assessment of hydrodynamic impacts from tidal power lagoons in the bay of fundy

International Journal of Marine Energy 1 (2013) 33–54

Contents lists available at SciVerse ScienceDirect

International Journal of Marine Energy

journa l homepage: www.e lsev ier .com/ locate / i jome

Assessment of hydrodynamic impacts from tidalpower lagoons in the Bay of Fundy

2214-1669/$ - see front matter Crown Copyright � 2013 Published by Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.ijome.2013.05.006

⇑ Corresponding author. Tel.: +1 613 993 6690.E-mail addresses: [email protected] (A. Cornett), [email protected] (J. Cousineau),

uottawa.ca (I. Nistor).

Andrew Cornett a,⇑, Julien Cousineau b, Ioan Nistor b

a National Research Council Canada, 1200 Montreal Rd, Building M32, Ottawa K1A 0R6, Canadab Dept of Civil Engineering, University of Ottawa, Ottawa, Canada

a r t i c l e i n f o a b s t r a c t

Keywords:Tidal lagoonTidal powerNumerical modellingHydrodynamicsWater levelTide rangeBay of Fundy

The Bay of Fundy located in eastern Canada is home to some of theworld’s largest tides. Currently there is renewed interest in harness-ing these very large tides for power generation in ways that avoidupsetting ecosystems, infrastructure and human activities thatare presently well adapted to existing conditions. This paper inves-tigates the hydrodynamic impacts due to tidal power lagoons, anapproach to power generation that involves temporarily storingseawater behind a circular engineered dyke and generating powerby gradually releasing the impounded seawater through conven-tional low-head hydroelectric turbines. This paper describes a studyin which a two-dimensional, depth-averaged hydrodynamic modelbased on the TELEMAC modelling system was developed, cali-brated, and applied to analyze, predict, and quantify the potentialchanges in tidal hydrodynamics (water levels, tide range, circula-tion patterns and tidal currents) throughout the Bay of Fundy andGulf of Maine due to the presence of a single tidal lagoon and multi-ple lagoons operating at various locations in the upper Bay of Fun-dy. The sensitivity of the hydrodynamic impacts to changes inlagoon type, size, location, the number of lagoons, and their operat-ing mode have also been investigated. The methods employed inthis study and the main findings are presented and discussedherein. These results will help inform future decisions concerningdevelopment of the vast tidal energy resources in the Bay of Fundy.

Crown Copyright � 2013 Published by Elsevier Ltd. All rightsreserved.

inistor@

http://crossmark.dyndns.org/dialog/?doi=10.1016/j.ijome.2013.05.006&domain=pdf

http://dx.doi.org/10.1016/j.ijome.2013.05.006

mailto:[email protected]

mailto:[email protected]

mailto:inistor@ uottawa.ca

mailto:inistor@ uottawa.ca

http://dx.doi.org/10.1016/j.ijome.2013.05.006

http://www.sciencedirect.com/science/journal/22141669

http://www.elsevier.com/locate/ijome

34 A. Cornett et al. / International Journal of Marine Energy 1 (2013) 33–54

Introduction

The Bay of Fundy (BoF), located on the Atlantic coast of North America, between the Canadian prov-inces of Nova Scotia and New Brunswick, is renowned for its large amplitude tides, which are amongthe world’s largest (Fig. 1). The Fundy tides are semi-diurnal; twice each day roughly 115 billion ton-nes of seawater flow in and out of the 255 km long Bay. The tidal range near Burntcoat Head, located inMinas Basin, can exceed 16 m during spring tides. The natural geometry and bathymetry of the Bay ofFundy are the main factors responsible for producing these very large tides. The BoF, together withGulf of Maine (GoM), forms a funnel with a natural period of approximately 13 h, close to the12.42 h period of the M2 tidal forcing. The large tides are a result of the near-resonant response ofthe BoF-GoM system to the M2 tidal forcing [1]. Because of the large tidal range and the strong tidalcurrents that arise in certain locations, the Bay of Fundy has long been recognized as an ideal site fortidal energy projects.

In the early 1970s, a series of technical and economic assessments were performed to investigatethe feasibility and environmental impacts of a potential large-scale tidal barrage in the Bay of Fundy[2]. From these studies, it has been shown that small changes in the geometry of the Bay, associatedwith the construction of a tidal barrage, could produce significant changes in tidal amplitudes as faraway as Boston, USA. It was concluded that the development of a large-scale tidal barrage was consid-ered potentially hazardous to the surrounding ecosystem, and the project was never implemented.Although a large-scale tidal barrage was not constructed, a smaller-scale tidal barrage was completedat Annapolis Royal, Nova Scotia. The Annapolis Royal tidal barrage, which began operations in 1984,remains one of three tidal power plants operating worldwide.

Tidal power lagoon concept

Presently, there is some interest in the idea of implementing tidal power lagoons in the upper Bayof Fundy. The tidal lagoon concept is a more recent approach to tidal power conversion that attemptsto solve some of the environmental problems associated with tidal barrages. Rather than blocking off asection of the Bay with a barrage, the tidal lagoon concept involves constructing a circular impound-ment structure (a rubble-mound or caisson-type dyke) and a power-house containing sluices and con-ventional low-head hydroelectric turbines, situated a mile or more offshore, in an area with shallow

Fig. 1. Map of the Gulf of Maine and Bay of Fundy showing spring tide range.

A. Cornett et al. / International Journal of Marine Energy 1 (2013) 33–54 35

depths and high tides. A tidal power lagoon may also be attached to the shore, so that the shorelineforms a part of the impoundment (Fig. 2). Although tidal power lagoons are generally thought to betechnically feasible, their environmental impacts have not been investigated in detail, and none havebeen constructed to date.

