# An Evaluation of the Large-Scale Implementation of Ocean Thermal Energy Conversion (OTEC) Using an Ocean General Circulation Model with Low-Complexity Atmospheric Feedback Effects

^{1}

^{2}

^{3}

^{*}

## Abstract

**:**

## 1. Introduction

^{−2}(24 h day basis) [8]; the area favorable for OTEC development may be as large as 140 million km

^{2}, and the net thermodynamic efficiency of OTEC systems is up to 3%. Combining these figures would yield an OTEC resource of 1000 TW, a staggeringly large number compared to humankind’s primary energy supply estimated at 18 TW in 2014 [9]. Offhand, it is not clear that invoking solar power alone is even appropriate since all heat fluxes at the ocean surface tend to balance one another [10]. More importantly, this approach fails to recognize that relatively large seawater flow rates are required to operate OTEC power plants, and that the availability of deep cold seawater in tropical areas depends on large-scale oceanic circulation. This fact actually provides a basic flow scale for the sustainable production of OTEC. From a 30 Sv (1 Sv = 10

^{6}m

^{3}s

^{−1}) estimate of the thermohaline circulation that ventilates ocean basins, and a typical 3 m

^{3}s

^{−1}requirement to generate 1 MW of OTEC electricity, Cousteau and Jacquier sized the global OTEC resource at 10 TW [11]. This extremely simple argument is attractive in principle, but it still assumes that the cold seawater intensity of OTEC power production does not change, i.e., that the thermal structure of the water column is unaffected.

## 2. Modelling Approach

#### 2.1. Ocean General Circulation Model with Sources and Sinks

^{−4}m

^{2}s

^{−1}for viscosity and 10

^{−5}m

^{2}s

^{−1}for diffusivity. The schemes of Redi [24], and Gent and McWilliams [25] are used for the parameterization of geostrophic eddies with the isoneutral and thickness diffusion coefficients set at 10

^{3}m

^{2}s

^{−1}. Sea ice is not explicitly modeled, but the condition that seawater temperature should not drop below the ice point (≈−2 °C) is enforced through compensating heat fluxes. As will be seen later in Section 4.2, this crude representation of ice dynamics leads to the lack of bottom water formation around Antarctica in the model simulations.

_{cw}, the volume flow rate of OTEC deep seawater per unit (latitude-longitude) area. The OTEC net power area density P

_{net}is calculated from the following formula [17]:

^{−3}, the seawater specific enthalpy c

_{p}as 4000 J kg

^{−1}K

^{−1}, and the OTEC combined turbo-generator efficiency ε

_{tg}as 0.75. The numerical coefficients account for a surface-to-deep seawater flow rate ratio of 1.5, and seawater pumping power losses equal to 30% of the turbo-generator output at standard conditions (T = 300 K, ΔT = 20 K). We note in passing that these somewhat arbitrarily defined standard conditions for the OTEC thermal resource are conservative, and correspond to the net power production of 1 MW with about 3 m

^{3}s

^{−1}of deep cold seawater (i.e., a local OTEC power intensity of 0.32 TW Sv

^{−1}, according to the global definition later used in Section 3.2, Table 1); these figures would be much better in most of the OTEC region, i.e., less seawater would produce the same amount of net power, or equivalently, pumping power losses would represent a smaller fraction of the turbo-generator output.

#### 2.2. The Ocean-Atmosphere Thermal Boundary Condition

_{obs}and T

_{obs}are climatological observations of the heat flux and of the ocean surface layer temperature T

_{0}, respectively. The form of Equation (2) is one example of the formalism initially proposed by Haney [27]. Contributions to Q involve short-wave solar radiation, incoming and outgoing longwave radiation, as well as the exchange of sensible and latent heat. By expanding heat flux components for small deviations around T

_{obs}, the restoring coefficient λ can be estimated and imparted with physical meaning in the process [28]. By contrast, we note that in the ocean-atmosphere boundary condition for salinity, a similar restoring coefficient is essentially a nudging parameter set to overcome inaccuracies in the freshwater precipitation forcing, and thus bring model predictions in satisfactory agreement with observations. λ is dominated by local sensible and latent heat transfer, and with an order of magnitude of about 50 W m

^{−2}K

^{−1}, provides a strong thermal coupling between the ocean and the atmosphere.

^{−2}K

^{−1}. Therefore, atmospheric temperatures in those large-scale scenarios essentially would adjust to those of the oceanic surface.

