Shallow-water hydrokinetic energy extraction has been proposed as a potentially transformative technology for rural electrification in developing regions [1], [2], [3], with theoretical potential estimates ranging from 50 – 100 GW globally based on irrigation canal and low-head river assessments [4], [5]. Yet after more than 15 years of intensive research activity – evidenced by over 200 peer-reviewed publications since 2010 identified through Scopus database analysis – commercial deployment remains minimal, limited to scattered demonstration projects totaling less than 5 MW cumulative installed capacity worldwide [6], [7], [8]. This striking disconnect between research intensity and market penetration suggests either: First, fundamental technical barriers not yet resolved [9], [10], Second economic uncompetitiveness versus alternative distributed generation technologies [11], [12], or third systematic methodological optimism in research literature overestimating real-world viability [13], [14]. Critical context for this disconnect emerges from comparison with other renewable technologies developed during the same period. Between 2010 and 2025, solar photovoltaic (PV) systems experienced 89% levelized cost reduction – from 0.38 USD/kWh to 0.04 USD/kWh – while achieving over 1,000 GW global installed capacity [11], [15], [16], despite similar or inferior capacity factors (solar: 15 – 25%; hydrokinetic: 30 – 40% claimed) [17], [18]. If shallow-water hydrokinetic technology possessed economic viability comparable to solar PV, exponential deployment growth would be expected given its superior capacity factor and 24-hour generation potential [19], [20]. The absence of such growth – indeed, the near-absence of commercial deployment beyond pilot projects – demands critical examination of underlying assumptions pervading research literature.

Three competing explanations may account for this deployment gap. First, unresolved technical barriers – ventilation in shallow submergence [21], debris accumulation effects [22], or laboratory-to-field performance degradation [23] – may prevent reliable operation in real-world conditions despite impressive laboratory results. Studies documenting field installations report systematic performance shortfalls ranging from 15 – 45% below laboratory predictions [1], [7], [8], with debris fouling, seasonal flow variation, and structural vibration identified as primary degradation mechanisms. Biofouling alone can reduce efficiency by 5 –15% annually in tropical climates [24], [25], yet zero systematic long-term monitoring studies quantify these effects across deployment scenarios. Second, economic factors may render the technology uncompetitive: systematic cost-benefit analyses comparing shallow hydrokinetic systems with rapidly declining solar PV and battery storage costs are conspicuously absent from reviewed literature [11], [26], [27]. The few techno-economic assessments available suggest levelized costs of electricity (LCOE) in the range 0.15 – 0.30 USD/kWh for hydrokinetic systems [25], [28], substantially higher than utility-scale solar PV (0.03–0.05 USD/kWh) or distributed solar-plus-storage (0.08 – 0.12 USD/kWh) in 2025 [11], [16], [29]. One notable exception documents LCOE as low as 0.018 – 0.021 USD/kWh for a specific floating water wheel design [12], though this estimate lacks independent validation and appears optimistic given 95% capacity factor claims unsupported by field data. This cost disadvantage may explain deployment stagnation despite technical maturity claims. Third, and most controversially, methodological optimism in research publications – performance overprediction through inappropriate blockage corrections [14], [30], neglect of free-surface effects [21], [31], or extrapolation from idealized laboratory conditions to complex field environments [13], [32] – may systematically inflate viability expectations. This hypothesis finds support in the observation that claimed laboratory efficiencies (C_p = 0.35–0.45) rarely translate to documented field performance (C_p = 0.18–0.28), suggesting a reproducibility crisis analogous to that documented in other engineering disciplines [33], [34], [35]. Figure 1. Deployment Gap Visualization. First Global renewable energy installed capacity comparison (2015 – 2023, log scale); note 10,000 × gap between established renewables and hydrokinetic capacity. Second Research publication growth versus commercial deployment for hydrokinetic technology (1980 – 2025); cumulative publications exceed 400 while installed capacity plateaus below 80 MW [11].

Figure 1.

Deployment gap visualization

Shallow-water environments – operationally defined in this review as water depths less than 3 meters or depth-to-diameter ratios (h/D) below 5 – present design challenges fundamentally distinct from deep-water marine and tidal applications dominating hydrokinetic literature [4], [19]. These constraints arise from proximity to the free surface, limited vertical clearance for rotor placement, and the influence of bottom and sidewall boundaries on flow dynamics [30]. However, the significance and even existence of some purported 'shallow-water effects' remain contested in literature, with contradictory claims emerging from different research groups [13], [14].

Shallow-water hydrokinetic literature exhibits chaotic nomenclature that impedes systematic comparison and technology transfer. The same physical configuration may be termed 'cross-flow turbine' [45], [46], 'transverse turbine' [10], 'horizontal-axis cross-flow turbine' [47], or 'waterfall turbine' [48] – terminology clarifying axis orientation but obscuring fundamental differences in energy extraction mechanism. Floating water wheels face worse taxonomic confusion, classified variously as 'undershot wheels' [22], 'stream wheels' [7], 'rotating hydrostatic pressure machines' [49], and simply 'water wheels' [8] – terms implying distinct operating principles without consistent definitions. This nomenclature chaos creates three practical problems undermining research synthesis and technology deployment. First, incomplete literature coverage: database searches using 'cross-flow turbine' terminology miss studies employing 'transverse turbine' nomenclature despite describing identical technology; a supplementary forward-citation search identified eight relevant studies absent from primary Boolean searches. Second, invalid performance comparisons: reviews comparing 'water wheel efficiency' across studies unknowingly aggregate kinetic-mode power coefficients (C_p = 0.12 – 0.20) with head-augmented hydraulic efficiencies (η = 0.39 – 0.69), inflating apparent performance variation and obscuring actual technology differences (Quaranta, 2018 Table 5 commits this error). Third, reinvention of existing technologies: the 'rotating hydrostatic pressure machine' described by Koondhar et al. (2024) is operationally identical to high-blockage water wheels studied by Paudel et al. (2013), yet mutual citation is absent, suggesting independent rediscovery rather than knowledge accumulation.

