Wave and Tidal Power Face a Hidden Test: Adding Up the Ocean’s Stresses

Marine renewable energy is having its moment. Nations racing toward net-zero are looking past offshore wind to the untapped power of waves, tides, ocean currents, and even gradients in salinity and temperature. But as devices move from single prototypes to commercial arrays, scientists are warning that the industry is being judged with an assessment tool it has barely learned to use: the cumulative effects assessment, or CEA. A sweeping new international review, led by researchers including Elizabeth Fulton of Australia’s CSIRO and Lenaïg Hemery and Andrea Copping of the United States’ Pacific Northwest National Laboratory, synthesizes more than two decades of literature to map where the field stands, and where it risks stumbling. The work, conducted under the International Energy Agency’s Ocean Energy Systems programme, draws on a systematic search that screened 926 documents and retained 145 relevant ones, all published since 2007.

The core problem is deceptively simple. Environmental regulators have long required developers to assess not just the direct footprint of a project but the combined burden it adds to everything else pressing on an ecosystem, from shipping noise to climate-driven warming. Yet the review finds that for marine renewable energy, effects are typically measured one by one: the underwater noise of a turbine, the electromagnetic field of an export cable, the collision risk posed to fish and marine mammals by rotating tidal blades. What is rarely asked is how these effects stack on top of each other, on top of other industries, and on top of natural variability. The authors argue that without rigorous cumulative assessment, the industry could trigger unintended consequences that undermine the very sustainability goals, from clean energy to climate action and healthier oceans, that motivate it.

The marine environment makes this accounting uniquely hard. Unlike land, the ocean is fluid and dispersive, spreading impacts far beyond the physical footprint of any installation. Many marine species shift habitats across their life histories, making some ecosystems exquisitely sensitive to disturbance at particular stages. Add the accelerating pace of ocean industrialization, what researchers have called the blue acceleration, and the potential for cascading, unpredictable effects multiplies. The review emphasizes that cumulative effects can arise not only from multiple projects of the same kind, such as a cluster of wave energy converters, but from interactions across sectors: fisheries, shipping, tourism, offshore wind, and the overarching pressure of climate change. Effects may also appear only after a lag, or emerge only when a second activity unlocks the harm of the first.

Not all cumulative effects are additive, and this is where the science gets genuinely tricky. The review catalogues several interaction types. Additive effects are the intuitive case: the combined outcome equals the sum of the parts, and this pattern tends to dominate in physical systems. But ecosystems frequently behave nonlinearly. Synergistic effects occur when two stressors amplify each other, as when mortality from boat strikes and fishing combines with climate-driven heat stress to push a population below the threshold where it can reproduce. Dominant effects arise when one process overwhelms everything else, such as sediment smothering a seagrass meadow regardless of how ideal other conditions are. Antagonistic effects, where stressors dampen one another, can even flip an expected harm into a benefit. Two decades of research suggest these nonlinear outcomes are the norm in ecosystems, not the exception, yet most assessment tools in routine use cannot represent them.

The history of the field explains some of the current awkwardness. Cumulative effects entered regulatory consciousness in the 1970s and 1980s as terrestrial industrialization accelerated, but progress stalled for years because ecosystem-scale analysis demanded large datasets and heavy computation. By the 1990s, tiered risk assessment had become standard practice, beginning with qualitative risk tables before progressing to quantitative methods. Marine cumulative assessment surged from the 2000s, often driven by European policy such as the Marine Strategy Framework Directive. Of 154 marine assessments reviewed in one earlier synthesis, the majority simply combined geographic information system maps of overlapping human activities with expert judgments about how those activities interact. Expert elicitation remains a workhorse, and the review notes it is currently the most widely used route for bringing Indigenous knowledge into cumulative effects assessment, though repeatability remains a persistent weakness that structured techniques such as the Delphi method try to address.

