Renewable energy system controls

Abstract

Physical and/or financial instruments may optimally hedge the cash flow of one or more renewable energy generators based on a desired risk and return profile of renewable infrastructure investors. Baseline revenues may be determined based on forward-looking electricity market price scenarios corresponding to qualified market products intended for sale from the renewable energy generators. Risk and return metrics of cash flows of the renewable energy generators may be determined. At least one physical hedge and/or financial hedge may be added. The size and operation of the renewable energy generators along with any physical hedges, or financial hedges, or both physical and financial hedges, may be optimized across multiple market price scenarios of qualified market products to optimize investor-tailored risk and return utility functions.

Claims

1. A method to manage operation of a renewable power asset, comprising: operating one or more computing devices controlling or in communication with one or more control devices of a renewable power asset; identifying, by the computing device, one or more market products to be sold by the renewable power asset; generating, by the one or more computing devices, forward looking market price scenarios representative of the one or more market products; calculating, by the one or more computing devices, baseline revenues of the renewable power asset in the forward looking market price scenarios; determining, by the one or more computing devices, a first cash flow associated with the baseline revenues; determining, by the one or more computing devices, a second cash flow by adding at least one physical hedge into the first cash flow, wherein each physical hedge comprises a long duration energy storage system or a short duration energy storage system, and each physical hedge has attributes of a maximum rated power, a maximum rated stored energy, a round-trip efficiency, a self-discharge, a ramp-rate, a response time, a calendar degradation, a throughput degradation, a capital cost, an operating cost, and a decommissioning cost; optimizing, by the one or more computing devices, a size of the at least one physical hedge and an operational strategy of the renewable power asset; sizing the at least one physical hedge at the optimized size, and charging and discharging the at least one physical hedge based on the optimized operational strategy. controlling, by the one or more computing devices or the one or more control devices in communication with the computing device, the renewable power asset to generate and deliver power based on the optimized operational strategy; and wherein optimizing further comprises optimizing based on a mean-variance optimization of risk and return metrics using $\underset{q_X,d_{RT},c_{\mathrm{RT}},P_{\mathrm{ESS}},E_{ESS}}{\max }\left\{{\mu }_{\tilde{\vartheta }}-\lambda \left(\underset{\stackrel{0\le m\le M}{0\le k\le K}}{.Math.}{{\gamma }_m\left(\right)}^2-{\mu }^{\stackrel{2}{\stackrel{~}{\vartheta }}}\right)\right\}$ where: q.sub.X is a real time curtailment of renewable generation, d.sub.RT is a real time optimal discharge schedule of the at least one physical hedge, c.sub.RT is a real time optimal charge schedule of the at least one physical hedge, P.sub.ESS is a rated power of the at least one physical hedge, E.sub.ESS is a rated energy of the at least one physical hedge, q.sub.RT is a real time renewable generation, q.sub.RT,m is a real time renewable generation forecast in an m.sup.th scenario, q.sub.DA is a day ahead commitment, q.sub.X is a real time curtailment of renewable generation, p.sub.RT is a real time price, p.sub.RT,k is a real time price forecast in a k.sup.th scenario, δ.sub.RT,m,k is a real time penalty, c.sub.P is a storage unit power cost amortized in an optimization horizon, c.sub.E is a storage unit energy cost amortized in the optimization horizon, γ.sub.m is a factor biasing optimization towards specific generation forecast scenarios, λ is a risk tolerance factor, is a joint probability of (q.sub.RT=q.sub.RT,m; p.sub.RT=p.sub.RT,k) given a realized DA price, M is a number of possible renewable generation scenarios, indexed by 1≤m≤M, K is a number of possible RT price scenarios, indexed by 1≤k≤K, μ.sub.{tilde over (ϑ)} is defined as $\underset{{0\le k\le K}_{}}{\underset{0\le m\le M}{.Math.}}{\gamma }_m,$ and is defined as (q.sub.RT,md.sub.RT−c.sub.RT−q.sub.Da−q.sub.X).sup.T.Math.p.sub.RT,k−δ.sub.RT,m,k−c.sub.PP.sub.ESS−c.sub.EE.sub.ESS.

2. The method of claim 1, further comprising holding the optimal size of a physical hedge as fixed at a size of an existing physical hedge associated with the renewable generator asset.