_{0}and the local atmospheric temperature T

_{A}. The definition of Q, in terms of all its components that are a priori functions of both T

_{0}and T

_{A}, is available as well. The two equations are linearized with respect to a reference temperature, so that T

_{0}and T

_{A}can be interpreted as deviations. The next step is to use the linearized atmospheric heat budget to solve for T

_{A}, and then substitute the result into the definition of Q. To obtain an explicit expression for T

_{A}, the operator (C

**I**−

**Q**) must be inverted, where C is a constant,

_{A}**I**is the identity and

**Q**the linear horizontal transport operator in the atmosphere. In general,

_{A}**Q**includes turbulent diffusion and advection (from the wind field); C is the sum (c + 2d), where c is a sensible and latent heat linear coupling constant at the ocean surface, with c >> d. The operator inverse (C

_{A}**I**−

**Q**)

_{A}^{−1}is approximated as ($\frac{I}{C}+\frac{{Q}_{A}}{{C}^{2}}$) on the ground that C

**I**dominates

**Q**. At this juncture, the definition of Q reads as:

_{A}_{obs}should correspond to the ocean surface-layer temperature field T

_{obs}:

**Q**is linear, subtracting Equation (4) from Equation (3) yields:

_{A}**Q**as a diffusive operator with constant coefficient k [20], i.e., k∇

_{A}^{2}, although an advective term involving the wind field could be added. ∇ is the well-known Nabla operator (∂/∂x, ∂/∂y), where partial derivatives are taken with respect to horizontal coordinates x and y (i.e., longitude and latitude). Combining k with the non-dimensional group FB/C

^{2}(≈ 0.9) as μ, but allowing this coefficient to spatially vary in the horizontal plane, Equation (5) can be written:

_{A}can be estimated from the following expression derived from Appendix A Equation (A4, second line):

## 3. Results and Discussion

#### 3.1. Implementation of the Modified Ocean-Atmosphere Boundary Condition

**Q**[20], since the approximation used to invert (C

_{A}**I**−

**Q**) would break down. They noted, however, that when running numerical models, the mesh size of the horizontal grid automatically provides a lower limit for perturbation scale. At that limiting scale, noted Δx (along a parallel) to fix ideas, they invoked a heuristic argument of continuity in thermal coupling between the Haney boundary condition, Equation (2), and the modified boundary condition, Equation (6). With λ >> γ, the approximation μ ≈ λ(Δx)

_{A}^{2}was therefore proposed. The authors noted that with a 4 degree (horizontal) resolution, this would lead to μ = 8 × 10

^{12}W K

^{−1}, or k = 9 × 10

^{12}W K

^{−1}. This diffusion coefficient can be related to the atmospheric thermal diffusivity ν

_{A}via the relationship ν

_{A}= k/(hρc

_{p}), where hρc

_{p}is the vertically averaged heat capacity of the atmosphere given as 10

^{7}J m

^{−2}K

^{−1}in Gill [29] (p. 22). Hence, Rahmstorf and Willebrand’s choice corresponds to ν

_{A}= 0.9 × 10

^{6}m

^{2}s

^{−1}, which is not significantly different from the median value of 2 × 10

^{6}m

^{2}s

^{−1}also found in Gill [29] (p. 591).

^{−1}which effectively corresponds to λ of the order of 50 W m

^{−2}K

^{−1}(the unit conversion involves the vertically averaged heat capacity of the 10 m thick upper ocean model layer, in J m

^{−2}K

^{−1}). Yet, the procedure outlined in the previous paragraph to select μ, with a horizontal mesh size four times smaller than in Rahmstorf and Willebrand [20], would result in a value corresponding to a very low atmospheric thermal diffusivity, i.e., ν

_{A}≈ 0.05 × 10

^{6}m

^{2}s

^{−1}. Therefore, a choice of μ involving grid size (squared) could lead to a substantial underestimation of diffusive effects, e.g., by a factor of up to 40 compared with Gill’s median estimate [29] (p. 591).

^{−2}K

^{−1}, and that the approximation used to invert (C

**I**−

**Q**) relies on surface latent and sensible heat coupling (C

_{A}**I**) dominating horizontal atmospheric diffusion (

**Q**), the strict mathematical validity of the new boundary condition would impose an upper limit on α in areas of sharp horizontal gradients, to ensure that CT

_{A}_{A}>>

**Q**(T

_{A}_{A}). This suggests that model simulations using Equation (6) and large values of α might be locally inaccurate where atmospheric and ocean surface temperatures exhibit strong horizontal variations. Such regions may include, for example, the vicinity of western boundary currents, even in the absence of OTEC flows, and when strong, widespread OTEC flows are considered, the edges of the OTEC zone. As long as our focus remains on large scale modeling results, which typically involve spatial integration, and more specifically, on differences between situations involving OTEC or not, a higher resolution model with potential local inaccuracies is deemed better than a coarser model with stronger internal consistency, but broadly less accurate.