This review proposes mechanism-based classification distinguishing technologies by energy extraction principle rather than geometric descriptors: (1) Cross-Flow Impulse employs momentum transfer via two-stage blade passage; (2) Cross-Flow Lift generates torque through hydrodynamic lift on rotating hydrofoils; (3) Floating Water Wheel (FWW) Kinetic operates via drag-dominated momentum transfer in free-flowing channels (BR < 0.40); and (4) FWW Rotating Hydrostatic Pressure Machine (RHPM) extracts energy from hydrostatic pressure differential in confined channels (BR > 0.40). Critically, FWW Kinetic and FWW RHPM employ identical hardware but operate in fundamentally different regimes, requiring distinct performance metrics (C_p versus η respectively) in Figure 1 and Figure 2. Literature treating these as a single technology type confounds performance comparison by mixing incompatible metrics – a methodological error perpetuated by eight of twelve reviewed FWW studies. Figure 3. Technology Taxonomy for Shallow-Water Hydrokinetic Turbines. Note: BR = 0.40 threshold for kinetic/RHPM mode transition is based on projected frontal area (D × L). For partially submerged FWW (h_i/D < 0.6), physical blockage may be 2–4× lower than calculated BR; field validation via head differential measurement (ΔH) is recommended for mode classification (see Section 4.2.3).

Figure 2.

Technology classification

Figure 3.

Technology taxonomy for shallow-water hydrokinetic turbines

Hydrokinetic energy conversion has been reviewed extensively, with at least twelve major reviews published between 2009 and 2024 accumulating over 5,000 collective citations [3], [4], [22], [50]. However, these reviews exhibit systematic limitations when applied to shallow-water applications (h < 3 m or h/D < 5), limiting their utility for canal-based and river hydrokinetic system design.

Given the deployment paradox outlined above, this systematic review addresses four interrelated research questions through critical evidence assessment:

RQ1: What is the strength of evidence supporting specific design parameter recommendations (blade count, tip-speed ratio, solidity, etc.) for shallow-water installations? Which parameters achieve multi-method validation versus single-study origin?

RQ2: How do laboratory-predicted performance metrics translate to field-validated operational data? What systematic degradation factors (biofouling, debris, seasonal variation, array effects) contribute to the documented 25 – 35% performance gap?

RQ3: Which methodological controversies (blockage correction methods, free-surface modeling approaches, ventilation prediction) remain unresolved, and what implications do these have for design guidance reliability?

RQ4: What explains the deployment stagnation despite claimed technical maturity? Do economic factors, unresolved technical barriers, or methodological optimism represent the primary limitation?

To address these questions, PRISMA 2020 systematic review guidelines are applied with dual independent screening (κ = 0.83), evidence quality hierarchies are established distinguishing strong from weak validation, laboratory-to-field performance gaps are quantified, and competing methodological approaches are critically assessed. The review emphasizes deployment readiness assessment over incremental parameter optimization, recognizing that additional laboratory studies of blade profile refinement offer diminishing returns compared to field validation campaigns, techno-economic assessments, and multi-year performance monitoring of actual installations.

This systematic review follows PRISMA 2020 guidelines [51] for transparent and reproducible evidence synthesis. The review protocol was not prospectively registered in PROSPERO due to engineering-specific focus and rapid evolution of shallow-water hydrokinetic literature, consistent with practices in energy technology reviews [52], [53]. However, all methodological decisions are documented a priori to minimize bias and enhance replicability [54]. The review addresses four primary research questions (stated in Section 1.5) focusing on evidence quality assessment for design parameters, laboratory-to-field performance translation, methodological controversies, and deployment barriers. Unlike conventional descriptive reviews, this critical systematic review explicitly evaluates evidence strength using five-domain quality assessment adapted from American Institute of Aeronautics and Astronautics (AIAA) CFD validation guidelines [32], [33] and hybrid energy system assessment frameworks [17].

Comprehensive literature searches were conducted across four bibliographic databases: Scopus, Web of Science Core Collection, IEEE Xplore, and Google Scholar (limited to first 200 results) between January 2015 to December 2025. Database selection follows established bibliometric practices for engineering systematic reviews [3], [4], [49]. Scopus provides extensive engineering coverage with reliable citation metrics [55], Web of Science ensures quality filtering through impact factor criteria, IEEE Xplore captures conference proceedings often missed by traditional databases, and Google Scholar supplements with grey literature and non-indexed sources in Table 1 [56].

The search strategy employed three complementary approaches to maximize sensitivity while maintaining precision [57], [58]. First, Boolean keyword searches combined technology terms.

Search strings were adapted for each database's syntax requirements following Cochrane Handbook recommendations [59]. Second, forward and backward citation tracking from 12 highly-cited seed papers (>100 citations each) identified studies missed by keyword searches, addressing nomenclature chaos discussed in Section 1.3 [60]. Third, hand-searching of specialized journals (Renewable Energy, Energy for Sustainable Development, Journal of Hydraulic Research) captured recent publications not yet indexed [61]. No language restrictions were applied initially, though non-English studies required translation for eligibility assessment. Publication date limits spanned January 2015 to December 2025, capturing modern computational and experimental capabilities while focusing on deployment-relevant research [6]. Earlier foundational works (pre-2015) were incorporated through citation tracking if directly relevant to shallow-water applications.

Studies were included if they met all of the following PICOS criteria (Population, Intervention, Comparator, Outcome, Study design) adapted for engineering reviews [54], [62] show in Table 2.

1. Population/Application: Shallow-water installations defined as h < 3 m absolute depth OR h/D < 5 depth-to-diameter ratio OR Froude number Fr > 0.15, distinguishing these from deep-water marine/tidal systems where free-surface effects are negligible [13], [14].