3. A method to manage operation of a renewable power asset, comprising: operating one or more computing devices controlling or in communication with one or more control devices of a renewable power asset; identifying, by the computing device, one or more market products to be sold by the renewable power asset; generating, by the one or more computing devices, forward looking market price scenarios representative of the one or more market products; calculating, by the one or more computing devices, baseline revenues of the renewable power asset in the forward looking market price scenarios; determining, by the one or more computing devices, a first cash flow associated with the baseline revenues; determining, by the one or more computing devices, a second cash flow by adding at least one physical hedge into the first cash flow, wherein each physical hedge comprises a long duration energy storage system or a short duration energy storage system, and each physical hedge has attributes of a maximum rated power, a maximum rated stored energy, a round-trip efficiency, a self-discharge, a ramp-rate, a response time, a calendar degradation, a throughput degradation, a capital cost, an operating cost, and a decommissioning cost; optimizing, by the one or more computing devices, a size of the at least one physical hedge and an operational strategy of the renewable power asset; sizing the at least one physical hedge at the optimized size, and charging and discharging the at least one physical hedge based on the optimized operational strategy. controlling, by the one or more computing devices or the one or more control devices in communication with the computing device, the renewable power asset to generate and deliver power based on the optimized operational strategy; and wherein optimizing further comprises: optimizing through a linear programming framework based on utility functions of risk and return metrics such that
=(q.sub.RT,m+d.sub.RT−c.sub.RT−q.sub.DA−q.sub.X).sup.T.Math.p.sub.RT,k−c.sub.PP.sub.ESSx.sub.EE.sub.ESS=d.sub.RT.sup.T.Math.p.sub.RT,k−c.sub.RT.sup.T.Math.p.sub.RT,k−q.sub.X.sup.T.Math.p.sub.RT,k−c.sub.PP.sub.ESS−c.sub.EE.sub.ESS+(q.sub.RT,m−q.sub.DA).sup.T.Math.p.sub.RT,k where q.sub.RT,m is a real time renewable generation forecast in an m.sup.th scenario, d.sub.RT is a real time optimal discharge schedule of the at least one physical hedge, c.sub.RT is a real time optimal charge schedule of the at least one physical hedge, q.sub.DA is a day ahead commitments, q.sub.X is a real time curtailment of renewable generation, p.sub.RT,k is a real time price forecast in a k.sup.th scenario, c.sub.P is a storage unit power cost amortized in an optimization horizon, P.sub.ESS is a rated power of the at least one physical hedge, c.sub.E is a storage unit energy cost amortized in the optimization horizon, E.sub.ESS is a rated energy of the at least one physical hedge, c.sub.RT is a real time optimal charge schedule of the at least one physical hedge, M is a number of possible renewable generation scenarios, indexed by 1≤m≤M, K is a number of possible RT price scenarios, indexed by 1≤k≤K, and J is a number of discretized time steps in the optimization horizon, indexed by 1≤j≤J; introducing an auxiliary optimization variable x and a unit returns r.sub.m,k such that $x\left[\begin{array}{c}{q}_X\\ d_{RT}\\ c_{RT}\\ P_{\mathrm{ESS}}\\ E_{\mathrm{ESS}}\\ x_{3J+1}\end{array}\right]$ and $r_{m,k}\left[\begin{array}{c}-p_{\mathrm{RT},k}\\ p_{\mathrm{RT},k}\\ -p_{\mathrm{RT},k}\\ -c_P\\ -c_E\\ {\left(q_{\mathrm{RT},m}-q_{\mathrm{DA}}\right)}^T.Math.p_{\mathrm{RT},k}\end{array}\right],$ and =x.sup.T.Math.r.sub.m,k; x.sub.3J+1=1; indexing scenarios to a variable t within a range of total scenario permutations, and expanding to components of x, r.sub.t such that {tilde over (ϑ)}.sub.t=x.sup.T.Math.r.sub.t=Σ.sub.1≤j≤3J+1x.sub.jr.sub.j,t; x.sub.3J+1=1; introducing an average return μ.sub.j at a given time step such that μ.sub.jΣ.sub.1≤t≤Tw.sub.tr.sub.j,t; and configuring a linear programming goal and applying one of: a mean absolute deviation (MAD), a minimax, a conditional value risk (CVaR), a Gini mean difference (GMD), or a weighted conditional value at risk (WCVaR).

4. The method of claim 3, wherein configuring a linear programming goal further comprises configuring the goal to optimize subject to a minimum return constraint as expressed in $\underset{x}{\max }\left\{\mu \left(x\right)-\rho \left(x\right):\mu \left(x\right)\ge {\mu }_0\right\}$ where ρ(x) is defined as Dispersion; and applying one of the MAD, the minimax, the CVaR, the GMD, or the WCVaR further comprises applying one of: the MAD as expressed in $\underset{x,d^{-}}{\max }\left\{\underset{j}{.Math.}{x}_j{\mu }_j-\underset{t}{.Math.}{d}_t^{-}{w}_t:d_t^{-}\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right)x_j;d_t^{-}\ge 0;\forall t=1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\right\},$ the minimax as expressed in $\underset{x,v}{\max }\left\{\underset{j}{.Math.}{x}_j{\mu }_j-v:v\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right)x_j;\forall t=1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\right\},$ the CVaR as expressed in $\underset{x,d^{-},\eta }{\max }\left\{\eta -\frac{1}{\beta }\underset{t}{.Math.}{d}_t^{-}{w}_t:d_t^{-}\ge \eta -\underset{j}{.Math.}{r}_{j,t}{x}_j;d_t^{-}\ge 0;\forall t=1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\right\}$ where η is defined as VaR at an optimum, and β is defined as a probability of returns ≤VaR, with 0 ≤β≤1, the GMD as expressed in $\underset{x,u}{\max }\left\{\begin{array}{c}\left(\underset{\begin{array}{c}t\\ j\end{array}}{.Math.}{r}_{j,t}{x}_j{w}_t^2+2\underset{\begin{array}{c}1\le t^{\prime }\le T-1\\ {t^{\prime }+1\le t^{″}\le T}^{}\end{array}}{.Math.}{u}_{t^{\prime },t^{″}}{w}_{t^{\prime }}{w}_{t^{″}}\right):\\ u_{t^{\prime },t^{″}}\le \underset{j}{.Math.}{r}_{j,t^{\prime }}{x}_j;u_{t^{\prime },t^{″}}\le \underset{j}{.Math.}{r}_{j,t^{″}}{x}_j;\forall t^{\prime }=1.Math.T-1;\\ t^{″}=t^{\prime }+1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\end{array}\right\},$ or the WCVaR as expressed in $\underset{x,d^{-}}{\max }\left\{\begin{array}{c}\underset{h}{.Math.}{\omega }_h\left({\eta }_h-\frac{1}{{\beta }_h}\underset{t}{.Math.}{d}_{h,t}^{-}{w}_t\right):\\ d_{h,t}^{-}\ge {\eta }_h-\underset{j}{.Math.}{r}_{j,t}{x}_j;d_{h,t}^{-}\ge 0;\forall t=1.Math.T;\\ {\omega }_h\in \mathrm{weights};\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\end{array}\right\}$ where: η is defined as VaR at an optimum, and β is defined as a probability of returns ≤VaR, with 0<β≤1.