**Q**. Errors may even stem from subtle discrepancies, such as the use of the field T

_{A}_{obs}as a basis for linearization to define thermal relaxation [28], instead of the constant T

_{ref}[20]. Understanding precisely why a simple replacement of Equation (2) with Equation (6) results in a less satisfactory simulated ocean, typically after a spin-up time of 2000 years, might be very difficult and time consuming.

_{obs}ab initio, such that subsequent simulations based on Equation (6) are stable and credible. We considered the heat flux from the 2001st year of MITgcm simulations using Equation (2), i.e., $Q={Q}_{obs}+\lambda ({T}_{obs}-{T}_{0}^{2001})$, and let it be equal to the flux $Q\text{}=\text{}{Q}_{obs}^{corrected}+\gamma ({T}_{obs}-{T}_{0}^{2001})-\nabla \cdot \{\mu \nabla ({T}_{obs}-{T}_{0}^{2001})\}$ representing a modified form of Equation (6). As a result, we obtained the following relationship:

_{obs}in the right-hand-side of Equation (9) include monthly ocean surface layer temperatures T

_{0}from the 2001st year of MITgcm simulations using Equation (2).

_{cw}= 20 m year

^{−1}[17]. The main difference from implementing different ocean-atmosphere boundary conditions is that the stiff thermal relaxation embodied in Equation (2), with an unchanging atmosphere, leads to a sharper buildup of the net heat input into the ocean followed by a more rapid drop until a new steady-state is reached (no net heat input, within numerical uncertainty). The modified R-W scheme essentially delays the response of the system to initial OTEC disturbances, with a weak relaxation coefficient bolstered by the additional adjustment mechanism based on horizontal atmospheric transport. The order of magnitude of the maximum transient heat input following the onset of OTEC operations is considerable in both cases, at 3 PW (1 PW = 10

^{15}W). Our MITgcm experimentation does not deal with Global Warming, since it is based on a seasonally variable steady-state background environmental configuration. Yet, it should be noted for the sake of comparison that current estimates of Earth’s energy imbalance, i.e., the cause of Global Warming, range from about 0.3 to 0.6 PW, 90% of which enter the ocean [31,32]. Clearly, the size of the peaks in Figure 2 mostly reflects the scale and impulsive nature of the selected OTEC scenario, which are unrealistic: any actual large-scale implementation of OTEC would proceed very slowly and extend very gradually. Yet, if we consider global heat addition into the ocean in terms of time-integrated values instead of the fluxes (rates) shown in Figure 2, it is clear that the effects of very large OTEC scenarios may not be trivial. More specifically, increases in the global ocean temperature when OTEC induced heat fluxes reach zero, would amount to 1.2 °C and 1.5 °C in the Haney and R-W × 40 cases, respectively.

#### 3.2. Overall OTEC Power Production

_{cw}, the volume flow rate of OTEC deep seawater per unit (latitude-longitude) area. The maxima of OTEC power for each curve occur at w

_{cw}= 20 m year

^{−1}(GLOBAL, Haney), 140 m year

^{−1}(100 KM, Haney), 15 m year

^{−1}(GLOBAL, R-W × 40), 100 m year

^{−1}(100 KM, R-W × 40), 10 m year

^{−1}(GLOBAL, R-W), and 70 m year

^{−1}(100 KM, R-W). Based on this information, a few selected characteristics of overall OTEC peak power production are shown in Table 1. The first row represents OTEC power intensity in TW Sv

^{−1}; from the perspective of a power plant engineer, the inverse of this expression indicates how many cubic meters per second of deep cold seawater would produce one megawatt of OTEC net power. The second row shows the relative OTEC power loss when a great number of plants would affect the OTEC thermal resource; it is simply the ratio of the ordinate over the abscissa in Figure 3, where a value of 100% would correspond to the red line. Looking at both rows in Table 1, differences among scenarios with different ocean-atmosphere boundary conditions and implementation areas do not appear to be significant. For those with a more practical bend, the third and fourth rows in Table 1 respectively display the number of ‘standard’ commercial OTEC plants, defined here by a cold seawater flow rate of 300 m

^{3}s

^{−1}(i.e., for an approximate power output of 100 MW with a 20 °C OTEC temperature difference), and their average spacing.