2. Intervention/Technology: Cross-Flow Impulse (Banki-type), Cross-Flow Lift (Darrieus/Gorlov/H-rotor), or Floating Water Wheel (undershot/breastshot/RHPM) turbines. Excluded: Axial-flow propeller turbines (fundamentally different physics; covered extensively elsewhere [50]), oscillating hydrofoils, vortex-induced vibration devices, and Archimedes screws (different operating principles).

3. Comparator: Not required (non-comparative review), though studies comparing different configurations or technologies were preferentially included for evidence synthesis [59].

4. Outcomes: Primary outcomes included power coefficient (C_p), hydraulic efficiency (η), design parameter relationships (blade count, tip-speed ratio, solidity), and field-validated performance data. Secondary outcomes included manufacturing feasibility, economic metrics (LCOE, capital cost), and environmental impacts. Studies reporting only theoretical potential estimates without performance data were excluded [10].

5. Study Design: Experimental (laboratory flume, field installation), computational (CFD with validation), or hybrid (combined experimental-numerical) studies. Excluded: Pure theoretical analyses without validation, patent descriptions without performance testing, conference abstracts without full-text availability, and review articles (though their references were tracked).

Critical justification for the h/D < 5 threshold: This boundary emerges from free-surface modeling studies showing >15% C_p overprediction by single-phase models when h/D drops below this value [21], [37]. The h < 3 m absolute criterion captures irrigation canals (typical depth 0.5 – 2.5 m) and small rivers applicable for distributed generation [7]. These thresholds intentionally exclude tidal and marine systems (h > 10 m, h/D > 20) where blockage, free-surface, and ventilation considerations differ fundamentally. These methodological considerations inform the systematic selection criteria applied across all candidate studies. Table 2 presents the complete inclusion and exclusion criteria, operationalizing the h/D < 5 threshold and technology scope definitions into actionable screening parameters. Seven criterion categories – water depth, application context, technology type, performance data, design details, blockage/depth reporting, and publication type – ensure comprehensive yet focused study selection aligned with shallow-water deployment realities.

Screening followed a two-stage approach: title/abstract review followed by full-text assessment. Dual independent screening was conducted using Rayyan QCRI, achieving inter-rater reliability of κ = 0.83 (substantial agreement). Discrepancies were resolved through consensus discussion [66]. The complete screening flow is presented in Figure 4 (PRISMA diagram) and summarized in Table 3.

Figure 4.

PRISMA 2020 Flow diagram

Notes: Flow diagram adapted from Page et al. (2021). Screening conducted using Rayyan QCRI with dual independent reviewers (κ = 0.83). *Legacy studies (n = 9) used as background physics only; not included in primary corpus or pooled statistics.

Evidence quality was assessed using a five-domain scoring rubric adapted from AIAA CFD validation guidelines [32], [33] and systematic review quality tools [67], [68]. This multi-domain approach recognizes that engineering research quality depends on methodological rigor, validation adequacy, and reporting transparency rather than traditional risk-of-bias metrics designed for clinical trials [54].

Domain 1: Experimental Methodology (for EFD studies): Scored 0-2 based on facility characterization (Reynolds number, turbulence intensity, boundary layer profiles), measurement uncertainty quantification following Guide to the Expression of Uncertainty in Measurement (GUM) guidelines [69], and repeatability demonstration (≥3 runs per condition). High-quality studies report full uncertainty budgets with 95% confidence intervals and document all systematic error sources [70].

Domain 2: Computational Methodology (for CFD studies): Scored 0-2 based on mesh independence demonstration using Richardson extrapolation or Grid Convergence Index [32], turbulence model validation against experimental data, and iterative convergence criteria (residuals <10^-4 or integrated quantities varying <1%). Free-surface modeling studies required VOF validation against measured surface elevations, not just C_p comparison [41], [71].

Domain 3: Validation Adequacy: Scored 0-3 distinguishing between validation strength levels. Score 3 (strong): Multi-method validation with independent datasets (e.g., Particle Image Velocimetry (PIV) flow field + force measurements + field installation data). Score 2 (moderate): Single-method experimental validation with reasonable agreement (<15% discrepancy for integral quantities). Score 1 (weak): Qualitative comparison only or validation against single published datapoint. Score 0 (absent): No validation provided [33].

Domain 4: Reporting Transparency: Scored 0-2 based on data accessibility (raw data availability in repositories), geometric specification completeness (CAD files or dimensioned drawings), and reproducibility potential (sufficient detail for replication without author contact). Studies scoring 2 provide supplementary data files or GitHub repositories; those scoring 0 omit critical parameters preventing reproduction [72], [73].

Domain 5: Field Relevance: Scored 0-1 based on Reynolds number matching field conditions (Re > 10⁵), debris tolerance assessment, seasonal variation consideration, or actual field deployment data. Laboratory studies at model scale (Re < 10⁴) without scale-effect discussion scored 0; field installations automatically scored 1 [4], [10].

Total scores (0-10) classified studies into evidence tiers: Strong (8-10), Moderate (5-7), Weak (3-4), Very Weak (0-2). Classification thresholds were established through pilot scoring of 10 studies with all authors achieving consensus, following GRADE working group recommendations adapted for engineering contexts [74], [75].