In 2006, Delta Marine Consultants (DMC) was retained by Tidal Electric Canada (TEC) to assess thefeasibility of constructing tidal power lagoons in the upper Bay of Fundy. DMC [3] proposed construct-ing a tidal lagoon on the tidal flats along the northern shore of Minas Basin, between Five Islands andEconomy Point. Various plant layouts were investigated, including lagoons with single and multiplebasins and lagoons with direct and rectified flow through the power station. They concluded that a sin-gle basin with direct flow through the power station would be most cost efficient for the Minas Basinsite. DMC developed conceptual designs for two lagoon types: (a) an offshore lagoon comprising apower station and a 12 km2 circular impoundment enclosed by a 11.9 km long dyke detached fromthe shore; and (b) a coastal lagoon comprising a power station and a 24 km2 impoundment formed be-tween a 10.2 km long dyke and the existing shoreline. According to DMC, a 12 km2 offshore lagoon fit-ted with fourteen 7.5 m diameter bulb turbine generators (up to 20 MW each) and 15 sluice gates(56 m2 area each) would have a peak output of 280 MW and an average power output of approximately124 MW. The larger coastal lagoon, equipped with twenty-four 20 MW bulb turbines and 15 sluices, isestimated to have an average power output of approximately 220 MW, and peak output of 480 MW.

Any large tidal power lagoon operating in the upper BoF will likely modify the tides and tidal cur-rents near the lagoon, and could perhaps create effects that are felt throughout the entire BoF-GoMregion. The nature of these hydrodynamic impacts will likely depend on the size of the tidal lagoon,its location, and its method of operation. Changes in the tidal hydrodynamics may also affect themovement of sediments and/or alter natural ecosystems and human activities that are presently welladapted to existing conditions.

Previous investigations

Several authors have developed numerical models to simulate tidal hydrodynamics in the GoM andBoF over the years. Greenberg [4] estimated changes in the tidal range throughout the BoF and GoMassociated with various proposals to construct large tidal barrages in the upper part of the BoF.Because of the near-resonant state of the existing system, it has been shown that small changes in

Fig. 2. Conceptual layout of offshore and coastal tidal power lagoons.


the geometry of the Bay associated with the construction of a tidal barrage could produce significantchanges in tidal amplitudes as far away as Boston. Sucsy et al. [5] used two and three-dimensionalmodels to simulate the M2 tide in the Gulf of Maine and estimate the changes caused by a large tidalbarrier located in the upper Bay of Fundy. Tidal amplitudes in the presence of a barrier increased by upto 50 cm for both models, corroborating Greenberg’s earlier results.

Karsten et al. [6] applied a three-dimensional model to investigate the energy that could in theorybe extracted from the tidal flows at Minas Passage, a 5.5 km wide channel located at the entrance toMinas Basin in the upper part of the BoF. They showed that the energy that could (in theory) be recov-ered exceeds the kinetic energy associated with the undisturbed flow through the Passage by a factorof three or more. They also developed relationships between the amount of energy dissipated at thePassage and the change in the tidal range at various locations throughout the BoF and the GoM.

Cornett et al. [7] developed detailed two- and three-dimensional models of the tidal flows in theBoF and applied these models to delineate and assess the considerable kinetic energy resourcesaround the Bay associated with the energetic tidal currents.

The Severn Estuary in the UK is another ideal site for tidal renewable-energy projects, since it hasthe third highest tide range in the world, with a spring tide amplitude approaching 7 m. Xia et al. [8]and Falconer et al. [9] have examined the potential hydrodynamic impacts on the Severn Estuary dueto several different proposed renewable-energy projects, including a large coastal lagoon. They used atwo-dimensional finite volume numerical model and applied it to forecast the hydrodynamic impactsdue to several different power projects. They modelled a large coastal lagoon with an area of approx-imately 86 km2 and an installed capacity of 1500 MW, and concluded that the lagoon would have rel-atively little influence on the hydrodynamic processes in the Severn Estuary.

Xia et al. [10] examined the impact of constructing a very large tidal barrage across the Severn Estu-ary. They refined a two-dimensional hydrodynamic model and used the model to simulate the tidalhydrodynamics with and without the barrage. They concluded that discharges, water levels, and tidalcurrents would all decrease significantly above the barrage. Xia et al. [11] also discussed the influenceof alternative operating modes on the hydrodynamic impacts of a large tidal barrage across the SevernEstuary. After considering various factors, they conclude that, for efficient electricity generation and re-duced flood risk, ebb-generation or bi-directional generation are preferred over flood-generation.

Present study objectives

Previous studies [4,5] have suggested that large tidal barrages located in the upper BoF could haveimportant impacts on the tides and tide-related processes in the BoF and GoM. However, recent stud-ies from the UK [8,9] suggest a large tidal power lagoon located in the Severn Estuary would have les-ser impact on the hydrodynamic processes in the Estuary. The scale and character of the potentialhydrodynamic impacts due to tidal lagoons operating in the upper Bay of Fundy have not been inves-tigated and remain unknown.

The main objective of the research reported herein was to investigate the nature of the potentialhydrodynamic impacts within the BoF and GoM due to various hypothetical tidal power lagoon devel-opments. A numerical model was developed and applied to investigate the changes in hydrodynamicsthroughout the BoF and GoM due to several hypothetical development scenarios involving the pres-ence of a single tidal lagoon, or multiple lagoons, operating in the upper BoF. The extent of the changesdue to different scenarios involving several number, size and location of lagoons, as well as their oper-ating mode was also investigated. The ultimate purpose of this novel study was to support decisionsconcerning the management and development of the vast tidal energy resources available in the BoF.Some preliminary results from this study were previously reported by Cornett et al. [12] and Cousi-neau et al. [13].

Model development and calibration

In this study, tidal hydrodynamics in the BoF and GoM have been simulated using a numericalmodel based on the TELEMAC modelling system, developed by Electricité de France [14]. Most


phenomena of importance in free-surface flows can be included in this model, such as the friction onthe bed and lateral boundaries, wind stress on the free surface, Coriolis force, turbulence, and densityeffects. The dynamic wetting and drying of the tidal flats influences the tidal hydrodynamics in theupper part of the Bay, and these processes were included in the numerical simulations.