5. The method of claim 3, wherein configuring a linear programming goal further comprises configuring the goal to optimize based on minimizing dispersion while subject to a minimum return constraint as expressed in $\underset{x}{\min }\left\{\rho \left(x\right):\mu \left(x\right)\ge {\mu }_0\right\}$ where ρ(x) is defined as Dispersion; and applying one of the MAD, the minimax, the CVaR, the GMD, or the WCVaR further comprises applying one of: the MAD as expressed in $\underset{x,d^{-}}{\min }\left\{\begin{array}{c}\underset{t}{.Math.}{d}_t^{-}{w}_t:d_t^{-}\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right)x_j;d_t^{-}\ge 0;\\ \forall t=1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\end{array}\right\},$ the minimax as expressed in $\underset{x,v}{\min }\left\{v:v\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right)x_j;\forall t=1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\right\},$ the CVaR as expressed in $\underset{x,d^{-},\eta }{\min }\left\{\begin{array}{c}\underset{j}{.Math.}{x}_j{\mu }_j-\eta +\frac{1}{\beta }\underset{t}{.Math.}{d}_t^{-}{w}_t:\\ d_t^{-}\ge \eta -\underset{j}{.Math.}{r}_{j,t}{x}_j;d_t^{-}\ge 0;\forall t=1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\end{array}\right\}$ where: η is defined as VaR at an optimum, and β is defined as a probability of returns≤VaR, with 0<β≤1, the GMD as expressed in $\underset{x,u}{\min }\left\{\begin{array}{c}\left(\underset{j}{.Math.}{x}_j{\mu }_j-\underset{\begin{array}{c}t\\ j\end{array}}{.Math.}{r}_{j,t}{x}_j{w}_t^2-2\underset{\begin{array}{c}1\le t^{\prime }\le T-1\\ {t^{\prime }+1\le t^{″}\le T}^{}\end{array}}{.Math.}{u}_{t^{\prime },t^{″}}{w}_{t^{\prime }}{w}_{t^{″}}\right):\\ u_{t^{\prime },t^{″}}\le \underset{j}{.Math.}{r}_{j,t^{\prime }}{x}_j;u_{t^{\prime },t^{″}}\le \underset{j}{.Math.}{r}_{j,t^{″}}{x}_j;\forall t^{\prime }=1.Math.T-1;\\ t^{″}=t^{\prime }+1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\end{array}\right\},$ or the WCVaR as expressed in $\underset{x,d^{-}}{\min }\left\{\begin{array}{c}\underset{j}{.Math.}{x}_j{\mu }_j-\underset{h}{.Math.}{\omega }_h\left({\eta }_h-\frac{1}{{\beta }_h}\underset{t}{.Math.}{d}_{h,t}^{-}{w}_t\right):\\ d_{h,t}^{-}\ge {\eta }_h-\underset{j}{.Math.}{r}_{j,t}{x}_j;d_{h,t}^{-}\ge 0;\forall t=1.Math.T;\\ {\omega }_h\in \mathrm{weights};\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\end{array}\right\}$ where: η is defined as VaR at an optimum, and β is defined as a probability of returns ≤VaR, with 0<β≤1.

6. The method of claim 3, wherein configuring a linear programming goal further comprises configuring the goal to optimize based relative to a no risk scenario as expressed in $\underset{x}{\max }\left\{\left(\mu \left(x\right)-r_0\right)/\rho \left(x\right)\right\}$ where r.sub.0 is defined as a risk free return and ρ(x) is defined as Dispersion; and applying one of the MAD, the minimax, the CVaR, the GMD, or the WCVaR further comprises applying one of: the MAD as expressed in $\underset{\tilde{x},\widetilde{d^{-}},z}{\max }\left\{\begin{array}{c}\underset{j}{.Math.}{\tilde{x}}_j{\mu }_j-r_{0}z:\underset{t}{.Math.}\widetilde{d_t^{-}}{w}_t=z;\\ \widetilde{d_t^{-}}\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right){\tilde{x}}_j;\\ \widetilde{d_t^{-}}\ge 0;\forall t=1.Math.T;\underset{j}{.Math.}{\tilde{x}}_j=z;{\tilde{x}}_j\ge 0\end{array}\right\}.Math.x=\tilde{x}/z,$ the minimax as expressed in $\underset{\tilde{x},\tilde{v}}{\max }\left\{\begin{array}{c}\underset{j}{.Math.}{\tilde{x}}_j{\mu }_j-r_{0}z:\\ \tilde{v}=z;\tilde{v}\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right){\tilde{x}}_j;\forall t=1.Math.T;\\ \underset{j}{.Math.}{\tilde{x}}_j=z;{\tilde{x}}_j\ge 0\end{array}\right\}.Math.x=\tilde{x}/z,$ or the CVaR as expressed in $\underset{\tilde{x},\widetilde{d^{-}},z,\eta }{\max }\left\{\begin{array}{c}\underset{j}{.Math.}{\tilde{x}}_j{\mu }_j-r_{0}z:\underset{j}{.Math.}{\tilde{x}}_j{\mu }_j-\eta +\frac{1}{\beta }\underset{t}{.Math.}\widetilde{d_t^{-}}{w}_t=z;\\ \widetilde{d_t^{-}}\ge \eta -\underset{j}{.Math.}{r}_{j,t}{\tilde{x}}_j;\widetilde{d_t^{-}}\ge 0;\forall t=1.Math.T;\\ \underset{j}{.Math.}{\tilde{x}}_j=z;{\tilde{x}}_j\ge 0\end{array}\right\}.Math.x=\tilde{x}/z$ where ƒ is defined as VaR at an optimum, and β is defined as a probability of returns≤VaR, with 0 <β≤1.