Structured data extraction employed standardized forms capturing 47 variables across five categories: bibliographic information, system specifications (turbine type, diameter, blade count), operating conditions (velocity, depth, blockage ratio), performance metrics (C_p, η, optimal λ), and study characteristics (experimental vs CFD, validation method). Extraction was performed independently by two reviewers (KR, YT) using Covidence systematic review software, with discrepancies resolved through discussion [76]. Critical methodological challenge: Performance metric harmonization across inconsistent reporting. Power coefficient C_p calculations require reference area specification, yet 18 of 38 studies (47%) failed to explicitly state whether projected (D×L), swept (πDL/4 for circular cross-sections), or physically submerged (h_i×L) area was used – areas differing by factors of 2–4× for partially submerged turbines [22]. Were standardized all C_p values to projected frontal area (D×L) following majority practice (32 of 38 studies, 84%), recalculating when sufficient raw data permitted. Studies providing insufficient information for recalculation were flagged with uncertainty notation. Blockage ratio definitions also required harmonization: 23 studies (61%) defined BR using projected area, 9 (24%) used physically submerged area, and 6 (16%) were ambiguous. Dual metrics were calculated where possible: BR_projected = (D×L)/(b×h) and BR_actual = (h_i×L)/(b×h), noting that the BR = 0.40 kinetic/RHPM threshold established in literature uses projected area convention. This definitional ambiguity and its implications for performance classification are critically analyzed in Section 4.2.3.

Quantitative meta-analysis was not performed due to high methodological heterogeneity in experimental conditions, turbine geometries, and performance definitions – factors precluding meaningful statistical pooling [59], [77]. Instead, narrative synthesis with tabulated evidence organized results by technology type, identifying convergent findings (consistent across ≥5 studies), divergent findings (contradictory results), and absent evidence (critical knowledge gaps). Critical analysis framework examined three validity dimensions adapted from Campbell & Stanley (1963) and Shadish et al. (2002): (1) Internal validity – Do study conclusions follow logically from methods and data? Assessed through consistency checks between reported uncertainties and claimed parameter sensitivities. (2) External validity – Do laboratory results generalize to field conditions? Evaluated by comparing laboratory-scale Reynolds numbers against field installations and documenting scale-effect discussions. (3) Construct validity – Do measured variables actually represent intended constructs? Examined through scrutiny of performance metric definitions, blockage correction applications, and free-surface modeling assumptions [78], [79]. Methodological controversies (blockage corrections, free-surface modeling, ventilation prediction) received dedicated analysis identifying competing approaches, quantifying prediction variation across methods, and assessing empirical support for each. This critical lens distinguishes the current review from prior descriptive syntheses, explicitly evaluating claim strength rather than uncritically aggregating reported results [35], [53].

Sensitivity analyses examined how conclusions changed under alternative methodological decisions [80]. First, varying quality threshold: Restricting to strong-evidence studies only (score ≥8) reduced sample to n = 12 (32%) but did not materially alter primary findings on optimal design parameters, suggesting conclusions are robust. Performance ranges narrowed slightly (C_p,max = 0.28 – 0.32 vs 0.25 – 0.35 for full sample) but general trends persisted. Second, excluding studies with ambiguous metric definitions (n = 18) primarily affected FWW performance ranges, expanding uncertainty intervals from ±8% to ±15% but not changing median values substantially. This highlights reporting quality as limiting factor in evidence synthesis [81]. Third, alternative blockage ratio thresholds: Testing BR boundaries of 0.30 – 0.50 instead of canonical 0.40 showed kinetic/RHPM classification remained stable for 32 of 38 studies (84%), with only partially submerged FWW exhibiting sensitivity. This analysis motivated detailed BR definition scrutiny in Section 4.2.3 and applicability boundary warnings in Section 5.1.

Several methodological limitations warrant acknowledgment [82]. First, publication bias likely inflates reported performance values, as negative results (e.g., ventilation-induced failures, debris-compromised installations) remain underreported. Funnel plot analysis was inappropriate given non-comparative nature, but an explicit search was conducted for field installation failures through grey literature and conference proceedings – finding only 3 documented negative outcomes versus 12 positive [83]. Second, language bias: While no language restrictions were applied, only 4 non-English studies (2 Chinese, 2 Spanish) met inclusion criteria after translation. Potential exclusion of relevant Asian or South American research may limit generalizability, though English-language publication dominance in engineering reduces this concern [84]. Third, grey literature coverage: Conference proceedings and technical reports were included when accessible, but proprietary industrial testing data remain unavailable. This gap particularly affects deployment readiness assessment, as commercial developers rarely publish detailed field performance or economic data [85]. Fourth, rapid literature evolution: The January 2015 to December 2025 window may miss relevant earlier studies, though backward citation tracking captured seminal works. Conversely, the 3-month lag between final search (March 2024) and manuscript submission means recent publications may be absent [86]. Mitigation strategies included: (1) Dual independent screening with high inter-rater reliability (κ = 0.83) minimizing selection bias. (2) Transparent documentation of all methodological decisions enabling replication. (3) Contact with 8 corresponding authors clarifying ambiguous methods or requesting supplementary data (5 responded). (4) Sensitivity analyses demonstrating conclusion robustness to alternative decisions. (5) Explicit acknowledgment of evidence limitations in synthesis sections rather than overstating certainty [87].

Cross-Flow Impulse turbines, commonly known as Banki or Michell turbines, extract energy through two-stage momentum transfer as flow passes radially through curved blade passages. Unlike reaction turbines, the flow undergoes atmospheric pressure throughout the rotor, simplifying mechanical design but limiting maximum efficiency [45], [88]. Shallow-water applications typically employ undershot configurations where the turbine intercepts surface flow without upstream impoundment, distinguishing these from traditional penstock-fed Banki installations [46], [47].

Cross-Flow Lift turbines extract energy through hydrodynamic lift on rotating hydrofoil blades, analogous to vertical-axis wind turbines but adapted for water flow [4]. Configurations include straight-bladed Darrieus (H-rotor), helical Gorlov, and variable-pitch cycloidal types. Shallow-water applications face critical ventilation constraints when blade trajectories approach the free surface [9].