TELEMAC-2D employs finite-element methods to solve the Saint-Venant equations over a compu-tational domain subject to initial conditions and boundary conditions. Since the computational do-main is represented by a flexible unstructured triangular mesh, a higher spatial resolution can beused in areas of special interest and where spatial gradients are large (i.e., near shorelines), while alower resolution can be used elsewhere. The Saint-Venant equations can be written as:

@h@tþ �u � �rðhÞ þ h � divð�uÞ ¼ Sh ð1Þ

@u@tþ �u � �rðuÞ ¼ �g

@Z@xþ Sx þ

1h

divðhv t�ruÞ ð2Þ

@v@tþ �u � �rðvÞ ¼ �g

@Z@yþ Sy þ

1h

divðhv t�rvÞ ð3Þ

Where h = total water depth (m); x, y = horizontal space coordinates (m); u, v = depth-averaged veloc-ities in x, y direction, respectively (m/s); g = gravity acceleration (m/s2); vt = momentum coefficients(m2/s); Z = free surface water elevation (m); t = time (s); Sh, Sx and Sy = source term (m/s2). In thisstudy, the effects of bed friction and Coriolis force were included through the source terms. The eddyviscosity was calculated using the Smagorinsky model [15]. The wetting and drying of inter-tidal flatswas also included in the model.

Fig. 3. Computational domain, model mesh and bathymetry.


Domain and bathymetry

The computational domain spans a 235,289 km2 area, extending from Providence, Rhode Island, toHalifax, Nova Scotia, and includes the Bay of Fundy and the entire Gulf of Maine (Fig. 3). The offshoreboundary was set beyond the edge of the continental shelf so that the model could be used to assessthe hydrodynamic impacts along the entire GoM-BoF coastline. The domain included inter-tidal areasup to the highest astronomical tide (HAT) shoreline. In general, smaller elements were used to providea higher resolution in shallow coastal areas and in the upper Bay of Fundy, while coarser elementswere used further away from the coast. Several different unstructured model grids were developedfor this study, representing both existing conditions without lagoons and hypothetical future condi-tions with one or more lagoons operating in the upper part of the Bay. The grid of existing conditionscomprised 108,993 nodes and 205,148 triangular elements with sides lengths varying from 50 m to10 km.

The seabed elevation that was specified at each node was developed by integrating data from thefollowing sources:

� Geobase Canadian Digital Elevation Data,� United States Geological Survey (USGS) National Elevation Dataset,� Canadian Hydrographic Service (CHS) Multibeam Bathymetry,� CHS Nautical Charts,� National Oceanic and Atmospheric Agency (NOAA) Soundings and Electronic Navigational Charts,

and� British Oceanographic Data Centre (BODC) Gridded Bathymetric Chart of the Oceans.

The six datasets, originally referenced to different tidal datums, were adjusted to Universal Trans-verse Mercator North American Datum (NAD) 83 (Zone 19) coordinates in the horizontal plane andreferenced to North American Vertical Datum of 1988 (NAVD 88) in the vertical plane.

Boundary conditions

Many previous researchers considered only the M2 tidal constituent when studying the hydrody-namic impacts due to tidal power projects in the Bay of Fundy [4,5]. The M2 constituent is certainlydominant and contributes over 80% of the total tidal energy in the upper Bay [16]. However, additionalconstituents must be included in order to achieve more realistic predictions and more accurate assess-ments. Hence, the tidal flows due to the M2 constituent and nine other leading constituents were con-sidered in the present study.

Sea surface elevations and initial depth-averaged currents were interpolated along the open oceanboundary using harmonic constants for 10 leading tidal constituents (M2, N2, S2, K1, 01, K2, L2, 2N2,NU2, M4). The boundary conditions were derived from results produced by an existing depth-inte-grated finite-element model of tidal hydrodynamics that spanned a larger region including theGoM, BoF and part of the Scotian Shelf [17]. This regional model assimilates tidal constituents derivedfrom sea level observations made by the TOPEX-Poseidon altimeter. Freshwater inflows from riversare insignificant relative to the tidal flows, and hence were neglected in the present study.

Model calibration and validation

The new model was calibrated by adjusting the bottom roughness in different parts of the domainto minimize the error in tide range and time-of-tide between the model’s predictions and water leveltime histories for 150 NOAA/DFO tidal stations distributed throughout the computational domain. Foreach tidal station, model results were compared to predictions synthesized from harmonic constantsestablished by DFO and NOAA. The same set of 10 constituents (M2, N2, S2, K1, 01, K2, L2, 2N2, NU2, M4)used to drive the numerical model were also used to generate the target water level time histories foreach tidal station. The Strickler’s roughness coefficient for different areas was adjusted with valuesranging from 20 to 40 in order to minimize the calibration errors throughout the domain. The physical


characteristics of the seabed were taken into consideration when selecting the bottom roughness coef-ficient for each area. A model time step of 10 seconds was chosen to ensure numerical stability. Tideswere simulated for periods of 28 days in order to capture two full spring-neap tidal cycles.

The model was first calibrated for the M2 tidal constituent alone, and then for all 10 constituentscombined together. The model was calibrated to predict tides over a 28 day period in 2007, carefullyselected to include average spring and neap tides. Once successfully calibrated, the model was vali-dated by predicting the tides over three different 28 day periods. Fig. 4 shows two examples of the

Fig. 4. Simulated and observed water levels at two sites.

Fig. 5. Simulated and observed current velocities (u and v) in Minas Basin.


results obtained. It can be seen that the numerical model predictions agree very well with astronom-ical tide projection; both the minimum and maximum water levels, as well as their timing, are pre-dicted with good precision. The depth averaged tidal currents predicted by the model were alsocompared with available ADCP data for several sites in Minas Passage. The agreement between themeasured and modelled currents was satisfactory (see Fig. 5). Based on these results, it was concludedthat the calibrated model is able to predict tidal fluctuations throughout the Bay of Fundy and Gulf ofMaine with acceptable accuracy.

Simulation of tidal power lagoons

Dykes, turbines, sluices

In addition to the grid representing existing conditions, thirteen other grids were developed to rep-resent various hypothetical scenarios including one or more tidal lagoons. The lagoons were modelledby refining the mesh locally to a resolution of 100 m along the impoundment dyke and around thepowerhouse. The mesh was only refined near the lagoon and remained unchanged away from the la-goon. The dyke and powerhouse were idealized in the model as vertical sided structures. Similar to Xiaet al. [8], the interior of the lagoon was modelled as a sub-domain that was disconnected from themain domain outside the lagoon. Multiple pairs or sources and sinks were defined to transfer seawaterbetween the two domains, simulating the flows through the sluices and turbines within thepowerhouse. For each lagoon layout, the powerhouse was assumed to have turbines located centrallybetween the sluices as shown in Fig. 6.