7. The method of claim 3, wherein configuring a linear programming goal further comprises configuring the goal expressed in $\underset{x}{\max }\left\{\mu \left(x\right)-\mathrm{\lambda \rho }\left(x\right)\right\}$ where:
λ is a risk tolerance factor, and
ρ(x) is defined as Dispersion; and applying one of the MAD, the minimax, the CVaR, the GMD, or the WCVaR further comprises applying one of: the MAD as expressed in $\underset{x,d^{-}}{\max }\left\{\underset{j}{.Math.}{x}_j{\mu }_j-\lambda \left(\underset{t}{.Math.}{d}_t^{-}{w}_t\right):d_t^{-}\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right)x_j;d_t^{-}\ge 0;\forall t=1.Math.T\right\},$ the minimax as expressed in $\underset{x,v}{\max }\left\{\underset{j}{.Math.}{x}_j{\mu }_j-\lambda v:v\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right)x_j;\forall t=1.Math.T\right\},$ the CVaR as expressed in $\underset{x,d^{-},\eta }{\max }\left\{\begin{array}{c}\underset{j}{.Math.}{x}_j{\mu }_j-\lambda \left(\underset{j}{.Math.}{x}_j{\mu }_j-\eta +\frac{1}{\beta }\underset{t}{.Math.}{d}_t^{-}{w}_t\right):\\ d_t^{-}\ge \eta -\underset{j}{.Math.}{r}_{j,t}{x}_j;d_t^{-}\ge 0;\forall t=1.Math.T\end{array}\right\}$ where: ƒ is defined as VaR at an optimum, and β is defined as a probability of returns ≤VaR, with 0≤β≤1, the GMD as expressed in $\underset{x,u}{\max }\left\{\begin{array}{c}\underset{j}{.Math.}{x}_j{\mu }_j-\lambda \left(\underset{j}{.Math.}{x}_j{\mu }_j-\underset{\underset{j}{t}}{.Math.}{r}_{j,t}{x}_j{w}_t^2-2\underset{\underset{t^{\prime }+1\le t^{\mathrm{\prime \prime }}\le T}{1\le t^{\prime }\le T-1}}{.Math.}{u}_{t^{\prime },t^{\mathrm{\prime \prime }}}{w}_{t^{\prime }}{w}_{t^{\mathrm{\prime \prime }}}\right):\\ u_{t^{\prime },t^{\mathrm{\prime \prime }}}\le \underset{j}{.Math.}{r}_{j,t^{\prime }}{x}_j;u_{t^{\prime },t^{\mathrm{\prime \prime }}}\ge \underset{j}{.Math.}{r}_{j,t^{\mathrm{\prime \prime }}}{x}_j;\forall t^{\prime }=1.Math.T-1;t^{\mathrm{\prime \prime }}=t^{\prime }+1.Math.T\end{array}\right\},$ or the WCVaR as expressed in $\underset{x,d^{-}}{\max }\left\{\begin{array}{c}\underset{h}{.Math.}{\omega }_h\left(\underset{j}{.Math.}{x}_j{\mu }_j-\lambda \left(\underset{j}{.Math.}{x}_j{\mu }_j-{\eta }_h+\frac{1}{{\beta }_h}\underset{t}{.Math.}{d}_{h,t}^{-}{w}_t\right)\right):\\ d_{h,t}^{-}\ge {\eta }_h-\underset{j}{.Math.}{r}_{j,t}{x}_j;d_{h,t}^{-}\ge 0;\forall t=1.Math.T;{\omega }_h\in \mathrm{weights}\end{array}\right\}$ where: η is defined as VaR at an optimum, and β is defined as a probability of returns ≤VaR, with 0≤β≤1.

8. The method of claim 3, further comprising holding the optimal size of a physical hedge as fixed at a size of an existing physical hedge associated with the renewable generator asset.