Floating water wheels represent the most successfully deployed shallow-water technology, with documented field installations totaling 2–3 MW cumulative capacity worldwide [7], [8], [22]. These systems employ undershot or breastshot configurations where water impacts paddle-like blades, extracting energy through combined drag and impulse mechanisms [23]. Critical distinction: FWW operate in two fundamentally different regimes depending on channel blockage, requiring separate performance metrics and design approaches.

Field-validated performance consistently underperforms laboratory predictions across all technology families, with degradation factors ranging from 15 – 45% [1], [7], [8]. This systematic gap undermines deployment viability assessments and highlights critical need for field validation campaigns.

Derating Factor Recommendations.

Conservative design practice requires applying derating factors to laboratory-validated C_p or η values:

• Optimistic scenario (clean water, frequent maintenance, low debris): Derate by 15 – 20%

• Realistic scenario (seasonal variation, moderate debris, bi-annual maintenance): Derate by 25 – 35%

• Conservative scenario (tropical climate, high debris, minimal maintenance): Derate by 40 – 50%

Example: Cross-Flow Impulse with C_p,lab = 0.32 (strong evidence, quality score 8 – 9) should be designed assuming C_p,field = 0.24 for realistic scenario (25% derate) or C_p,field = 0.19 for conservative scenario (40% derate). Economic viability must account for reduced capacity factor.

The systematic performance gap between laboratory predictions and field measurements is visualized in Figure 5, which compares C_p values across all three technology families under controlled (laboratory) versus deployed (field) conditions. This comparison reveals consistent 15 – 45% performance degradation, with Cross-Flow Impulse turbines showing 25 – 35% reduction, FWW kinetic mode 15 – 30% reduction, and Cross-Flow Lift lacking field validation entirely. The shaded degradation bands represent the derating factor ranges recommended for design practice, translating optimistic laboratory results into realistic deployment expectations.

Figure 5.

Laboratory versus field performance comparison

Synthesis across technology families reveals distinct trade-offs in performance, complexity, and deployment readiness:

Highest Performance (Laboratory): Cross-Flow Lift turbines demonstrate superior C_p,max = 0.35 – 0.49 under idealized conditions but ventilation constraints, Reynolds sensitivity, and zero shallow-water field validations disqualify these for h/D < 2.0 applications.

Best Field-Validated: Floating water wheels (RHPM mode) achieve η = 0.55 – 0.69 with documented multi-year operation [23], [112]. However, infrastructure requirements (confined channels, BR > 0.45) and regulatory considerations limit deployment flexibility.

Most Deployable: Floating water wheels (kinetic mode) and Cross-Flow Impulse turbines offer deployment flexibility with BR < 0.40 requiring minimal channel modifications. Field-validated C_p,field = 0.18 – 0.28 represents acceptable performance-versus-complexity trade-off for distributed generation (<50 kW) [7], [8], [92].

Critical research gap: Economic viability assessments comparing shallow-water hydrokinetic LCOE (0.15 – 0.30 USD/kWh estimated) with declining solar PV costs (0.03 – 0.05 USD/kWh) are conspicuously absent from reviewed literature despite being deployment-critical information.

Channel confinement accelerates flow through turbine swept areas, artificially inflating measured C_p values compared to free-stream operation. Correcting for this blockage effect requires estimating the velocity increase – yet eight competing methodologies identified in reviewed literature yield prediction variations of ±15 – 25% for identical geometries [113]. This methodological chaos prevents reliable performance comparison across studies and undermines design guidance transferability.

Blockage Ratio Definition Ambiguity.

Beyond correction method selection, fundamental ambiguity exists in BR definition itself, particularly for partially submerged turbines. Literature convention defines BR using projected frontal area (D×L), yet physical flow blockage depends on actual submerged area – quantities differing by factors of 2–4× for floating water wheels with immersion ratios h_i/D = 0.20 – 0.40 [22], [23], [107].

Example: FWW with D = 2.0 m, L = 3.0 m, h_i = 0.5 m, channel dimensions h = 1.5 m, W = 5.0 m yields BR_projected = (2.0×3.0)/(1.5×5.0) = 0.80 suggesting RHPM operation (BR > 0.40), but BR_physical = (0.5×3.0)/(1.5×5.0) = 0.20 indicating kinetic free-stream conditions. This 4× discrepancy directly affects performance metric selection (C_p vs η), blockage correction application, and technology classification.

Only 2 of 12 FWW studies (17%) validate operating mode through field measurement of head differential ΔH rather than relying solely on calculated BR [23], [112]. This evidence gap prevents definitive resolution of which BR definition better predicts the kinetic-to-RHPM transition threshold, leaving practitioners without reliable classification criteria for shallow-water installations. The methodological inconsistencies in blockage correction are illustrated in Figure 6, which compares velocity correction factors from seven methods across BR = 0.10 – 0.50, revealing ±18% variation at typical shallow-water blockage ratios.

Figure 6.

Blockage correction method comparison

Volume-of-Fluid (VOF) methods have become standard for simulating shallow-water turbines, appearing in 18 of 38 reviewed studies (47%). Yet systematic validation against measured surface elevation profiles – the fundamental output distinguishing VOF from single-phase models – occurs in only 4 studies (22% of VOF implementations) [21], [38], [117], [118], [119]. This validation deficit creates circular reasoning: VOF models calibrated against integrated quantities (C_p, C_T) cannot confirm if free-surface effects are correctly captured or merely compensated by other model parameters.

Ventilation – air entrainment from free surface into blade passages causing catastrophic torque loss – represents the most severe shallow-water failure mode for Cross-Flow Lift turbines, yet prediction capability remains entirely empirical [9], [43], [102]. No physics-based models validated against controlled experiments exist despite this phenomenon's deployment-critical importance.

Confusion between kinetic power coefficient C_p = P/(0.5ρAV³) and hydraulic efficiency η = P/(ρgQΔH) pervades shallow-water hydrokinetic literature, undermining comparative assessment and technology selection. These metrics differ fundamentally in physical meaning, numerical magnitude (typically 2–3×), and applicability conditions – yet 8 of 38 reviewed studies (21%) compare values directly or use terms interchangeably [3], [22].

