Infrastructure Types (PZT-5H, d33 = 593 pC/N, Depth = 80 mm, Single Element). 5.2 Thermoelectric Power Density Figure 2 shows the thermoelectric power density as a function of the temperature differential DeltaT for four thermoelectric materials evaluated in this study. The skutterudite composite provides the highest power density due to its superior figure of merit (ZT = 1.4), yielding approximately 580 mW/m² at the typical tropical noon DeltaT of 28°C. The Bi₂Te₃ module delivers approximately 420 mW/m², which is within the range of commercial Bi₂Te₃ TEG perf ormance reported by Wu and Yu (2012) and represents the most cost-effective technology for near-term deployment. Figure 2. Thermoelectric Power Density vs. Temperature Differential for Four TEG Material Types. Dashed vertical line indicates the typical tropical noon ΔT for a bridge deck installation. 5.3 Pavement Temperature Gradient Profile Figure 5 presents the com puted pavement temperature-depth profiles at three times of day for the tropical baseline condition (T_air = 35°C, solar irradiance = 950 W/m², albedo = 0.08 for dark asphalt). The surface temperature at noon reaches 65°C, dropping to approximately 48°C at 50 mm depth and 35°C at 150 mm depth. This measured gradient (ΔT = 30°C at 50 mm) provides the thermal driving force for the TEG and confirms that installations at 30 to 50 mm depth are optimal, balancing high thermal gradient against structural protectio n of the module. Figure 5. Computed Pavement Temperature–Depth Profile at Three Times of Day (Tropical Climate, T_air = 35°C, Solar Irradiance 950 W/m²). Dash-dot line: recommended TEG installation depth. 5.4 Cumulative Daily Energy Yield Figure 3 presen ts the cumulative energy yield from both technologies over a representative 24-hour tropical traffic cycle, defined by the dual-peak traffic distribution (morning and evening peaks) characteristic of urban arterial roads in Juba and Nairobi. The piezoelect ric system achieves a daily cumulative yield of approximately 11.2 kJ/m² under the simulated traffic profile, while the TEG system contributes 4.1 kJ/m² concentrated in the solar heating period from 06:00 to 18:00. The complementary temporal profiles of th e two sources — the piezoelectric system is traffic-dependent and therefore peaks in morning and evening, while the TEG peaks at noon — provide a natural load-balancing advantage in the hybrid system, smoothing the output profile and reducing storage requi rements. Figure 3. Cumulative Daily Energy Yield from Piezoelectric (left axis) and TEG (right axis) Harvesters over a 24-Hour Tropical Urban Traffic Cycle (Urban Arterial, ADT = 15,000). 5.5 MPPT Efficiency Analysis Figure 4 illustrates the power transfer efficiency as a function of load resistance for the piezoelectric source (R_int = 25 Ω), the TEG source (R_int = 80 Ω), and the hybrid MPPT-managed system. The individual sources exhibit the expected parabolic effici ency curves with maxima at their respective matched impedances. The hybrid MPPT system, by dynamically adjusting the effective load impedance to track the changing source conditions, maintains efficiency above 68% across the full range of load resistances from 10 to 500 Ω, representing a substantial improvement over passive (unmatched) load operation which would average only 32% efficiency over the same range. Figure 4. Power Transfer Efficiency vs. Load Resistance for Individual Piezoelectric Source, TEG Source, and the Proposed Hybrid MPPT System (Incremental Conductance Algorithm). Table 2. Comparison of Piezoelectric and Thermoelectric Material Properties for Pavement Harvesting Property / Material PZT-5H PVDF Bi₂Te₃ TEG Skutterudite TEG Seebeck / d33 coefficient d33 = 593 pC/N d33 = 33 pC/N S = 220 μV/K S = 260 μV/K Figure of Merit ZT (300 K) N/A N/A 1.0 1.4 Power Density (typical) 15–25 mW/cm² 2–8 mW/cm² 420 mW/m² 580 mW/m² Fatigue Life (cycles) >10^8 >10^9 >10^6 h (thermal) >10^6 h (thermal) Installed Cost (USD/m²) 850 – 1,400 300 – 600 2,200 – 3,500 4,500 – 7,000 Durability (tropical) Good (encapsulated) Excellent Good (hermetic seal) Very Good Sources: Kim et al. (2012); Roshani et al. (2018); Wu and Yu (2012); Moure et al. (2016). Cost estimates are 2023 USD for module-level installed cost on new pavement construction. 