In these simulations, the discharge through each sluice QS is given by:

Q s ¼ CsAffiffiffiffiffiffiffiffiffiffiffiffiffi2gDH

pð4Þ

where Cs is the discharge coefficient for a sluice; A is the cross-sectional area (56 m2); g is the accel-eration due to gravity (9.806 m/s2); and DH is the difference in water level across the powerhouse.Although not entirely physically accurate, a constant discharge coefficient Cs = 0.75 was assumed inthis study for simplicity.

The performance of a turbine is often defined as a ‘‘hill chart’’, relating specific discharge, unit speedand efficiency. These charts are based largely on physical and computational fluid dynamics models.

Fig. 6. Domain decomposition and grid refinement near the dyke.

Fig. 7. Assumed relationships between head, discharge and power output for 20 MW low-head turbines.


However, detailed information on the performance of turbines and generators for tidal power schemesis difficult to obtain due to a desire by manufacturers to maintain confidentiality. An idealized butrealistic turbine performance curve was developed for use in this study, based on data from [18],which has been adapted to suit the case where the maximum power output to the generator is20 MW, the minimum operating head is 2 m, and the normal operating head is 6 m. The turbine per-formance curves employed in this study are shown in Fig. 7. As seen in this figure, the power outputincreases with increasing water head up to a head of 6 m, and then remains constant at 20 MW forheads above 6 m. The discharge through the turbine also reaches its maximum when the water headis around 6 m, and decreases for both smaller and larger heads. In this study, efficiencies of 90% onlagoon outflow and 80% on lagoon inflow were assumed. The turbine performance curves shown inFig. 7 were employed in the numerical simulations to regulate the discharge through each turbine

Fig. 8. Water levels and power output for the two-way power generation protocol.


as a function of the water head across the powerhouse, and to compute the power output at each timestep.

Operating modes

Tidal lagoons can be operated in several different modes, including ebb generation, flood genera-tion and two-way generation. Ebb-generation is used at the Annapolis Royal Tidal Generating Station.For ebb-generation, the sluices are used to fill the lagoon during flood tide, and power is generatedonly during ebb tide. For flood generation, power is generated during flood tide only, and the lagoonis drained during ebb tide. For two-way generation, power is generated when filling the lagoon duringthe flood tide and when draining the lagoon during the ebb tide. Two-way generation requires that theturbines operate in both forward and reverse directions. The two-way power generation protocol isdepicted in Fig. 8. Each tidal cycle comprises eight stages: holding, generating, generating-releasing,

Fig. 9. Location and layout of hypothetical tidal lagoons: (a) coastal lagoons, (b) offshore lagoons.


releasing, holding, generating, generating-filling, and filling. Each generating stage begins when thehead difference across the powerhouse exceeds a minimum value, assumed to be 2 m in this study;and ends when the head difference falls below a minimum level, also assumed to be 2 m.

Modelled scenarios

The numerical model has been used to study the effects that one or more operating tidal lagoonsmight have on the tidal hydrodynamics in the Bay of Fundy and Gulf of Maine. Nineteen differenthypothetical scenarios with one or more tidal power lagoons located in Minas Basin and/or in Chign-ecto Bay have been modelled. These nineteen scenarios were developed to allow the influence of var-ious parameters to be assessed, including: lagoon type; lagoon location; lagoon size; number oflagoons; and operating protocol. Single lagoons were simulated in some scenarios, while up to six la-goons operating simultaneously were modelled in others. In this way, the relationship between thescale of lagoon development and the scale of potential impacts has been ascertained.

Coastal and offshore tidal lagoons were implemented at six different sites. These sites, named sitesA to F, are mapped in Fig. 9. Site A, which was proposed by DMC [3], is located along the northernshore of Minas Basin between Five Islands and Economy Point. Site B is also located along the northernshore of Minas Basin, east of site A near Upper Economy. Site C is located along the southern shore ofMinas Basin near Cambridge. Site D is located along the western shore of Chignecto Bay near Mount-ville. Site E is located on the eastern shore of Chignecto Bay near Ragged Reef. Finally, site F is locatedon the south-western shore of Chignecto Bay between Alma and Cape Enrage. Several key parametersfor the hypothetical development scenarios considered in this study are summarized in Table 1.

The same boundary forcing, based on 10 harmonic tidal constituents, was employed in every sce-nario. All the simulations spanned the same 15 day period for which the spring and neap tides werevery close to long term average conditions. In addition to modelling hypothetical scenarios, the tidalflows for existing conditions without lagoons were also modelled for the same 15 day period using thesame boundary conditions. The changes in tidal hydrodynamics due to each scenario were obtained bydifferencing the model results for each scenario from those obtained for existing conditions. The im-pacts of each scenario on water levels, tide range, circulation patterns and tidal currents have all beenassessed in this manner.

Table 1Key parameters for hypothetical development scenarios.

Scenario # oflagoons

Site (s) Lagoontype

Total impoundmentarea (km2)

Total # ofturbines

Total # ofsluices

Operatingmode

S1 1 A Coastal 26.7 24 15 2-WayS2 1 A Offshore 12.0 14 15 2-WayS3 1 A Coastal 26.7 22 31 2-WayS4 1 A Coastal 26.7 20 26 EbbS5 1 A Coastal 26.7 16 17 FloodS6 1 A Coastal 26.7 – – OpenS7 1 A Coastal 26.7 – – ClosedS8 1 A Coastal 35.1 30 41 2-WayS9 1 A Coastal 57.7 52 72 2-Way

S10 1 A Offshore 12.0 12 23 2-WayS11 1 A Offshore 18.0 19 29 2-WayS12 1 A Offshore 24.0 23 39 2-WayS13 1 B Coastal 34.2 29 31 2-WayS14 1 C Coastal 34.0 31 40 2-WayS15 3 A, B, C Coastal 94.8 82 102 2-WayS16 3 A, B, C Offshore 36.0 36 69 2-WayS17 3 D, E, F Coastal 84.0 78 109 2-WayS18 3 D, E, F Offshore 36.0 36 69 2-WayS19 6 A, B, C, D, E, F Offshore 72.1 72 138 2-Way