The methodological controversies documented above share common deficiencies: (1) Limited experimental validation datasets constraining model calibration, (2) Single-study origins for widely-cited thresholds lacking independent verification, (3) Technology-specific correlations erroneously generalized across turbine types, (4) Inadequate reporting transparency preventing method reproduction. Addressing these requires systematic experimental campaigns prioritizing.

1. Surface Elevation Validation Datasets: Comprehensive free-surface profile measurements (≥20 spatial locations, multiple Fr, BR, turbine types) enabling VOF model systematic validation and Fr-threshold verification.

2. Blockage Correction Verification: Controlled experiments measuring upstream velocity profiles with and without turbine across BR = 0.10 – 0.50, enabling correction method accuracy assessment and uncertainty quantification.

3. Ventilation Onset Mapping: Systematic reduction of submergence ratio c_top/c at fixed λ, solidity, documenting performance degradation curves and identifying transition mechanisms (vortex formation vs tip cavitation) through high-speed visualization.

4. Head Differential Field Surveys: Instrumentation of existing shallow-water installations with upstream/downstream pressure monitoring validating BR thresholds and metric selection criteria under real operating conditions including seasonal variation.

Until such systematic validation campaigns materialize, designers must acknowledge that shallow-water hydrokinetic performance predictions carry uncertainties of ±20 – 40% arising from methodological ambiguities – substantially larger than typical measurement uncertainties (±5 – 10%) and potentially dominating economic viability assessments. The cumulative impact of methodological uncertainties is quantified in Figure 7, demonstrating that total methodological uncertainty (46%) exceeds measurement uncertainty by 6.1×, explaining why laboratory Cp predictions overestimate field performance by 1.3–2.0×.

Figure 7.

Cumulative uncertainty waterfall in hydrokinetic turbine performance prediction

Cumulative uncertainty waterfall showing how methodological uncertainties accumulate in CFD-based performance predictions for hydrokinetic turbines. Starting from measurement uncertainty (±7.5%, blue baseline), successive methodological uncertainties are added: blockage correction (±15%, orange), free-surface modeling (±8%, green), turbulence modeling (±5%, red), mesh resolution (±3%, purple), and field derating factor (15%, brown). Total methodological uncertainty (46%) exceeds measurement uncertainty by 6.1×, explaining why laboratory C_p predictions overestimate field performance by 1.3–2.0× (Section 3.4). Blue shaded region indicates typical measurement uncertainty range (±5-10%). This analysis demonstrates that improving experimental accuracy provides diminishing returns compared to resolving methodological controversies in blockage correction and free-surface modeling.

Section 3 documented systematic 15 – 45% performance degradation between laboratory testing and field operation across all technology families. Conservative design practice requires incorporating this degradation explicitly rather than applying laboratory C_p values directly to deployment planning. Table 7 synthesizes field-adjusted ranges by technology, incorporating derating factors, blockage correction uncertainties, and deployment complexity.

Critical design implication: Power predictions using laboratory C_p,max without field adjustment overestimate energy yield by 1.3 – 2.0×, directly undermining economic viability assessments and deployment planning. Conservative approach applies lower bound of field-adjusted range plus additional 10 – 15% safety margin for site-specific uncertainties (flow measurement accuracy, debris loading, seasonal variation).

Technology selection for shallow-water applications requires balancing performance potential against deployment complexity, validation status, and risk tolerance. The following present decision framework incorporating evidence strength, field validation, and deployment barriers.

Despite 40+ years of shallow-water hydrokinetic research and documented laboratory performance (Section 3), global installed capacity remains <100 MW with limited commercial traction [3], [4], [133]. This deployment gap reflects systematic underestimation of non-technical barriers dominating real-world implementation.

Bridging the deployment gap requires acknowledging that technology maturation follows established progression: laboratory validation → pilot demonstration → field deployment → commercial scale-up. Shallow-water hydrokinetics currently stalls between pilot and field deployment phases.

This systematic review applied rigorous evidence quality assessment (Section 2.5) to 38 studies. Table 8 summarizes findings strength by topic area, highlighting where robust evidence supports design guidance versus where critical knowledge gaps persist.

Priority research needs emerging from this assessment: (1) Comprehensive techno-economic analysis comparing hydrokinetic LCOE against competing renewables across deployment contexts; (2) Long-term (10+ year) field monitoring documenting actual versus predicted performance, O&M costs, failure modes; (3) Systematic validation of methodological controversies identified Section 4 – blockage corrections, VOF accuracy, ventilation prediction; (4) Standardization efforts developing testing protocols and certification frameworks.

This systematic review examined shallow-water hydrokinetic turbines (h < 3 m) through critical synthesis of 38 studies published 2000 – 2025, providing the first comprehensive assessment integrating performance validation, methodological controversies, and deployment realities. Three technology families emerged with distinct trade-offs: Cross-Flow Impulse turbines achieve field-validated C_p,field = 0.18 – 0.28 with moderate deployment complexity; Floating Water Wheels (FWW) demonstrate robust field operation at C_p,field = 0.12 – 0.18 (kinetic mode) or η = 0.45 – 0.60 (RHPM mode) with lowest installation barriers; Cross-Flow Lift turbines show laboratory potential C_p,lab = 0.35 – 0.49 but lack shallow-water field validation and face ventilation constraints at h/D < 2.0 (Section 3).

Critical methodological analysis revealed that performance predictions carry ±20 – 40% uncertainty from competing blockage correction methods (±18% variation at BR = 0.30), unvalidated VOF free-surface modeling (only 22% of implementations validate surface profiles), and absent ventilation prediction frameworks (Section 4). This uncertainty substantially exceeds typical measurement errors (±5 – 10%) and directly impacts economic viability assessments. Furthermore, systematic 15 – 45% performance degradation between laboratory testing and field deployment – documented across all technologies – demands explicit derating factors in design practice: 25 – 35% for Cross-Flow Impulse, 15 – 30% for FWW kinetic mode (Section 5.1).