^{3}s

^{−1}MW

^{−1}), and to fix ideas, roughly 15,000 ‘standard’ plants spread 30 km apart in the 100 KM scenarios could accomplish such power output. Alternatively, Point B corresponds to moderate thermal degradation (power ratio of about 80%), for an overall power production of about 6.7 TW in most simulations (R-W cases lie a little below); here, OTEC power intensity would drop slightly below 0.40 and roughly 60,000 ‘standard’ plants spread 15 km apart in the 100 KM scenarios could generate such power output.

## 4. Large Scale OTEC Environmental Effects

#### 4.1. Ocean Temperature

_{cw}= 15 m year

^{−1}and the modified ocean-atmosphere boundary condition R-W × 40, which we label as experiment W15. This choice for the volume flow rate of OTEC deep seawater per unit (latitude-longitude) area corresponds to overall OTEC peak power among R-W × 40 simulation scenarios. The computed temperature difference δT between W15 and CTL is referred to as temperature anomaly.

_{θ}from CTL (contours). σ

_{θ}is defined as the seawater potential density (referenced to sea surface pressure) minus 1000 in units of kg m

^{−3}. Striking features in Figure 4 are the cooling within the equatorial region and warming in the high latitudes. It turns out that the equatorial cooling begins within one year of OTEC activity. The high latitude warming emerges as a weak signal south of Australia and against the eastern and northern land boundaries of the Pacific and Atlantic basins within 10 years, and enhances with time.

_{θ}from CTL). This can be understood as follows: the volume of OTEC effluent of a certain density, discharged at a depth of neutral buoyancy, would expand water of that density relative to the reference state (CTL), thus displacing potential density surfaces and resulting in positive (denser) and negative (lighter) potential density anomalies above and below the discharge depth, respectively. Since potential density is a function of temperature and salinity, this would also mean that anomalies of opposite signs across the discharge depth would also emerge in the temperature and salinity fields.

_{θ}= 26.5 for the dynamical component (upper panel) and σ

_{θ}= 25.0 for the spiciness component (lower panel). These are the surfaces where the components have their respective maxima along the Equator in the Pacific sector, positive for the dynamical anomaly and negative for the spiciness anomaly (middle and lower panels in Figure 6). The snapshots in the sequence correspond to simulation times representing 1, 3, 5, 7 and 10 years after the onset of OTEC deployment, respectively. Arrows represent the ocean circulation in CTL. In Figure 7a, merely one year after the start of OTEC operations, the warm dynamical anomaly is confined within the OTEC zone. By Year 10 (Figure 7e, upper panel), its magnitude has gradually increased and the region of anomaly greatly expanded. In particular, there is clear indication of its poleward extension along the eastern boundary in each basin, opposite to the direction of flow. This represents the action of Kelvin waves. Within the equatorial wave guide (in a latitude range approximately 4° wide across the Equator), equatorial Kelvin waves propagate eastward, spreading the dynamical component to the eastern boundary, where they split into poleward-travelling coastal Kelvin waves to reach the coast of Alaska and the tip of South America in the Pacific Ocean, the southern coast of Australia in the Indian Ocean, and South Africa in the Atlantic. In the high latitudes of the North Atlantic where the surface mixed layer water is heavier than this density surface (light grey, see also the contours in Figure 4), the warm dynamical anomaly reaches the high latitudes along denser surfaces (e.g., σ

_{θ}= 27.5, not shown). The signatures of Rossby waves can also be seen in this figure. Rossby waves are reflected from the coastal Kelvin waves to propagate due west. Since the phase speeds of Rossby waves are fastest in the tropical latitudes and decrease sharply towards the high latitudes, the westward extension of the warm anomaly narrows sharply poleward, which is seen most clearly in the Pacific Ocean. Within the OTEC zone and outside the equatorial wave guide, OTEC activity can also generate Rossby waves that carry temperature anomalies to the western boundary, where they turn equatorward as coastal Kelvin waves, and then eastward as equatorial Kelvin waves which then split into poleward coastal Kelvin waves at the eastern boundary. Rossby and Kelvin waves would propagate cool dynamical anomalies on lighter potential density surfaces in the same manner as for the warm anomalies on denser potential density surfaces. The main difference is that lighter σ

_{θ}surfaces rise to intersect the ocean surface at lower latitudes, bringing a cool anomaly to the tropical region, while the denser σ

_{θ}surfaces affect the ocean surface temperature at higher latitudes. This explains the temperature anomaly distribution shown in Figure 4 at the ocean surface.