Results and discussion

Single lagoon at site A

The tidal power lagoons modelled in scenarios S1 and S2 were developed to replicate the singlecoastal and offshore lagoons proposed by DMC [3]. Fig. 10 shows the depth-averaged velocity fieldaround a 26.7 km2 coastal lagoon located along the north shore of Minas Basin (scenario S1) that isfollowing a bi-directional operating scheme. The numerical schematisation of this coastal lagoon isshown in Fig. 6. Fig. 10 shows typical flow patterns while the tide is flooding and ebbing. Duringthe flood phase, seawater flows into the lagoon mainly through the turbines (some water flowsthrough the sluices near the end of the flood stage). Within the lagoon, velocities are highest nearthe powerhouse, and diminish with increasing distance away from the powerhouse. Outside the la-goon, relatively strong currents occur near the powerhouse and along the southern perimeter of theimpoundment dyke. During the ebb phase, seawater flows out of the lagoon mainly through the tur-bines, although some water passes through the sluices near low water. Relatively strong velocities oc-cur near the powerhouse, both inside and especially outside the lagoon. The peak velocities near thepowerhouse tend to be stronger during the ebb than during the flood, due to the shallower waterdepths outside the lagoon near the end of the ebb phase.

A tidal lagoon will induce local changes in the direction and strength of the tidal currents bothwithin the lagoon and outside it, especially near the powerhouse and along the external perimeterof the impoundment dyke. Depending on seabed conditions, scour protection may be required in theseareas to mitigate the high energy flows emerging from the sluice gates and turbines. Scour protectionmay also be required to stabilize and protect the toe of the impoundment dyke.

In this study the maximum tide range is defined as the difference between the maximum and min-imum water levels over the 15 day simulation period. The numerical simulations indicate that a26.7 km2 coastal lagoon at site A (scenario S1) will cause changes in maximum tide range varying fromnil in the eastern part of Minas Basin, up to +5 cm at Minas Passage, and ranging from +1 to +4 cm overa large portion of the BoF and GoM shoreline (see Fig. 11a). For the smaller 12 km2 offshore lagoon(scenario S2), except for areas very close to the lagoon, the predicted impact on maximum tide rangeis less than +2 cm at Minas Passage, and ranges from 0 to +2 cm along the shore of the BoF and GoM(see Fig. 11b).

The depth-averaged root-mean-square (RMS) velocity, URMS, is used to characterize speed of thetime-varying tidal currents. URMS is defined as:

Fi

URMS ¼

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiPni¼1U2

i

n

sð5Þ

g. 10. Velocity field near a 26 km2 coastal lagoon during: (a) flood and (b) ebb (bi-directional generation mode).

Fig. 11. Change in maximum tide range for single coastal and offshore lagoons at site A (scenarios S1 and S2).


where Ui is the depth-averaged velocity at time step i, and n is the number of steps in the simulation.Fig. 12 shows the predicted change in depth-averaged RMS current velocity (URMS) near the coastaland offshore lagoons modelled in scenarios S1 and S2. For the coastal lagoon, significant changes inflow speed (URMS more than 10 cm/s) are confined to local areas around the lagoon and within thelagoon. It can be seen that, relative to present conditions, flow speeds are significantly greater nearthe powerhouse outside the lagoon. There is also a large area inside the lagoon where speeds arereduced significantly. At the center of Minas Passage, the RMS current velocity is predicted to decreaseby up to 6 cm/s for scenario S1 and by up to 3 cm/s for scenario S2. The tidal currents in distant loca-tions beyond Minas Passage are virtually unchanged by the lagoons in Minas Basin.

Fig. 12. Change in RMS current speed for single coastal and offshore lagoons at site A (scenarios S1 and S2).


Relationship between power output and lagoon size

The power output, P, produced by a tidal power lagoon will vary considerably over time dependingon whether or not there is flow through the turbines and the head forcing the flow (Fig. 8). The root-mean-square (RMS) value of the power output, PRMS, can be used to characterise the time-varyingpower output. PRMS is defined as:

PRMS ¼

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiPmj¼1

Pni¼1Pi

� �2

m

sð6Þ

where Pi is the instantaneous power produced by turbine i, n is the number of turbines, j is the sim-ulation step number, and m is the number of time steps in the simulation. The power output for thetidal power lagoons considered in this study is found to be directly proportional to the combinedimpoundment area, A. Results for the bi-directional generation mode are shown in Fig. 13. In this

Fig. 13. Relationship between combined lagoon area and power output.


figure, the triangular and circular symbols denote results from hypothetical scenarios with coastal andoffshore tidal lagoons, respectively. For offshore lagoons in Minas Basin, PRMS = 9.33A provides a gooddescription of the relationship between RMS power output (MW) and impoundment area (km2), whilefor coastal lagoons in Minas Basin, PRMS = 8.32A. For a fixed lagoon size, the power output in ChignectoBay is slightly less than in Minas Basin due to the lower tide range.

Hydrodynamic impacts at reference sites

As expected, the authors’ simulations suggest that the scale of the changes in tidal hydrodynamicsincreases with increases in the scale of lagoon development. While multiple lagoons or larger lagoonslocated in the upper BoF will generate more electrical power, they also constitute a larger perturbationto a near resonant system, and can therefore be expected to induce larger changes in tidal hydrody-namics. The results of this study delineate the trade-off between power generation from tidal lagoonsand hydrodynamic impact.

The hydrodynamic changes (relative to existing conditions) at several locations in the Bay of Fundyand Gulf of Maine, derived from the numerical simulations of scenarios S1–S19, are summarised be-low in Table 2. The change in maximum tide range is shown for five tide stations distributed across thestudy domain, namely: Boston, Bar Harbor, Saint John, Chignecto and Five Islands. Similarly, thechange in RMS tidal current speed is shown for five reference sites, namely: Minas Basin, Minas Pas-sage, Minas Basin Entrance (MB Entrance), Chignecto Bay Entrance (CB Entrance) and Chignecto. Thesereference stations and reference sites are mapped in Fig. 14.

In Table 2, the change in maximum tide range is expressed as a percentage of the maximum tiderange at that station for existing conditions without tidal power lagoons. These values were obtainedfrom simulation S0, and are also shown in Table 2. The predicted changes in RMS tidal current speedfor each reference site are also expressed as a percentage of URMS for existing conditions, as predictedin simulation S0.