9. A method to manage operation of a renewable power asset, comprising: operating one or more computing devices controlling or in communication with one or more control devices of a renewable power asset; identifying, by the computing device, one or more market products to be sold by the renewable power asset; generating, by the one or more computing devices, forward looking market price scenarios representative of the one or more market products; calculating, by the one or more computing devices, baseline revenues of the renewable power asset in the forward looking market price scenarios; determining, by the one or more computing devices, a first cash flow associated with the baseline revenues; determining, by the one or more computing devices, a second cash flow by adding at least one financial hedge into the first cash flow, wherein each financial hedge is any of a fixed for floating revenue swap, a fixed for floating renewable volume swap, or a fixed for floating price swap, applied in optimal amounts to revenues or components of revenues of the renewable power asset, and wherein each financial hedge has attributes of fixed revenues, fixed volumes, fixed prices, upper and lower bounds, and financial hedge premiums; optimizing, by the one or more computing devices, a size of the at least one financial hedge and an operational strategy of the renewable power asset; transacting, by the one or more computing devices, exchange of the at least one financial hedge at the optimized size through a financial transaction interface; controlling, by the one or more computing devices or the one or more control devices in communication with the computing device, the renewable power asset to generate and deliver power based on the optimized operational strategy; and wherein optimizing further comprises optimizing based on a mean-variance optimization of risk and return metrics wherein optimizing further comprises optimizing based on a mean-variance optimization of risk and return metrics using $\underset{q_X,{\alpha }_{s,\tau },{\delta }_{s,\tau }}{\max }\left\{{\mu }_{\tilde{\vartheta }}-\lambda \left(\underset{\begin{array}{c}0\le m\le M\\ {0\le k\le K}^{}\end{array}}{.Math.}{{\gamma }_m\left(\right)}^2-{\mu }^{\stackrel{2}{\stackrel{~}{\vartheta }}}\right)\right\}$ where: q.sub.X is a real time curtailment of renewable generation, α.sub.S,τ is a multiplier of a hedge volume shape perturbation in a season s, and a time block τ, δ.sub.S,τ is a multiplier of a hedge price shape perturbation in the season s, and the time block τ, γ.sub.m is a factor biasing optimization towards specific generation forecast scenarios, λ is a risk tolerance factor, q.sub.RT is a real time renewable generation, q.sub.RT,m is a real time renewable generation forecast in an m.sup.th scenario, q.sub.DA is a day ahead commitment, q.sub.X is a real time curtailment of renewable generation, q.sub.H is defined to be a baseline hedge volume shape, q.sub.H,S,τ is a hedge volume shape perturbation in the season s, and the time block τ, p.sub.RT is a real time price, p.sub.RT,k is a real time price forecast in a k.sup.th scenario, p.sub.H is a baseline hedge price shape, p.sub.H,S,τ is a hedge price shape perturbation in the season s, and the time block τ, δ.sub.RT,m,k is a real time penalty, is a joint probability of (q.sub.RT=q.sub.RT,m; p.sub.RT=p.sub.RT,k), M is a number of possible renewable generation scenarios, indexed by 1≤m≤M, K is a number of possible RT price scenarios, indexed by 1≤k≤K, μ.sub.{tilde over (ϑ)} is defined as $\underset{\begin{array}{c}0\le m\le M\\ {0\le k\le K}^{}\end{array}}{.Math.}{\gamma }_m,$ and is defined as ${\left(q_{\mathrm{RT},m}-q_{\mathrm{DA}}-q_X\right)}^T.Math.p_{\mathrm{RT},k}-{\delta }_{\mathrm{RT},m,k}-q_H^T.Math.\left(p_{\mathrm{RT},k}-p_H\right)-\underset{\underset{\tau }{s}}{.Math.}{\alpha }_{s,\tau }{q}_{H,s,\tau }^T.Math.\left(p_{\mathrm{RT},k}-p_H\right)+\underset{\underset{\tau }{s}}{.Math.}{\delta }_{s,\tau }{q}_H^T.Math.p_{H,s,\tau }.$

10. A method to manage operation of a renewable power asset, comprising: operating one or more computing devices controlling or in communication with one or more control devices of a renewable power asset; identifying, by the computing device, one or more market products to be sold by the renewable power asset; generating, by the one or more computing devices, forward looking market price scenarios representative of the one or more market products; calculating, by the one or more computing devices, baseline revenues of the renewable power asset in the forward looking market price scenarios; determining, by the one or more computing devices, a first cash flow associated with the baseline revenues; determining, by the one or more computing devices, a second cash flow by adding at least one financial hedge into the first cash flow, wherein each financial hedge is any of a fixed for floating revenue swap, a fixed for floating renewable volume swap, or a fixed for floating price swap, applied in optimal amounts to revenues or components of revenues of the renewable power asset, and wherein each financial hedge has attributes of fixed revenues, fixed volumes, fixed prices, upper and lower bounds, and financial hedge premiums; optimizing, by the one or more computing devices, a size of the at least one financial hedge and an operational strategy of the renewable power asset; transacting, by the one or more computing devices, exchange of the at least one financial hedge at the optimized size through a financial transaction interface; controlling, by the one or more computing devices or the one or more control devices in communication with the computing device, the renewable power asset to generate and deliver power based on the optimized operational strategy; and wherein optimizing further comprises: optimizing through a linear programming framework based on utility functions of risk and return metrics such that:
=(q.sub.RT,m−q.sub.DA−q.sub.X).sup.T.Math.p.sub.Rt,k−q.sub.H.sup.T.Math.(p.sub.RT,k−p.sub.H)−Σ.sub.s.sub.τα.sub.S,τq.sub.H,S,τ.sup.T.Math.(p.sub.RT,kp.sub.H)+Σ.sub.s.sub.ΣΕ.sub.S,τq.sub.H.sup.T.Math.p.sub.H,S,τwhere q.sub.RT,m is a real time renewable generation forecast in an m.sup.th scenario, q.sub.DA is a day ahead commitments, q.sub.X is a real time curtailment of renewable generation, q.sub.H is defined to be a baseline hedge volume shape, q.sub.H,S,τ is a hedge volume shape perturbation in the season s, and the time block τ, p.sub.RT,k is a real time price forecast in a k.sup.th scenario, p.sub.H is a baseline hedge price shape, p.sub.H,S,τ is a hedge price shape perturbation in the season s, and the time block τ, α.sub.S,τ is a multiplier of a hedge volume shape perturbation in a season s, and a time block τ, δ.sub.S,τ is a multiplier of a hedge price shape perturbation in the season s, and the time block τ, M is a number of possible renewable generation scenarios, indexed by 1≤m≤M, K is a number of possible RT price scenarios, indexed by 1≤k≤K, and J is a number of discretized time steps in the optimization horizon, indexed by 1≤j≤J; introducing an auxiliary optimization variable x and a unit returns r.sub.m,k such that for S seasons and N time blocks per day, $x\stackrel{\mathrm{def}}{=}\left[\begin{array}{c}{q}_X\\ {\alpha }_{1,1}\\ .Math.\\ {\alpha }_{1,N}\\ .Math.\\ {\alpha }_{S,1}\\ .Math.\\ {\alpha }_{S,N}\\ {\delta }_{1,1}\\ .Math.\\ {\delta }_{1,N}\\ .Math.\\ {\delta }_{S,1}\\ .Math.\\ {\delta }_{S,N}\\ x_{J+1+\left(2\star S\star N\right)}\end{array}\right],$ $r_{m,k}\stackrel{\mathrm{def}}{=}\left[\begin{array}{c}-p_{\mathrm{RT},k}\\ -q_{H1,1}^T.Math.\left(p_{\mathrm{RT},k}-p_H\right)\\ .Math.\\ -q_{H1,N}^T.Math.\left(p_{\mathrm{RT},k}-p_H\right)\\ .Math.\\ -q_{H,S,1}^T.Math.\left(p_{\mathrm{RT},k}-p_H\right)\\ .Math.\\ -q_{H,S,N}^T.Math.\left(p_{\mathrm{RT},k}-p_H\right)\\ q_H^T.Math.p_{H,1,1}\\ .Math.\\ q_H^T.Math.p_{H,1,N}\\ .Math.\\ q_H^T.Math.p_{H,S,1}\\ .Math.\\ q_H^T.Math.p_{H,S,N}\\ {\left(q_{\mathrm{RT},m}-q_{\mathrm{DA}}\right)}^T.Math.p_{\mathrm{RT},k}-q_H^T.Math.\left(p_{\mathrm{RT},k}-p_H\right)\end{array}\right],$ and
=x.sup.T.Math.r.sub.m,k; x.sub.J+1+2(+2*S*N)=1; indexing scenarios to a variable t within a range of total scenario permutations, and expanding to components of x, r.sub.t such that {tilde over (ϑ)}.sub.t=x.sup.T.Math.r.sub.t=Σ.sub.1≤j≤J+(2*S*N)x.sub.jr.sub.j,t; x.sub.J+1+(2*S*N)=1; introducing an average return μ.sub.j at a given time step such that μ.sub.j Σ.sub.1≤t≤Tw.sub.tr.sub.j,t; considering a hedge counter-party's cash flow to be
r.sub.H,k=q.sub.H.sup.t.Math.(p.sub.RT,k−p.sub.H)+Σ.sub.S.sub.τα.sub.S,τq.sub.H,S,τ.sup.T.Math.(p.sub.RT,k−p.sub.H)−Σ.sub.S.sub.τδ.sub.S,τq.sub.H.sup.T.Math.p.sub.H,S,τ; and configuring a linear programming goal and applying one of: a mean absolute deviation (MAD), a minimax, a conditional value risk (CVaR), a Gini mean difference (GMD), or a weighted conditional value at risk (WCVaR).