#### 4.2. Ocean Circulation

_{cw}= 15 m year

^{−1}at maximum OTEC power production), and covers the entire so-called OTEC region (GLOBAL). Therefore, specific effects directly resulting from OTEC activity would be superimposed on the existing circulation.

^{6}km

^{2}for the Atlantic versus 95 × 10

^{6}km

^{2}for the Indo-Pacific; therefore, the single value of the OTEC cold seawater flow rate per horizontal area used here, w

_{cw}= 15 m year

^{−1}, corresponds to relatively much larger overall OTEC flow rates in the Indo-Pacific. This might skew or complicate a priori a comparison between basins, and simulations with two different values of w

_{cw}might be considered in the future. When separate OTEC power production maxima were determined for the two regions [18], however, with a value of w

_{cw}twice as large in the Atlantic as in the Indo-Pacific, reported environmental effects did not qualitatively contradict the foregoing results [18] (Figure 5 and Figure 6).

## 5. Conclusions

^{−1}(with deep cold seawater being the flow reference); in other words, and in an average sense across the OTEC region, 5 m

^{3}s

^{−1}of deep cold seawater would produce 1 MW of net OTEC power. As noted in an earlier study that imposed geographic limitations on the OTEC region [18], the approximate invariance of OTEC power intensity at peak OTEC power production suggests that the degradation of the thermal resource under such conditions is relatively similar across the OTEC implementation region in all cases; this can be inferred from Equation (1). It also means that the ratio of actual peak OTEC power over nominal power does not change significantly among simulations, from 40% to below 50%. OTEC power curves (Figure 3) also showed that an overall OTEC power production of about 2 TW would not have large-scale environmental effects, and that 6 to 7 TW might be produced provided that associated effects remain acceptable.

## Acknowledgments

## Author Contributions

## Conflicts of Interest

## Appendix A

_{U}, Q

_{D}and Q

_{B}) are shown here as red arrows, and the incoming shortwave solar radiation flux Q

_{S}is depicted in yellow. Other fluxes involve latent and sensible heat exchange at the ocean-atmosphere interface (Q

_{C}) and horizontal atmospheric heat transport (Q

_{A}). e and f are fractions representing heat absorption in the atmosphere.

**Figure A1.**Local atmospheric heat fluxes considered in Rahmstorf and Willebrand [20].

_{C}is formally linear from the outset, and equal to c(T

_{0}− T

_{A}). The longwave radiation fluxes are linearized for temperatures around T

_{ref}= 273 K, and written as:

## References

- D’Arsonval, A. Utilisation des forces naturelles, avenir de l’électricité. Rev. Sci.
**1881**, 17, 370–372. (In French) [Google Scholar] - Claude, G. Power from the tropical seas. Mech. Eng.
**1930**, 52, 1039–1044. [Google Scholar] - Avery, W.H.; Wu, C. Renewable Energy from the Ocean; Oxford University Press: New York, NY, USA, 1994; p. 446. ISBN 0195071999. [Google Scholar]
- Vega, L.A. Ocean Thermal Energy Conversion. In Encyclopedia of Sustainability Science and Technology; Meyers, R.A., Ed.; Springer: New York, NY, USA, 2012; pp. 7296–7328. ISBN 9780387894690. [Google Scholar]
- Nihous, G.C. Ocean Thermal Energy Conversion (OTEC). In The World Scientific Handbook of Energy; Crawley, G., Ed.; World Scientific: Singapore, 2013; pp. 359–372. ISBN 9789814343510. [Google Scholar]
- Nihous, G.C.; Gauthier, M. Ocean Thermal Energy Conversion: A historical perspective. In Marine Renewable Energy Handbook; Multon, B., Ed.; Wiley: New York, NY, USA, 2011; pp. 367–401. ISBN 9781848213326. [Google Scholar]
- Vega, L.A. Economics of Ocean Thermal Energy Conversion (OTEC): An Update. In Proceedings of the Offshore Technology Conference, Houston, TX, USA, 3–6 May 2010; Paper OTC 21016. p. 18. [Google Scholar]
- Loster, M. Solar Land Area. Wikimedia Commons Image File. 2006. Available online: https://commons.wikimedia.org/wiki/File:Solar_land_area.png (accessed on 18 September 2010).
- International Energy Agency. Key World Energy Statistics; IEA Publications: Paris, France, 2016; p. 80. ISBN 9789264266520. Available online: https://www.iea.org/publications/freepublications/publication/KeyWorld2016.pdf (accessed on 26 July 2017).
- Da Silva, A.; Young, A.C.; Levitus, S. Atlas of Surface Marine Data 1994 Vol. 1: Algorithms and Procedures, NOAA Atlas NESDIS 6; U.S. Government Printing Office: Washington, DC, USA, 1994.
- Cousteau, J.Y.; Jacquier, H. Énergie des mers: Plan-plan les watts. In Français, on a Volé ta Mer; R. Laffont: Paris, France, 1981; p. 192. ISBN 2221007654. (In French) [Google Scholar]
- Nihous, G.C. An order-of-magnitude estimate of Ocean Thermal Energy Conversion resources. J. Energy Resour. Technol.
**2005**, 127, 328–333. [Google Scholar] [CrossRef] - Nihous, G.C. A preliminary assessment of Ocean Thermal Energy Conversion (OTEC) resources. J. Energy Resour. Technol.
**2007**, 129, 10–17. [Google Scholar] [CrossRef] - Nihous, G.C. An estimate of Atlantic Ocean Thermal Energy Conversion (OTEC) resources. Ocean Eng.
**2007**, 34, 2210–2221. [Google Scholar] [CrossRef] - Marshall, J.; Adcroft, A.; Hill, C.; Perelman, L.; Heisey, C. A finite-volume, incompressible Navier-Stokes model for studies of the ocean on parallel computers. J. Geophys. Res.
**1997**, 102, 5753–5766. [Google Scholar] [CrossRef] - Rajagopalan, K.; Nihous, G.C. Estimates of global Ocean Thermal Energy Conversion (OTEC) resources using an Ocean General Circulation Model. Renew. Energy
**2013**, 50, 532–540. [Google Scholar] [CrossRef] - Rajagopalan, K.; Nihous, G.C. An assessment of global Ocean Thermal Energy Conversion (OTEC) resources with a high-resolution Ocean General Circulation Model. J. Energy Resour. Technol.
**2013**, 135, 041202. [Google Scholar] [CrossRef] - Rajagopalan, K.; Nihous, G.C. An assessment of global Ocean Thermal Energy Conversion resources under broad geographical constraints. J. Renew. Sustain. Energy
**2013**, 5, 063124. [Google Scholar] [CrossRef] - Kwiatkowski, L.; Ricke, K.L.; Caldeira, K. Atmospheric consequences of disruption of the ocean thermocline. Environ. Res. Lett.
**2015**, 10, 034016. [Google Scholar] [CrossRef] - Rahmstorf, S.; Willebrand, J. The role of temperature feedback in stabilizing the thermohaline circulation. J. Phys. Oceanogr.
**1995**, 25, 787–805. [Google Scholar] [CrossRef] - Adcroft, A.; Campin, J.-M.; Dutkiewicz, S.; Evangelinos, C.; Ferreira, D.; Forget, G.; Fox-Kemper, B.; Heimbach, P.; Hill, C.; Hill, E.; et al. MITgcm User Manual. Available online: http://mitgcm.org/public/r2_manual/latest/online_documents/manual.pdf (accessed on 14 July 2016).
- Forget, G. Mapping ocean observations in a dynamical framework: A 2004–2006 ocean atlas. J. Phys. Oceanogr.