The review organizes fifteen assessment methods along a gradient of increasing computational complexity. At the simple end sit interaction matrices and map-based overlays, including the weighted cumulative pressure mapping approach popularized by a landmark 2008 global study of human impacts on the ocean, variants of which account for more than half of all published assessments. These methods are transparent and digestible, which explains their appeal to stretched regulatory agencies, but they treat effects as linear and additive and cannot capture feedbacks or indirect pathways. In the middle tier are qualitative system models, notably loop analysis using signed digraphs, which have been applied to offshore wind in the Bay of Seine and elsewhere. These networks of positive and negative links between pressures and ecosystem components can be tested for stability and probed with perturbation analysis, and they pair naturally with Bayesian belief networks that attach conditional probabilities to each causal link.

At the demanding end lie fully dynamic numerical models: models of intermediate complexity focused on a handful of interacting species, and whole-of-system ecosystem models that track age-structured populations or biomass pools across space and time. Originally built for fisheries and climate research, these platforms are now being redirected at marine energy. Ecospace, an ecosystem model with fisheries roots, has been used to project cumulative impacts of combined aquaculture and marine energy platforms off Scotland, while spatiotemporal analyses have mapped offshore wind development scenarios across the North Sea basin. The review also flags an incoming wave of artificial intelligence: machine learning models already predict hydrodynamic and water quality responses relevant to energy siting, and large language models are being used to assemble ecosystem models far faster than manual parameterization allows. The authors predict AI and machine learning will make substantial contributions to cumulative assessment within a decade, while cautioning that models must still remain interpretable to the human decision-makers who rely on them.

Project-level practice, however, lags far behind the research frontier. Many consenting documents still amount to a cursory statement that no significant cumulative effects are anticipated, a formula the review’s authors describe as opaque and high-level. The critique echoes twenty years of published complaints: shifting baselines, single-industry focus, superficial expert assurances, and poor communication of uncertainty. Part of the fix, the authors argue, is structural. Nesting project-level assessments within strategic regional assessments, as maritime spatial planning frameworks increasingly attempt, would spread the analytical burden and ensure system-wide effects are not missed. A Baltic Sea case study, which embedded cumulative assessment at multiple scales into spatial planning and explicitly included marine energy among the activities considered, is held up as a template that balances rigor with pragmatism.

The recommendations that close the review are pointed. Assessments should use standardized terminology and terms of reference across jurisdictions; risk criteria should be set with stakeholders before analysis begins; predictive models should underpin assessments wherever possible; and every assumption, limitation, and uncertainty should be documented transparently, ideally communicated through IPCC-style confidence matrices. Centralized, interoperable data platforms, from the United States’ PRIMRE to Europe’s MARENDATA and France’s ResCORE, should be maintained and linked to reduce duplicated monitoring effort. Field studies must disentangle single-pressure from multi-pressure responses, and region-specific reference points defining acceptable levels of effect need to be developed, potentially borrowing methods from fisheries management. Crucially, the authors stress that assessments cannot remain a planning formality: they must be validated against post-deployment monitoring, shifting from hazard analysis to genuine evaluation of realized risk. For an industry still small enough to learn cheaply, the message is clear: borrow the hard-won methods of offshore wind, fisheries, and even terrestrial mining now, before the arrays grow too large for trial and error.

Subject of Research: Cumulative effects assessment methods for marine renewable energy development

Article Title: Marine renewable energy and cumulative effects assessments: Summary of the state of play

Article References: Fulton, E. A., Hemery, L. G., Copping, A. E., Farr, H. K., Fox, J., Garavelli, L., Hasselman, D. J., Miller, R., & O’Hagan, A. M. (2026). Marine renewable energy and cumulative effects assessments: Summary of the state of play. Environmental Challenges, 25, Article 101674. https://doi.org/10.1016/j.envc.2026.101674

Image Credits: AI Generated

DOI: 10.1016/j.envc.2026.101674

Keywords: marine renewable energy, cumulative effects assessment, environmental impact assessment, offshore energy, marine spatial planning, ecosystem modeling, tidal energy, wave energy, ocean stressors, Bayesian networks, marine ecosystems, regulatory policy