11. The method of claim 10, wherein configuring a linear programming goal further comprises configuring the goal expressed in $\underset{x}{\max }\left\{\mu \left(x\right)-\mathrm{\lambda \rho }\left(x\right)\right\}$ where:
λ is a risk tolerance factor, and
ρ(x) is defined as Dispersion; and applying one of the MAD, the minimax, the CVaR, the GMD, or the WCVaR further comprises applying one of: the MAD as expressed in $\underset{x,d^{-}}{\max }\left\{\underset{j}{.Math.}{x}_j{\mu }_j-\lambda \left(\underset{t}{.Math.}{d}_t^{-}{w}_t\right):d_t^{-}\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right)x_j;d_t^{-}\ge 0;\forall t=1.Math.T\right\},$ the minimax as expressed in $\underset{x,v}{\max }\left\{\underset{j}{.Math.}{x}_j{\mu }_j-\lambda v:v\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right)x_j;\forall t=1.Math.T\right\},$ the CVaR as expressed in $\underset{x,d^{-},\eta }{\max }\left\{\begin{array}{c}\underset{j}{.Math.}{x}_j{\mu }_j-\lambda \left(\underset{j_{}}{.Math.}{x}_j{\mu }_j-\eta +\frac{1}{\beta }\underset{t}{.Math.}{d}_t^{-}{w}_t\right):\\ d_t^{-}\ge \eta -\underset{j}{.Math.}{r}_{j,t}{x}_j;d_t^{-}\ge 0;\forall t=1.Math.T\end{array}\right\}$ where: η is defined as VaR at an optimum, and β is defined as a probability of returns ≤VaR, with 0≤β≤1, the GMD as expressed in $\underset{x,u}{\max }\left\{\begin{array}{c}\underset{j}{.Math.}{x}_j{\mu }_j-\lambda \left(\underset{j}{.Math.}{x}_j{\mu }_j-\underset{\underset{j}{t}}{.Math.}{r}_{j,t}{x}_j{w}_t^2-2\underset{\underset{t^{\prime }+1\le t^{\mathrm{\prime \prime }}\le T}{1\le t^{\prime }\le T-1}}{.Math.}{u}_{t^{\prime },t^{\mathrm{\prime \prime }}}{w}_{t^{\prime }}{w}_{t^{\mathrm{\prime \prime }}}\right):\\ u_{t^{\prime },t^{\mathrm{\prime \prime }}}\le \underset{j}{.Math.}{r}_{j,t^{\prime }}{x}_j;u_{t^{\prime },t^{\mathrm{\prime \prime }}}\le \underset{j}{.Math.}{r}_{j,t^{\mathrm{\prime \prime }}}{x}_j;\forall t^{\prime }=1.Math.T-1;t^{\mathrm{\prime \prime }}=t^{\prime }+1.Math.T\end{array}\right\},$ or the WCVaR as expressed in $\underset{x,d^{-}}{\max }\left\{\begin{array}{c}\underset{h}{.Math.}{\omega }_h\left(\underset{j}{.Math.}{x}_j{\mu }_j-\lambda \left(\underset{j}{.Math.}{x}_j{\mu }_j-{\eta }_h+\frac{1}{{\beta }_h}\underset{t}{.Math.}{d}_{h,t}^{-}{w}_t\right)\right):\\ d_{h,t}^{-}\ge {\eta }_h-\underset{j_{}}{.Math.}{r}_{j,t}{x}_j;d_{h,t}^{-}\ge 0;\forall t=1.Math.T;{\omega }_h\in \mathrm{weights}\end{array}\right\}$ where: η is defined as VaR at an optimum, and β is defined as a probability of returns ≤VaR, with 0≤β≤1.