Deployment gap analysis identified that despite 40+ years of research, global installed capacity remains <100 MW due to convergent economic, regulatory, and technical maturity barriers. Field capital costs (5,000 – 12,000 USD/kW) exceed literature estimates by 2–3×, yielding LCOE 0.15 – 0.30 USD/kWh facing severe competition from solar PV (0.03 – 0.05 USD/kWh) and onshore wind (0.04 – 0.06 USD/kWh). Regulatory approval timelines (12 – 36 months typical) and absence of international standards create financing challenges incompatible with commercial deployment scale (Section 5.3).

Based on evidence quality assessment and field validation status, this review provides differentiated recommendations by technology and application context:

Floating Water Wheels (Kinetic Mode, BR < 0.40): Recommended for distributed off-grid applications (1 – 10 kW) where simplicity and portability outweigh modest efficiency. Strong evidence (12 studies, 7 with field validation) supports reliable operation at C_p,field = 0.12 – 0.18 in shallow flows (h = 1.0–2.0 m, V = 0.8 – 1.5 m/s). Economics favorable where baseline diesel generation >0.25 USD/kWh. Design priorities: conservative immersion ratio h_i/D = 0.20 – 0.30, blade count N_b = 8 – 12 (conservative), hybrid integration with solar PV for diurnal diversity. Critical gap: long-term reliability beyond 5 years undocumented (Section 5.2.1).

Cross-Flow Impulse Turbines: Recommended for community-scale installations (10 – 100 kW) where higher efficiency (C_p,field = 0.18 – 0.28) justifies moderate deployment complexity. Strong evidence (18 studies) validates design parameters: blade count N_b = 15 – 26, diameter ratio d_i/D = 0.66 – 0.68, blade angles β₁ = 30 – 35°, β₂ = 90 – 105°. However, free-surface effects at Fr > 0.25 remain poorly validated, requiring site-specific flow modeling. Fixed mounting necessitates civil works and extended permitting (12 – 24 months). Economically competitive only against micro-hydro alternatives where head development infeasible (Section 3.1, 5.2.1).

Cross-Flow Lift Turbines (Darrieus/Gorlov): Not recommended for shallow-water deployment (h/D < 2.0) given zero field validation, unpredictable ventilation onset, and Reynolds sensitivity degrading performance 15 – 40% at typical shallow-water velocities (V = 0.8 – 1.5 m/s, blade chord Re_c = 0.5 – 3 × 10⁵). Conditional consideration only for deep-water applications (h/D ≥ 2.5) with verified top clearance c_top ≥ 1.0 c maintained seasonally, requiring pilot demonstration with performance monitoring before commercial scale-up. Laboratory performance advantages (C_p,lab = 0.35 – 0.49) do not justify deployment risk absent validated shallow-water operation protocols (Section 3.2, 5.2.1).

Competitive Positioning: Honest assessment positions shallow-water hydrokinetics as niche technology serving specialized contexts where (1) off-grid locations with diesel baseline >0.25 USD/kWh, (2) hybrid systems valuing baseload stability at premium pricing, or (3) environmental constraints prohibit dam construction. Mainstream renewable energy deployment will continue dominated by solar PV, wind, and conventional hydropower given their superior economics (2 – 6 × lower LCOE), faster deployment timelines (1 – 6 months vs 12 – 36 months), and established supply chains. Without technological breakthrough reducing LCOE below 0.10 USD/kWh – requiring 40 – 60% cost reduction from current field reality – global installed capacity likely plateaus at hundreds of megawatts rather than gigawatt scale (Section 5.4.3).

Bridging the gap between laboratory success and commercial deployment requires systematic research addressing four priority areas:

Priority 1: Long-Term Field Validation Studies (10 – 20 Years): Current maximum documented continuous operation spans only 3 – 5 years [8], [23], insufficient to validate claimed 15 – 20 year design lifetimes or establish actuarial databases enabling insurance products and bankable performance guarantees. Coordinated research programs should instrument 20 – 30 installations across diverse contexts (flow regimes, debris loading, climatic conditions) monitoring:

• Actual versus predicted capacity factors quantifying prediction accuracy and seasonal variation impacts

• Component failure modes, rates, and costs distinguishing design deficiencies from inherent operational challenges

• Maintenance requirements and schedules optimizing reliability versus operational costs

• Environmental impacts including fish interaction, sediment transport alteration, riparian habitat effects

International cooperation through organizations like IEA Hydropower or IRENA could coordinate data collection protocols ensuring cross-study comparability and accelerating evidence accumulation. Estimated program cost 5–10 million USD over 15 years represents modest investment potentially unlocking billion-dollar markets if economic viability confirmed [11], [135].

Priority 2: Comprehensive Techno-Economic Analysis: Only 3 of 38 reviewed studies conducted economic analysis, and none provided comprehensive LCOE sensitivity assessment across deployment contexts. Critical knowledge gaps include:

• Site-specific LCOE modeling incorporating flow regime statistics, civil works requirements, grid connection distances, local labor costs, permitting timelines – enabling developers to rapidly assess project viability without costly feasibility studies

• Competitive benchmarking against solar PV, wind, diesel, and micro-hydro alternatives across power scales (1 kW to 1 MW) and deployment contexts (grid-connected, off-grid, hybrid) identifying competitive niches

• Learning curve analysis quantifying cost reduction potential through volume manufacturing, standardization, and supply chain maturation versus inherent technological constraints

• Market size estimation for viable deployment niches informing R&D investment decisions and industry development strategies

Open-access tools and databases (similar to NREL's System Advisor Model for solar/wind) would democratize techno-economic analysis enabling small developers and community organizations to evaluate projects without expensive consultants [136], [137].