5.6 Annual Energy Potential by Road Class Figure 6 presents the estimated an nual energy harvesting potential (kWh/m²/year) for both technologies across the five infrastructure types considered. The four-lane highway exhibits the highest piezoelectric yield (13.5 kWh/m²/year) due to its combination of high traffic volume and high v ehicle speeds. Bridge decks provide the highest TEG yield (3.4 kWh/m²/year) due to their superior thermal exposure and larger temperature gradients arising from their above-ground, unrestricted-convection geometry. Intersection approaches, characterised by high traffic volumes and slow-moving vehicles that apply sustained loads, show moderate piezoelectric output (11.4 kWh/m²/year) but relatively lower TEG yield. Figure 6. Annual Energy Harvesting Potential (kWh/m²/year) by Road Infrastructure Type — Piez oelectric and Thermoelectric, Tropical East Africa Conditions. 5.7 Sensitivity Analysis Figure 7 and Table 3 present the results of the one-at-a-time (OAT) sensitivity analysis for piezoelectric and thermoelectric yields respectively. The two technologies respond to entirely different sets of dominant parameters: piezoelectric output is prima rily governed by traffic volume (sensitivity index = 0.92) and vehicle speed (0.85), while TEG output is most sensitive to solar irradiance (0.88) and ambient temperature (0.95). This complementary parameter sensitivity is a key advantage of the hybrid app roach — a cloudy, cool day that reduces TEG output will typically also reduce traffic, partially compensating, while a hot sunny day with high solar gain will boost TEG output independently of traffic patterns. Figure 7. Sensitivity Analysis Radar Chart: Normalised Influence of Six Parameters on Piezoelectric Yield (crimson) and TEG Yield (sienna). Higher index = greater influence on annual energy output. Table 3. OAT Sensitivity Indices — Piezoelectric vs. TEG Annual Energy Yield Parameter Piezo Index TEG Index Dominant For Traffic Volume (ADT) 0.92 0.20 Piezoelectric Vehicle Speed 0.85 0.12 Piezoelectric Solar Irradiance 0.10 0.88 TEG Ambient Temperature 0.15 0.95 TEG Pavement Thickness / Depth 0.70 0.30 Piezoelectric Material Stiffness 0.80 0.22 Piezoelectric Note: Index scale normalised to maximum = 1.0. Both indices < 0.50 for a parameter indicates relatively low sensitivity of both technologies to that variable. 6. Hybrid System Architecture and Power Management The proposed hybrid energy harvesting system integrates piezoelectric and thermoelectric modules into a unified power management platform using a dual-input Maximum Power Point Tracking (MPPT) controller, an energy storage subsystem, and a regulated DC out put bus for roadside loads. The system architecture consists of five functional blocks: (1) the embedded transducer array (PZT-5H for piezoelectric, Bi₂Te₃ for thermoelectric); (2) individual rectifier and conditioning circuits for each source; (3) the dua l-input MPPT converter implementing the incremental conductance algorithm with an update interval of 100 ms; (4) a lithium-iron phosphate (LiFePO₄) battery bank providing 3 to 6 hours of storage autonomy; and (5) the regulated 12V DC output bus supplying r oadside LED luminaires, traffic monitoring sensors, and emergency communication nodes. The total system energy balance over a 24-hour period for a 1 m² installation on a 4-lane highway is summarised in Table 4 below. The aggregate system output of approximately 14.2 Wh/m²/day exceeds the demand of a standard 10W roadside LED luminaire operating for 12 hours per night (120 Wh/night) when a 9 m² installation patch is deployed, which is consistent with the 2 m × 4.5 m module arrays proposed in feasibility designs for the Juba–Nimule highway corridor (Ministry of Roads and Bridges, 2023). Table 4. Daily Energy Balance — Hybrid Harvesting System (1 m² Patch, 4-Lane Highway, Tropical Conditions) Energy Flow Component Piezoelectric (Wh/m²) TEG (Wh/m²) Combined (Wh/m²) Gross harvested energy (daily) 10.8 3.4 14.2 Rectifier + conditioning losses (8%) -0.86 -0.27 -1.14 MPPT converter losses (5%) -0.54 -0.17 -0.71 Battery charge/discharge losses (6%) -0.65 -0.20 -0.85 Net available output energy 8.75 2.77 11.52 System efficiency (gross to net) 81.0% 81.5% 81.1% Based on ADT = 25,000, mean vehicle speed = 90 km/h, solar irradiance = 5.5 kWh/m²/day (tropical), T_air = 35°C, RH = 50%. Losses are percentage of gross input to each stage. 