**2010**, 40, 1201–1221. [Google Scholar] [CrossRef] - Large, W.C.; McWilliams, J.C.; Doney, S.C. Oceanic vertical mixing: A review and a model with a nonlocal boundary layer parameterization. Rev. Geophys.
**1994**, 32, 363–403. [Google Scholar] [CrossRef] - Redi, M.H. Oceanic isopycnal mixing by coordinate rotation. J. Phys. Oceanogr.
**1982**, 12, 1154–1158. [Google Scholar] [CrossRef] - Gent, P.R.; McWilliams, J.C. Isopycnal mixing in ocean circulation models. J. Phys. Oceanogr.
**1990**, 20, 150–155. [Google Scholar] [CrossRef] - Rajagopalan, K.; Nihous, G.C. Predictions of water-column properties under widespread artificial upwelling scenarios in the North Pacific Subtropical Gyre using an ocean general circulation model. J. Mar. Environ. Eng.
**2014**, 9, 285–304. [Google Scholar] - Haney, R.L. Surface thermal boundary condition for ocean circulation models. J. Phys. Oceanogr.
**1971**, 1, 241–248. [Google Scholar] [CrossRef] - Barnier, B.; Siefridt, L.; Marchesiello, P. Thermal forcing for a global ocean circulation model using a three-year climatology of ECMWF analysis. J. Mar. Syst.
**1995**, 6, 363–380. [Google Scholar] [CrossRef] - Gill, A.E. Atmosphere-Ocean Dynamics; Academic Press: San Diego, CA, USA, 1982; p. 662. ISBN 0122835220. [Google Scholar]
- Griffies, S.M.; Biastoch, A.; Böning, C.; Bryan, F.; Danabasoglu, G.; Chassignet, E.P.; England, M.H.; Gerdes, R.; Haak, H.; Hallberg, R.W.; et al. Coordinated Ocean-Ice Reference Experiments (COREs). Ocean Model.
**2009**, 26, 1–46. [Google Scholar] [CrossRef] - Hansen, J.; Sato, M.; Kharecha, P.; von Schuckmann, K. Earth’s energy imbalance and implications. Atmos. Chem. Phys.
**2011**, 11, 13421–13449. [Google Scholar] [CrossRef] [Green Version] - Trenberth, K.E.; Fasullo, J.T.; Balmaseda, M.A. Earth’s energy imbalance. J. Clim.
**2014**, 27, 3129–3144. [Google Scholar] [CrossRef] - Furue, R.; Jia, Y.; McCreary, J.P.; Schneider, N.; Richards, K.J.; Müller, P.; Cornuelle, B.D.; Martínez Avellaneda, N.; Stammer, D.; Liu, C.; et al. Impacts of regional mixing on the temperature structure of the equatorial Pacific Ocean. Part 1: Vertically uniform vertical diffusion. Ocean Model.
**2015**, 91, 91–111. [Google Scholar] [CrossRef] - Cunningham, S.A.; Kanzow, T.; Rayner, D.; Baringer, M.O.; Johns, W.E.; Marotzke, J. Temporal variability of the Atlantic Meridional Overturning Circulation at 26.5° N. Science
**2007**, 317, 935–938. [Google Scholar] [CrossRef] [PubMed] - Danabasoglu, G.; Yeager, S.G.; Bailey, D.; Behrens, E.; Bentsen, M.; Bi, D.; Biastoch, A.; Böning, C.; Bozec, A.; Canuto, V.M.; et al. North Atlantic simulations in Coordinated Ocean-ice Reference Experiments phase II (CORE-II). Part I: Mean states. Ocean Model.
**2014**, 73, 76–107. [Google Scholar] [CrossRef] [Green Version] - McCreary, J.P.; Lu, P. Interaction between the subtropical and equatorial ocean circulations: The Subtropical Cell. J. Phys. Oceanogr.
**1994**, 24, 466–497. [Google Scholar] - Mazloff, M.; Heimbach, P.; Wunsch, C. An eddy-permitting Southern Ocean state estimates. J. Phys. Oceanogr.
**2010**, 40, 880–899. [Google Scholar] [CrossRef] - Talley, L. Freshwater transport estimates and the global overturning circulation: Shallow, deep and throughflow components. Prog. Oceanogr.
**2008**, 78, 257–303. [Google Scholar] [CrossRef] - Farneti, R.; Downes, S.M.; Griffies, S.M.; Marsland, S.J.; Behrens, E.; Bentsen, M.; Bi, D.; Biastoch, A.; Böning, C.; Bozec, A.; et al. An assessment of Antarctic Circumpolar Current and Southern Ocean meridional overturning circulation during1958–2007 in a suite of interannual CORE-II simulations. Ocean Model.
**2015**, 93, 84–120. [Google Scholar] [CrossRef] - Marshall, J.; Speer, K. Closure of the meridional overturning circulation through Southern Ocean upwelling. Nat. Geosci.
**2012**, 5, 171–180. [Google Scholar] [CrossRef]