Impacts on tide range

The predicted impacts of the various hypothetical scenarios considered in this study on the max-imum tide range at Boston are plotted in Fig. 15a versus the RMS power output. Blue diamonds areused to denote results for scenarios with tidal lagoons in Minas Basin (scenarios S1–S3 and S8–S16), violet squares denote results for scenarios S17 and S18, with tidal lagoons in Chignecto Bay,and a red diamond denotes the result for scenario S19, where lagoons in Minas Basin and Chignecto

Table 2Summary of hydrodynamic changes at reference sites for scenarios S1–S19.

Boston(8443970)

Bar harbor(8413320)

Saint John(65)

Chignecto(40139)

FiveIslands (260)

MinasBasin

MinasPassage

MBEntrance

CBEntrance

Chignecto RMS poweroutput,PRMS(MW)

Long. (�) �71.052 �68.205 �66.067 �64.983 �64.067 �64.211 �64.443 �64.784 �65.043 �64.652

Lat. (�) 42.355 44.392 45.267 45.483 45.383 45.314 45.349 45.224 45.450 45.636

Scenario Maximum tide range (m) RMS current velocity, URMS (m/s)

S0 – 3.60 3.89 7.32 10.14 12.76 0.22 2.01 0.37 0.58 0.70 0S1 D 0.4% 0.5% 0.5% 0.4% 0.0% �13.4% �2.2% �5.1% 0.2% 0.4% 265S2 D 0.3% 0.3% 0.2% 0.2% �0.1% �7.9% �1.4% �3.2% 0.2% 0.1% 156S3 D 0.4% 0.5% 0.5% 0.4% 0.1% �14.4% �2.4% �5.6% 0.2% 0.4% 264S4 D 0.5% 0.4% 0.5% 0.4% 0.0% �12.0% �1.9% �4.3% 0.3% 0.4% 246S5 D 0.4% 0.4% 0.5% 0.4% 0.0% �11.6% �1.9% �4.3% 0.3% 0.4% 172S6 D 0.3% 0.3% 0.2% 0.1% �0.7% �7.4% �1.4% �3.7% 0.0% 0.1% 0S7 D 0.4% 0.5% 0.5% 0.5% 0.4% �11.6% �1.8% �3.5% 0.5% 0.6% 0S8 D 0.7% 0.8% 0.7% 0.6% 0.2% �20.8% �3.5% �8.0% 0.3% 0.6% 373S9 D 1.2% 1.6% 1.2% 1.0% – �38.0% �6.4% �14.4% 0.7% 1.0% 658

S10 D 0.3% 0.3% 0.2% 0.2% �0.1% �8.3% �1.4% �3.5% 0.0% 0.1% 142S11 D 0.4% 0.5% 0.4% 0.3% �0.1% �13.9% �2.4% �5.6% 0.2% 0.3% 239S12 D 0.6% 0.7% 0.6% 0.5% �0.1% �19.4% �3.3% �7.5% 0.3% 0.4% 296S13 D 0.5% 0.6% 0.8% 0.7% 0.9% �16.2% �2.4% �4.8% 0.7% 0.7% 360S14 D 0.6% 0.8% 0.5% 0.4% 0.0% �12.5% �3.9% �9.4% 0.2% 0.4% 385S15 D 1.7% 2.2% 1.7% 1.4% 0.8% �45.8% �9.0% �20.3% 0.9% 1.4% 988S16 D 0.7% 1.0% 0.7% 0.5% 0.0% �20.8% �4.6% �11.0% 0.2% 0.4% 411S17 D 2.4% 0.8% 0.0% �0.9% �0.4% �2.3% �0.5% �2.1% �12.0% �18.5% 747S18 D 1.3% 0.4% �0.2% �0.9% �0.6% �2.8% �0.6% �2.7% �8.2% �14.1% 312S19 D 2.0% 1.5% 0.5% �0.4% �0.8% �23.1% �5.2% �13.6% �8.0% �13.6% 723

48A

.Cornettet

al./InternationalJournal

ofM

arineEnergy

1(2013)

33–54

Fig. 14. Map of reference stations and reference sites.

Fig. 15. Influence of tidal power lagoons on the maximum tide range at: (a) Boston, (b) Bar Harbor, (c) Saint John and (d)Chignecto.



Bay were modelled together. For tidal power lagoons in Minas Basin, these results reveal a strong lin-ear relationship between the power output and the change in amplitude of the Boston tides. Althoughlagoons in Chignecto Bay were only considered in two scenarios, these results also suggest a linearrelationship between power output and the change in the amplitude of the tides at Boston. Greaterpower generation is clearly linked to stronger hydrodynamic impacts. Moreover, for equal power out-put, the impact on the Boston tides due to tidal power lagoons located in Chignecto Bay will beroughly double the impact of lagoons located in Minas Basin. In other words, the tides at Bostonare roughly twice as sensitive to power generation in Chignecto Bay compared to equivalent powergeneration in Minas Basin.

The predicted change in the maximum tide range at Bar Harbor due to the various hypothetical sce-narios considered in this study is plotted in Fig. 15b as a function of RMS power output. As at Boston,separate linear relationships are found between power output and the change in tide range at Bar Har-bor for lagoons in Minas Basin and in Chignecto Bay. However, unlike at Boston, these results indicatethat the tides at Bar Harbor are more sensitive to power generation in Minas Basin than in ChignectoBay. For equal power output, the impact on the Bar Harbor tides due to tidal power lagoons in Chign-ecto Bay will be roughly half as large as the impact due to lagoons in Minas Basin.

The predicted change in the maximum tide range at Saint John is plotted versus RMS power outputin Fig. 15c. Again, the results indicate a strong linear relationship between power generation in MinasBasin and the impact on the tide range at Saint John. Greater power generation is again clearly linkedto stronger hydrodynamic impact. However, results for scenarios with lagoons in Chignecto Bay do notfollow this rule. In fact, although only two data points are available, they suggest that producingpower in Chignecto Bay has virtually no impact on the tide range at Saint John.