Priority 3: Methodological Controversy Resolution: Section 4 identified three critical controversies undermining design reliability. Systematic experimental campaigns should address:

Blockage correction validation: Controlled flume experiments measuring upstream velocity profiles with/without turbines across BR = 0.10–0.50, Fr = 0.15 – 0.60, aspect ratios L/D = 0.5 – 3.0. Seven competing correction methods currently yield ±18% prediction variation [113]; systematic validation could reduce uncertainty to ±5 – 8% enabling confident multi-turbine array design.

Free-surface modeling verification: Comprehensive surface elevation profile measurements (≥20 spatial locations, 5D upstream to 10D downstream, temporal resolution ≥10× blade passage frequency) enabling VOF model validation. Currently only 22% of VOF implementations validate surface predictions; robust validation datasets would establish when simplified single-phase models suffice versus requiring computationally expensive multiphase approaches.

Ventilation prediction frameworks: Systematic submergence ratio reduction experiments (c_top/c = 2.0 to 0.2) at fixed tip-speed ratios documenting performance degradation curves and identifying transition mechanisms through high-speed visualization. Current empirical correlations provide only binary onset/no-onset indicators; quantitative degradation prediction enables rational design trade-offs balancing clearance requirements versus installation constraints.

Priority 4: Standardization and Certification Development: Industry collaboration through technical committees (e.g., IEC TC 114 Marine Energy, extending scope to river applications) should develop:

• Power performance testing protocols specifying flow measurement requirements, blockage correction procedures, reporting standards enabling verifiable capacity ratings

• Design standards addressing structural integrity, safety factors for flood loading, corrosion protection, electrical safety in wet environments

• Certification frameworks providing third-party validation of manufacturer claims and design compliance reducing insurance costs and enabling bankable projects

Wind and solar photovoltaic industries demonstrate that standardization accelerates deployment by reducing transaction costs, enabling supply chain economies, and providing regulatory confidence [138]. Hydrokinetic sector maturation requires similar institutional infrastructure.

Government policies can accelerate deployment by addressing non-technical barriers while maintaining appropriate environmental safeguards:

Streamlined Regulatory Frameworks: Establish clear permitting pathways distinguishing low-impact installations (BR < 0.15, single units, no channel modifications) from high-impact deployments (BR > 0.40, arrays, civil works). Low-impact projects meeting defined criteria could receive expedited approval (30 – 90 days) similar to rooftop solar PV, while high-impact projects undergo comprehensive environmental assessment. Current regulatory ambiguity – classifying hydrokinetic installations inconsistently as hydropower, water diversion, or novel technology – creates 12 – 36 month approval timelines deterring private investment [132], [134].

Demonstration Project Support: Targeted grant programs (40 – 60% capital cost sharing) for pilot installations in representative contexts (diverse flow regimes, debris conditions, climatic zones) generating long-term performance data addressing knowledge gaps identified Priority 1. Successful models include Canada's Smart Grid Demonstration Program and European Commission's Horizon Europe supporting 5 – 10 MW pilot projects de-risking technologies before commercial scale-up [139], [140]. Hydrokinetic programs targeting 50–100 MW cumulative demonstration capacity over 10 years (50 – 100 USD million investment) could validate economic viability and inform deployment strategies.

Hybrid System Integration Incentives: Hydrokinetic systems provide complementary generation profiles to solar PV (nocturnal baseload when solar unavailable) and wind (flow-driven versus weather-driven). Feed-in tariffs or renewable energy credits recognizing baseload value (e.g., 1.5–2.0× standard renewable credit for dispatchable generation) improve economic competitiveness. Germany's Renewable Energy Act and California's Self-Generation Incentive Program demonstrate how differentiated incentives can support diverse renewable portfolios rather than lowest-LCOE monocultures [141], [142].

Shallow-water hydrokinetic energy conversion represents mature laboratory technology facing persistent deployment barriers. Four decades of research have established robust design knowledge for two field-validated technologies (Cross-Flow Impulse, Floating Water Wheels) achieving respectable performance (C_p,field = 0.12 – 0.28) in appropriate contexts. However, this review's critical synthesis reveals that technical performance alone determines neither deployment success nor societal value. Economic reality (LCOE 3–10× competing renewables), regulatory complexity (12 – 36 month approvals), and technical maturity gaps (absent standardization, <5 year field validation) explain why global installed capacity remains orders of magnitude below early projections.

The path forward requires honest acknowledgment that shallow-water hydrokinetics will not revolutionize global renewable energy supply. Solar photovoltaic and wind technologies have achieved cost reductions, deployment scales, and supply chain maturation incompatible with hydrokinetic competition at mainstream level. Rather, this technology's value proposition lies in serving specialized niches where geographic constraints prohibit alternatives or where baseload generation commands premium pricing. Distributed off-grid applications represent the most promising near-term opportunity, leveraging FWW simplicity and portability for remote communities where diesel baselines exceed 0.25 USD/kWh.

Realizing even this modest potential demands systematic research addressing critical knowledge gaps – particularly long-term field validation and comprehensive techno-economic analysis – coupled with policy frameworks streamlining low-impact deployment while maintaining environmental safeguards. The research community must resist exaggerating deployment prospects or perpetuating laboratory-derived optimism disconnected from field reality. Instead, focusing efforts on contexts where hydrokinetic technologies offer genuine advantages can deliver meaningful contributions to renewable energy portfolios without misallocating scarce resources chasing unattainable gigawatt-scale dreams.

Success should be measured not by installed capacity rivaling conventional hydropower, but by delivering reliable, cost-effective energy access to communities and applications where alternatives fail – a worthwhile objective requiring evidence-based deployment strategies grounded in systematic review findings rather than aspirational projections.