**Figure 1.**Time history of global ocean temperature from MITgcm with different ocean-atmosphere thermal boundary conditions.

**Figure 2.**Time history of the additional net heat flux into the ocean when implementing a global Ocean Thermal Energy Conversions (OTEC) scenario (w

_{cw}= 20 m year

^{−1}) in MITgcm with different ocean-atmosphere thermal boundary conditions.

**Figure 3.**Long-term OTEC power production as a function of nominal OTEC power for different implementation scenarios and ocean-atmosphere thermal boundary conditions.

**Figure 4.**Temperature anomaly δT (°C) between the control (CTL) and an OTEC perturbation (W15) simulations in the model’s uppermost layer (mid-depth of 5 m) 10 years (

**upper**) and 1000 years (

**lower**) after OTEC activation. Note that the color scale increment is 0.2 °C between −1.0 °C and 1.0 °C to bring out the weak warming in the high latitudes after 10 years, and 1.0 °C otherwise. Also shown are σ

_{θ}contours from CTL.

**Figure 5.**Temperature anomaly δT (°C) between CTL and W15 simulations at latitude 0.5° N after 1 year of OTEC activity (

**upper**); components of δT (determined with the method of Furue et al. [33]) are also shown: dynamical (

**middle**) and spiciness (

**lower**). Color scale is the same as in Figure 4. The light grey shading in the middle and lower panels indicates areas where the components are undefined; this is because the decomposition requires matching density surfaces between W15 and CTL, and the near-surface cooling from OTEC activity has resulted in the absence of light density water in W15. Also shown are σ

_{θ}contours from CTL.

**Figure 6.**Same as Figure 5 except that it is after 10 years of OTEC activity, and that the color scale has a 0.5 °C constant increment.

**Figure 7.**(

**a**) Components of temperature anomaly δT (°C) between CTL and W15 simulations after 1 year of OTEC activity (determined with the method of Furue et al. [33]): dynamical along CTL isopycnal σ

_{θ}= 26.5 kg m

^{−3}(upper panel); spiciness along CTL isopycnal σ

_{θ}= 25 kg m

^{−3}(lower panel). The light grey shading indicates regions where the density surfaces do not exist. Also shown are velocity vectors from CTL. (

**b**) Same as Figure 7a after 3 years of OTEC activity. (

**c**) Same as Figure 7a after 5 years of OTEC activity. (

**d**) Same as Figure 7a after 7 years of OTEC activity. (

**e**) Same as Figure 7a after 10 years of OTEC activity.

**Figure 8.**Average streamfunctions (Sv) from Equation (10) in the Atlantic (

**upper**) and the Indo-Pacific (

**lower**) with OTEC (W15,

**right**), or without (CTL,

**left**).

**Figure 9.**Vertical velocity (m day

^{−1}) averaged over a 4° wide band across the Equator, with OTEC (W15,

**lower**) or without (CTL,

**upper**).

**Figure 10.**March and September averaged potential density profiles with or without OTEC at selected northern high-latitude locations: Pacific Ocean (

**left**) and Atlantic Ocean (

**right**).

**Figure 11.**Average streamfunctions (Sv) from Equation (10) in the Atlantic (

**upper**) and in the Indo-Pacific (

**lower**); OTEC is active in the Atlantic only (

**left**), or in the Indo-Pacific only (

**right**).

Quantity | GLOBAL Haney | GLOBAL R-W | GLOBAL R-W × 40 | 100 KM Haney | 100 KM R-W | 100 KM R-W × 40 |
---|---|---|---|---|---|---|

OTEC power intensity ^{a} (TW Sv^{−1}) | 0.19 | 0.22 | 0.18 | 0.19 | 0.23 | 0.21 |

Power ratio ^{b} (%) | 43 | 49 | 41 | 40 | 42 | 44 |

Number of 300 m^{3} s^{−1} plants ^{c} (thousands) | 245 | 123 | 184 | 207 | 104 | 148 |

Plant-to-plant spacing (km) | 22 | 31 | 25 | 8 | 12 | 10 |

^{a}OTEC power divided by OTEC cold seawater flow rate (w

_{cw}times OTEC area).

^{b}Ratio of actual OTEC net power over nominal OTEC net power.

^{c}OTEC cold seawater flow rate divided by 300 (value for ‘standard’ 100 MW plant).

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Jia, Y.; Nihous, G.C.; Rajagopalan, K.
An Evaluation of the Large-Scale Implementation of Ocean Thermal Energy Conversion (OTEC) Using an Ocean General Circulation Model with Low-Complexity Atmospheric Feedback Effects. *J. Mar. Sci. Eng.* **2018**, *6*, 12.
https://doi.org/10.3390/jmse6010012

**AMA Style**

Jia Y, Nihous GC, Rajagopalan K.
An Evaluation of the Large-Scale Implementation of Ocean Thermal Energy Conversion (OTEC) Using an Ocean General Circulation Model with Low-Complexity Atmospheric Feedback Effects. *Journal of Marine Science and Engineering*. 2018; 6(1):12.
https://doi.org/10.3390/jmse6010012

**Chicago/Turabian Style**

Jia, Yanli, Gérard C. Nihous, and Krishnakumar Rajagopalan.
2018. "An Evaluation of the Large-Scale Implementation of Ocean Thermal Energy Conversion (OTEC) Using an Ocean General Circulation Model with Low-Complexity Atmospheric Feedback Effects" *Journal of Marine Science and Engineering* 6, no. 1: 12.
https://doi.org/10.3390/jmse6010012