Finally, Fig. 15d shows the predicted change in the maximum tide range at the Chignecto referencestation plotted as a function of the RMS power output. In this case, the results confirm a that the tiderange at Chignecto will increase linearly in response to increases in the power output from lagoons inMinas Basin. However, for the case of lagoons in Chignecto Bay, the results suggest that the tide rangeat Chignecto will decrease in response to increases in power output. Generating power in Minas Basinincreases the tides at the Chignecto station, while generating power in Chignecto Bay reduces them.Based on these observations it seems reasonable to conclude that changes in the tide range at Chign-ecto might be minimized by generating power in both Minas Basin and Chignecto Bay at the sametime. The results for the ‘mixed’ scenario S19 support this conclusion. This interesting result is partic-ular to the Chignecto reference station, and does not apply elsewhere.

In summary, lagoon development in Chignecto Bay will increase the tide range at Boston and BarHarbor, but not at Saint John, whereas lagoon development in Minas Basin will increase the tide rangeat all three cities.

Influence of operating mode

Fig. 16 illustrates the variation in hydrodynamic impact (change in maximum tide range) with dis-tance for a 26.7 km2 coastal lagoon in Minas Basin at site A (scenarios S3–S7). Also shown is the influ-ence of five different operating modes (flood generation, ebb generation, 2-way generation, gatesopen, gates closed) on the scale of the impacts. The impact at Five Islands, located only 15 km fromsite A, is relatively sensitive to the operating mode, whereas the impact at more distant sites, suchas Bar Harbor and Boston, are fairly insensitive to changes in operating mode. These results show thatthe far-field change in tide range is minimized for scenario S6, where the sluice gates are always open,and maximized for scenario S7, where the sluices are always closed. The hydrodynamic impact at eachstation is similar for the three different power generating modes modelled in scenarios S3, S4 and S5(2-way generation, ebb-generation and flood-generation, respectively), which confirms that the far-field impacts are fairly insensitive to the mode of power generation.

Impacts on tidal currents

The predicted change in RMS current speed at Minas Passage for each of the various hypotheticalscenarios considered in this study is plotted in Fig. 17 as a function of RMS power output. These results

Fig. 16. Change in maximum tide range versus distance for scenarios S3–S7.

Fig. 17. Influence of power generation by tidal lagoons on current speed at Minas Passage.


reveal the presence of a strong linear relationship between power generation from tidal lagoons inMinas Basin and reductions in current speed at Minas Passage. The linear relationship between poweroutput and hydrodynamic impact is virtually identical for both offshore lagoons and coastal lagoons.These results also show that the flows at Minas Passage are insensitive to power generation from tidallagoons in Chignecto Bay.

The changes in maximum depth-averaged current speed at the five reference sites mapped inFig. 14 are plotted in Fig. 18 as a function of the distance from site A. At each site, the impact is small-est for scenario S6, where the sluice gates are always open, and greatest for scenario S3, where a 2-waygeneration scheme was modelled. The predicted changes in tidal flows at distant locations are verysmall and quite insensitive to changes in operating mode. Closer to the lagoon, the changes are largerand more sensitive to the operating mode.

Fig. 18. Change in maximum depth-averaged current speed versus distance for scenarios S3–S7.


Conclusions

This article presents results from a comprehensive study assessing the hydrodynamic impactsthroughout the Gulf of Maine and the Bay of Fundy resulting from tidal power lagoons located inthe upper part of the Bay, in Minas Basin and Chignecto Bay. A detailed two-dimensional (depth-aver-aged) hydrodynamic model based on the TELEMAC modelling system has been developed, verified,and applied to investigate the changes in water levels and tidal currents for a range of hypotheticallagoon development scenarios. The model was successfully calibrated and validated against water le-vel observations for many tide stations throughout the region, and can be used to predict tidal hydro-dynamics with acceptable accuracy. Lagoons are simulated by partitioning the model domain intosub-domains, and using source-sink pairs to simulate the exchange of seawater through sluices andturbines. The flows through sluices are estimated using an orifice equation, while the flows throughturbines and the resulting power output are based on performance curves typical of double-regulatedbulb turbines. This modelling approach is able to provide reasonable simulations of the hydrodynamicprocesses at man-made tidal power lagoons.

The model was subsequently applied to predict tidal hydrodynamics for present conditions with-out lagoons, and for nineteen different hypothetical scenarios featuring from one to six individual la-goons. Lagoons were simulated at three sites in Minas Basin and at three sites in Chignecto Bay. Thescenarios included variations in lagoon size, lagoon type (coastal or offshore), lagoon location, thenumber of lagoons, and the operating mode (2-way generation, ebb-generation or flood-generation).Hydrodynamic impacts have been estimated by differencing results for each hypothetical scenariofrom equivalent results for existing conditions without lagoons. The sensitivities to changes in lagoonlocation, lagoon size, lagoon type, operating mode and the number of lagoons have been investigatedby comparing results for various scenarios.

As expected, the numerical results indicate that any tidal power lagoon will induce some largechanges in the pattern and strength of the local tidal currents, particularly near the powerhouseand possibly along portions of the impoundment dyke. Numerical results indicate that the changesin water levels and tidal current velocities are largest near the lagoon and generally diminish withincreasing distance. Notably, however, a small change in tide range is predicted throughout the entireGulf of Maine, even for the smallest development scenario.

For the lagoons considered in this study, the power output for a given operating mode was foundto be proportional to the surface area of the impoundment. However, because the tide range in MinasBasin is slightly larger than in Chignecto Bay, a lagoon located in Minas Basin can generate slightly


more energy per unit surface area. While larger tidal power lagoons and more numerous lagoons willproduce more energy, they also constitute a larger perturbation to a near resonant system, and cantherefore be expected to cause larger changes in the tidal hydrodynamics. The present results con-firm this hypothesis; the magnitude of the changes in tidal hydrodynamics increase with the scaleof lagoon development, with larger lagoons and multiple lagoons inducing greater hydrodynamicchanges.