12. The method of claim 10, wherein configuring a linear programming goal further comprises configuring the goal to optimize subject to a minimum return constraint as expressed in $\underset{x}{\max }\left\{\mu \left(x\right)-\rho \left(x\right):\mu \left(x\right)\ge {\mu }_0\right\}$ where ρ(x) is defined as Dispersion; and applying one of the MAD, the minimax, the CVaR, the GMD, or the WCVaR further comprises applying one of: the MAD as expressed in $\underset{x,d^{-}}{\max }\left\{\underset{j}{.Math.}{x}_j{\mu }_j-\underset{t}{.Math.}{d}_t^{-}{w}_t:d_t^{-}\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right)x_j;d_t^{-}\ge 0;\forall t=1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\right\},$ the minimax as expressed in $\underset{x,v}{\max }\left\{\underset{j}{.Math.}{x}_j{\mu }_j-v:v\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right)x_j;\forall t=1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\right\},$ the CVaR as expressed in $\underset{x,d^{-},\eta }{\max }\left\{\eta -\frac{1}{\beta }\underset{t}{.Math.}{d}_t^{-}{w}_t:d_t^{-}\ge \eta -\underset{j}{.Math.}{r}_{j,t}{x}_j;d_t^{-}\ge 0;\forall t=1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\right\}$ where η is defined as VaR at an optimum, and β is defined as a probability of returns ≤VaR, with 0<β≤1, the GMD as expressed in $\underset{x,u}{\max }\left\{\begin{array}{c}\left(\underset{\underset{j}{t}}{.Math.}{r}_{j,t}{x}_j{w}_t^2+2\underset{\underset{t^{\prime }+1\le t^{\mathrm{\prime \prime }}\le T}{1\le t^{\prime }\le T-1}}{.Math.}{u}_{t^{\prime },t^{\mathrm{\prime \prime }}}{w}_{t^{\prime }}{w}_{t^{\mathrm{\prime \prime }}}\right):\\ u_{t^{\prime },t^{\mathrm{\prime \prime }}}\le \underset{j}{.Math.}{r}_{j,t^{\prime }}{x}_j;u_{t^{\prime },t^{\mathrm{\prime \prime }}}\le \underset{j}{.Math.}{r}_{j,t^{\mathrm{\prime \prime }}}{x}_j;\forall t^{\prime }=1.Math.T-1;t^{\mathrm{\prime \prime }}=t^{\prime }+1.Math.T;\underset{J}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\end{array}\right\},$ or the WCVaR as expressed in $\underset{x,d^{-}}{\max }\left\{\begin{array}{c}\underset{h}{.Math.}{\omega }_h\left({\eta }_h-\frac{1}{{\beta }_h}\underset{t}{.Math.}{d}_{h,t}^{-}{w}_t\right):\\ d_{h,t}^{-}\ge {\eta }_h-\underset{j}{.Math.}{r}_{j,t}{x}_j;d_{h,t}^{-}\ge 0;\forall t=1.Math.T;{\omega }_h\in \mathrm{weights};\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\end{array}\right\}$ where: η is defined as VaR at an optimum, and β is defined as a probability of returns ≤VaR, with 0 <β≤1.

13. The method of claim 10, wherein configuring a linear programming goal further comprises configuring the goal to optimize based on minimizing dispersion while subject to a minimum return constraint as expressed in $\underset{x}{\min }\left\{\rho \left(x\right):\mu \left(x\right)\ge {\mu }_0\right\}$ where ρ(x) is defined as Dispersion; and applying one of the MAD, the minimax, the CVaR, the GMD, or the WCVaR further comprises applying one of: the MAD as expressed in $\underset{x,d^{-}}{\min }\left\{\underset{t}{.Math.}{d}_t^{-}{w}_t:d_t^{-}\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right)x_j;d_t^{-}\ge 0;\forall t=1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\right\},$ the minimax as expressed in $\underset{x,v}{\min }\left\{v:v\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right)x_j;\forall t=1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\right\},$ the CVaR as expressed in $\underset{x,d^{-},\eta }{\min }\left\{\begin{array}{c}\underset{j}{.Math.}{x}_j{\mu }_j-\eta +\frac{1}{\beta }\underset{t}{.Math.}{d}_t^{-}{w}_t:\\ d_t^{-}\ge \eta -\underset{j}{.Math.}{r}_{j,t}{x}_j;d_t^{-}\ge 0;\forall t=1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\end{array}\right\}$ where: η is defined as VaR at an optimum, and β is defined as a probability of returns≤VaR, with 0<β≤1, the GMD as expressed in $\underset{x,u}{\min }\left\{\begin{array}{c}\left(\underset{j}{.Math.}{x}_j{\mu }_j-\underset{\underset{j}{t}}{.Math.}{r}_{j,t}{x}_j{w}_t^2-2\underset{\underset{t^{\prime }+1\le t^{\mathrm{\prime \prime }}\le T}{1\le t^{\prime }\le T-1}}{.Math.}{u}_{t^{\prime },t^{\mathrm{\prime \prime }}}{w}_{t^{\prime }}{w}_{t^{\mathrm{\prime \prime }}}\right):\\ u_{t^{\prime },t^{\mathrm{\prime \prime }}}\underset{j}{.Math.}{r}_{j,t^{\prime }}{x}_j;u_{t^{\prime },t^{\mathrm{\prime \prime }}}\le \underset{j}{.Math.}{r}_{j,t^{\mathrm{\prime \prime }}}{x}_j;\forall t^{\prime }=1.Math.T-1;t^{\mathrm{\prime \prime }}=t^{\prime }+1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\end{array}\right\},$ or the WCVaR as expressed in $\underset{x,d^{-}}{\min }\left\{\begin{array}{c}\underset{j}{.Math.}{x}_j{\mu }_j-\underset{h}{.Math.}{\omega }_h\left({\eta }_h-\frac{1}{{\beta }_h}\underset{t}{.Math.}{d}_{h,t}^{-}{w}_t\right):\\ d_{h,t}^{-}\ge {\eta }_h-\underset{j}{.Math.}{r}_{j,t}{x}_j;d_{h,t}^{-}\ge 0;\forall t=1.Math.T;{\omega }_h\in \mathrm{weights};\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\end{array}\right\}$ where: η is defined as VaR at an optimum, and β is defined as a probability of returns ≤VaR, with 0<β≤1.