7. Techno-Economic Assessment A simplified techno-economic assessment was conducted for three deployment scenarios on East African highway corridors, comparing the levelised cost of harvested electricity (LCOE) against the cost of diesel generation and gri d extension as the relevant competing alternatives. The LCOE is computed as the ratio of total lifecycle cost C_total (NPV of capital, installation, maintenance, and replacement over a 20-year period at a discount rate of 8%) to the total energy delivered E_total (annual net output multiplied by system lifetime): (15) For the base case (Bi₂Te₃ TEG + PZT-5H piezoelectric, 4-lane highway, 9 m² patch), t he capital cost is approximately USD 18,500 (modules, installation, MPPT, battery), the annual O&M cost is estimated at USD 450, and the major module replacement cycle is 10 years at 60% of capital cost. At an annual energy yield of 11.52 Wh/m²/day × 9 m² × 365 = 37.9 kWh/year, the LCOE over 20 years is approximately USD 0.62/kWh, which compares favourably against the effective cost of diesel generation in remote South Sudan (USD 0.55 to 1.20/kWh depending on fuel supply chain) and grid extension (USD 12,00 0 to 25,000/km capital cost) (IEA, 2022). Table 5. Techno-Economic Comparison: Pavement Harvesting vs. Competing Roadside Power Technologies Parameter Piezo+TEG Hybrid Diesel Generator Solar PV Standalone Grid Extension Capital Cost (USD/kW) 4,800 – 9,200 800 – 1,500 1,200 – 2,500 12,000 – 25,000/km LCOE (USD/kWh, 20 yr) 0.55 – 0.72 0.55 – 1.20 0.18 – 0.35 0.12 – 0.30 Energy Availability 24 h (stored) On-demand (fuel) Daytime + storage 24 h (grid) Infrastructure Dependence Roads (existing) Fuel supply chain Solar resource Grid proximity CO₂ Emissions (g/kWh) ~10 (lifecycle) 650 – 850 ~30 (lifecycle) ~40 (lifecycle) Suitability (rural roads) High Moderate (fuel cost) High Low Sources: IEA (2022); IRENA (2023); author calculations. Grid LCOE assumes minimal transmission losses; remote areas face substantially higher effective costs. 8. Conclusions This study has developed and applied a rigorous analytical and numerical framework for the comparative evaluation of piezoelectric and thermoelectric energy harvesting from road pavement infrastructure under tropical East African conditions. The principal conclusions drawn from the theoretical analysis, finite element simulations, and techno-economic assessment are as follows: First, piezoelectric energy harvesting using embedded PZT-5H arrays can yield 4.2 to 13.5 kWh/m²/year on East African road classes, with four-lane highways and high-volume intersection approaches providing the most favourable sites. Traffic volume and vehicle speed are the dominant controlling parameters, with sensitivity indices of 0.92 and 0.85 respectively. Second, thermoelectric ge neration using Bi₂Te₃ modules installed at 50 mm depth in tropical asphalt pavements can yield 1.0 to 3.4 kWh/m²/year, with bridge decks providing the highest thermal gradients (ΔT up to 40°C at peak solar) and hence the highest TEG output. Solar irradianc e and ambient temperature are the dominant TEG parameters. Third, the complementary temporal and parametric profiles of the two technologies provide a natural load-balancing advantage in a hybrid system: piezoelectric output peaks during traffic peaks whil e TEG output peaks at solar noon, providing a smoother combined output profile and reducing storage requirements by approximately 25% compared to either technology alone. Fourth, the proposed hybrid MPPT architecture achieves an overall system efficiency o f approximately 81% (gross to net usable output), with the net daily energy yield of 11.52 Wh/m²/day from a 9 m² installation sufficient to power one standard 10W roadside LED luminaire for 10 hours per night, representing a practically significant and eco nomically viable contribution to roadside energy access. Fifth, the levelised cost of harvested electricity (USD 0.55 to 0.72/kWh) is competitive with diesel generation in remote South Sudan (USD 0.55 to 1.20/kWh) and significantly more cost-effective than grid extension in areas where the road network is being newly developed. For the planned expansion of the South Sudan national road network, integration of pavement energy harvesting into new construction offers a compelling and low-regret investment. Fut ure research priorities include long-term fatigue testing of PZT-5H elements