The trade-off between power generation and hydrodynamic impact is quantified by way of a set oflinear relationships linking power output and hydrodynamic change for several communities aroundthe Bay of Fundy and Gulf of Maine, including, Boston, Bar Harbor and Saint John. For example thenumerical results indicate that a single 26.7 km2 coastal lagoon operating in Minas Basin with anRMS power output of 264 MW will cause the tide range in Boston to increase by �1.4 cm, while threecoastal lagoons operating in Minas Basin with a combined area of 94.8 km2 and a combined RMSpower output of 988 MW will lead to a �7.2 cm increase in the Boston tides. Interestingly, the tiderange at Boston is found to be more sensitive to lagoon development in Chignecto Bay than in MinasBasin, whereas the tide range at Bar Harbor, is more sensitive to lagoon development in Minas Basin.Furthermore, at Saint John the tide range is found to be sensitive to lagoon development in Minas Ba-sin, but insensitive to lagoon development in Chignecto Bay. In summary, lagoon development inChignecto Bay will increase the tide range at Boston and Bar Harbor, but not at Saint John, whereaslagoon development in Minas Basin will increase the tide range at all three cities. While the scaleof these changes represent a small fraction of the tide range at each community, their potential impacton communities and ecosystems warrants careful consideration and further investigation. Withoutfurther study it is difficult to comment on whether the potential benefits of tidal power lagoons mightoutweigh the drawbacks associated with these changes in tidal hydrodynamics.

A strong linear relationship is also found linking the scale of the lagoon development in Minas Ba-sin to the attenuation in the speed of the depth-averaged currents at Minas Passage, with more ambi-tious developments causing larger attenuation (up to �9% reduction in current speed for an RMSpower output of 988 MW). The speed of the current at Minas Passage is found to be insensitive to la-goon development in Chignecto Bay.

Three different operating modes have been modelled and assessed in this study, namely: flood-generation, ebb-generation and 2-way generation. The flood-generation mode is found to yield lessenergy than either the ebb-generation or 2-way generation modes, which produce similar amountsof energy. While the choice of operating mode has considerable influence on local velocities nearthe lagoon, particularly near the powerhouse, the operating mode appears to have little influenceon the magnitude of far-field hydrodynamic impacts.

In this study, methods for simulating the presence and operation of man-made tidal power lagoonswithin a 2D hydrodynamic model have been developed and applied to discover and assess the mag-nitude and character of the changes in tidal hydrodynamics throughout the Bay of Fundy and Gulf ofMaine that would result from tidal power lagoons operating in Minas Basin and in Chignecto Bay. Thestudy has generated new information concerning the relationships between power generation andhydrodynamic change, and the influences on these relationships of variations in lagoon size, lagoonlocation, lagoon type, operating mode and the number of lagoons. These results can be used to informpolicies and decision concerning the development of the vast tidal energy resources in the Bay ofFundy.

Acknowledgements

The authors gratefully acknowledge funding from the Nova Scotia Association for Offshore EnergyEnvironmental and Technical Research, and from Natural Resources Canada.

References

[1] C. Garret, Tidal resonance in the Bay of Fundy and Gulf of Maine, Nature 238 (1972) 441–443.[2] Bay of Fundy Tidal Power Review Board. 1977. Reassessment of Fundy Tidal Power. Ottawa, Report prepared for the

Minister of Supply and Services (Canada), under, Contract No. 02KX 23384–5207.

http://refhub.elsevier.com/S2214-1669(13)00007-6/h0005


[3] Delta Marine Consultants, Tidal Power Plant: Bay of Fundy, Technical Feasibility Study. 2007.[4] D.A. Greenberg, A numerical model investigation of tidal phenomena in the Bay of Fundy and Gulf of Maine, Mar. Geod. 2

(2) (1979) 161–187.[5] P.V. Sucsy, B.R. Pearce, V.G. Panchang, Comparison of two- and three-dimensional model simulation of the effect of a tidal

barrier on the Gulf of Maine tides, J. Phys. Oceanogr. 23 (1992) 1231–1248.[6] R.H. Karsten, J.M. McMillan, M.J. Lickley, R.D. Haynes, Assessment of tidal current energy in the Minas Passage, Bay of

Fundy, in: Proc. IMechE Vol. 222 Part A: J. Power Energy, 2008, 493–507.[7] A. Cornett, N. Durand, M. Serrer, 3-D Modelling and assessment of tidal current resources in the Bay of Fundy, Canada, in:

Proc. 3rd Int. Conf. on Ocean Energy, 2010, Bilbao.[8] J. Xia, R. Falconer, B. Lin, Impact of different tidal renewable energy projects on the hydrodynamic processes in the Severn

Estuary, UK, Ocean Modell. 32 (2010) (2010) 86–104.[9] R. Falconer, J. Xia, B. Lin, R. Ahmadian, The Severn barrage and other tidal energy options: hydrodynamic and power output

modelling, Sci. China Ser. E-Tech. Sci. 52 (11) (2009) 3413–3424.[10] J. Xia, R.A. Falconer, B. Lin, Hydrodynamic impact of a tidal barrage in the Severn Estuary, UK, Renewable Energy 35 (2010)

1455–1468.[11] J. Xia, R.A. Falconer, B. Lin, Impact of different operating modes for a Severn Barrage on the tidal power and flood

inundation in the Severn Estuary, UK, Appl. Energy 87 (2010) 2374–2391.[12] A. Cornett, J. Cousineau, I. Nistor, Hydrodynamic Impacts due to Tidal Power Lagoons in the Upper Bay of Fundy, Canada,

in: Proc. 2011 European Wave and Tidal Energy Conference (EWTEC), Southampton, UK.[13] J. Cousineau, I. Nistor, A. Cornett, Hydrodynamic Impacts of Tidal Power Lagoons in the Bay of Fundy, in: Proc. 2012

International Conference on Coastal Engineering (ICCE), Santander, Spain.[14] J.-M. Hervouet, Hydrodynamics of Free-Surface Flows – Modelling with the Finite Element Method, John Wiley & Sons Ltd,

2007.[15] J. Smagorinsky, General circulation experiments with the primitive equations I: the basic experiment, Mon. Weather Rev.

91 (1963) 99–164.[16] A. Zhang, E. Wei, B.B. Parker, Optimal estimation of tidal open boundary conditions using predicted tides and adjoint data

assimilation technique, Cont. Shelf Res. 23 (11–13) (2003) 1055–1070.[17] F. Dupont, C.G. Hannah, D.A. Greenberg, Modelling the sea level of the upper Bay of Fundy, Atmosphere-Ocean 43 (1)

(2005) 33–47.[18] A.C. Baker, Tidal Power, Peter Peregrinus Ltd., London, United Kingdom, 1991.





















assessment of hydrodynamic impacts from tidal power lagoons in the bay of fundy

Documents