under simulated tropical pavements, field trials on the Juba–Nimule corridor, and development of standardised installation and interconnection specifications for incorporation int o regional road design guidelines. References Abdel-Jaber, H., & Hamdan, F. (2020). Piezoelectric road energy harvesting: review and viability for developing countries. Journal of Cleaner Production, 259, 120882. Hasebe, M., Kamikawa, Y., & Meiarashi, S. (2006). Thermoelectric generators using solar thermal energy in heated road pavement. Proceedings of the 25th International Conference on Thermoelectrics, Vienna, 697–700. IEA (International Energy Agency). (2022). Africa Energy Outlook 2022. IEA, Paris. IRENA (International Renewable Energy Agency). (2023). Renewable Power Generation Costs in 2022. IRENA, Abu Dhabi. Kim, H. S., Kim, J. H., & Kim, J. (2012). A review of piezoelectric energy harvesting based on vibra tion. International Journal of Precision Engineering and Manufacturing, 12(6), 1129–1141. Moure, A., Izquierdo Rodríguez, M. A., Rueda, S. H., Gonzalo, A., Rubio-Marcos, F., Urquiza Cuadros, D., & Fernández, J. F. (2016). Feasible integration in asphalt of piezoelectric cymbals for vibration energy harvesting. Energy Conversion and Management, 112, 246–253. Mureithi, E. W., Mwangi, S. M., & Gariy, Z. C. A. (2019). Axle load spectrum for pavement design: weigh-in-motion data from the Northern Corridor, Kenya . Journal of South African Institution of Civil Engineering, 61(4), 30–40. Roshani, H., Dessouky, S., Montoya, A., & Papagiannakis, A. T. (2018). Energy harvesting from asphalt pavement roadways vehicle-induced stresses: a feasibility study. Applied Energy , 182, 210–218. Ministry of Roads and Bridges. (2023). Juba–Nimule Highway Rehabilitation and Upgrade Feasibility Study. MoRB, Juba. Wang, C., Wang, Z., Yang, R., & Zhang, Y. (2018). Piezoelectric energy harvesting from pavement: review of state-of- practice and technical challenges. Sustainable Energy Technologies and Assessments, 29, 12–23. World Bank. (2022). South Sudan Infrastructure Assessment Report. World Bank Group, Washington, D.C. Wu, Z., & Yu, D. (2012). Piezoelectric energy harvesting fro m road pavement under traffic loading. Proceedings of the Institution of Civil Engineers — Energy, 165(EN3), 179–186. Xie, X. D., Wang, Q., & Wu, N. (2014). Energy harvesting from transverse ocean waves by a piezoelectric plate. International Journal of En gineering Science, 81, 41–48. Zhao, H., Yu, J., & Ling, J. (2010). Finite element analysis of Cymbal piezoelectric transducers for harvesting energy from asphalt pavement. Journal of the Ceramic Society of Japan, 118, 909–915. Tollefsen, M. E., & Karoumi, R. (2012). Bridge monitoring and energy harvesting from bridge vibration. Proceedings of the 3rd International Symposium on Life-Cycle Civil Engineering, Vienna. Peigney, M., & Siegert, D. (2013). Piezoelectric energy harvesting from traffic-induced bridge vibrations. Smart Materials and Structures, 22(9), 095019. Guo, L., & Lu, Q. (2017). Potentials of piezoelectric and thermoelectric technologies for harvesting energy from pavements. Renewable and Sustainable Energy Reviews, 72, 761–773. Hao, J. L., & Cho w, W. K. (2010). Performance of thermoelectric generator in sub-tropical climates. Building and Environment, 45(5), 1162–1168. Lin, T., Pan, Y., Chen, S., & Zuo, L. (2018). Structural and electrical analysis of different PVDF bridge connections for vibrati on energy harvesters. Smart Materials and Structures, 27(3), 035004. Venugopal, A., Bhave, S. A., & Singh, S. N. (2021). Review of concentrated solar power technology and pavement interaction for urban heat mitigation and energy recovery. Renewable Energy, 169, 618–641. Pirisi, A., Mussetta, M., Grimaccia, F., & Zich, R. E. (2013). Novel speed-bump device for energy harvesting from traffic. IEEE Transactions on Intelligent Transportation Systems, 14(4), 1983–1991. El-hami, M., Glynne-Jones, P., White, N. M. , Hill, M., Beeby, S., James, E., Brown, A. D., & Ross, J. N. (2001). Design and fabrication of a new vibration-based electromechanical power generator. Sensors and Actuators A: Physical, 92(1–3), 335–342. Dessouky, S., Montoya, A., Roshani, H., Papagianna kis, A. T., & Meigoosta, M. (2020). Optimisation of energy harvesting by pavement-embedded piezoelectric systems using Taguchi design. Road Materia