14. The method of claim 10, wherein configuring a linear programming goal further comprises configuring the goal to optimize based relative to a no risk scenario as expressed in $\underset{x}{\max }\left\{\left(\mu \left(x\right)-r_0\right)/\rho \left(x\right)\right\}$ where r.sub.0 is defined as a risk free return and ρ(x) is defined as Dispersion; and applying one of the MAD, the minimax, the CVaR, the GMD, or the WCVaR further comprises applying one of: the MAD as expressed in $\underset{\tilde{x},{\tilde{d}}^{-},z}{\max }\left\{\begin{array}{c}\underset{j}{.Math.}{\tilde{x}}_j{\mu }_j-r_{0}z:\underset{t}{.Math.}{\tilde{d}}_t^{-}{w}_t=z;{\tilde{d}}_t^{-}\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right){\tilde{x}}_j;\\ {\tilde{d}}_t^{-}\ge 0;\forall t=1.Math.T;\underset{j}{.Math.}{\tilde{x}}_j=z;{\tilde{x}}_j\ge 0\end{array}\right\}.Math.x=\tilde{x}/z,$ the minimax as expressed in $\underset{\tilde{x},\tilde{v}}{\max }\left\{\begin{array}{c}\underset{j}{.Math.}{\tilde{x}}_j{\mu }_j-r_{0}z:\\ \tilde{v}=z;\tilde{v}\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right){\tilde{x}}_j;\forall t=1.Math.T;\underset{j}{.Math.}{\tilde{x}}_j=z;{\tilde{x}}_j\ge 0\end{array}\right\}.Math.x=\tilde{x}/z,$ or the CVaR as expressed in $$\begin{gathered} \underset{\tilde{x},{\tilde{d}}^{-},z,\eta }{\max }\left\{\underset{j}{.Math.}{\tilde{x}}_j{\mu }_j-r_{0}z:\underset{j}{.Math.}{\tilde{x}}_j{\mu }_j-\eta +\frac{1}{\beta }\underset{t}{.Math.}{\tilde{d}}_t^{-}{w}_t=z;{\tilde{d}}_t^{-}\ge \eta -\underset{j}{.Math.}{r}_{j,t}{\tilde{x}}_j;{\tilde{d}}_t^{-}\ge \\ 0;\forall t=1.Math.T;\underset{j}{.Math.}{\tilde{x}}_j=z;{\tilde{x}}_j\ge 0\right\}.Math.x=\tilde{x}/z \end{gathered}$$ where η is defined as VaR at an optimum, and β is defined as a probability of returns ≤VaR, with 0<β≤1.

15. A method to manage operation of a renewable power asset, comprising: operating one or more computing devices controlling or in communication with one or more control devices of a renewable power asset; identifying, by the computing device, one or more market products to be sold by the renewable power asset; generating, by the one or more computing devices, forward looking market price scenarios representative of the one or more market products; calculating, by the one or more computing devices, baseline revenues of the renewable power asset in the forward looking market price scenarios; determining, by the one or more computing devices, a first cash flow associated with the baseline revenues; determining, by the one or more computing devices, a second cash flow by adding at least one physical hedge and at least one financial hedge into the first cash flow; optimizing, by the one or more computing devices, a size of the at least one physical hedge, a size of the at least one financial hedge, and an operational strategy of the renewable power asset; sizing the at least one physical hedge at the optimized size, and charging and discharging the at least one physical hedge based on the optimized operational strategy; transacting, by the computing device, exchange of the at least one financial hedge at the optimized size through an energy market exchange interface; controlling, by the one or more computing devices or the one or more control devices in communication with the computing device, the renewable power asset to generate and deliver power based on the optimized operational strategy; and wherein optimizing further comprises optimizing based on a mean-variance optimization of risk and return metrics using $\underset{q_X,d_{\mathrm{RT}},c_{\mathrm{RT}},P_{\mathrm{ESS}},E_{\mathrm{ESS}},{\alpha }_{s,\tau },{\delta }_{s,\tau }}{\max }\left\{{\mu }_{\tilde{\vartheta }}-\lambda \left(\underset{\underset{0\le k\le K}{0\le m\le M}}{.Math.}{{\gamma }_m\left(\right)}^2-{\mu }_{\tilde{\vartheta }}^2\right)\right\}$ where: q.sub.X is a real time curtailment of renewable generation, q.sub.RT,m is a real time renewable generation forecast in an m.sup.th scenario, q.sub.DA is a day ahead commitments, q.sub.H is defined to be a baseline hedge volume shape, q.sub.H,S,τ is a hedge volume shape perturbation in a season s, and a time block τ, d.sub.RT is a real time optimal discharge schedule of the at least one physical hedge, c.sub.RT is a real time optimal charge schedule of the at least one physical hedge, c.sub.P is a storage unit power cost amortized in an optimization horizon, c.sub.E is a storage unit energy cost amortized in the optimization horizon, p.sub.RT,k is a real time price forecast in a k.sup.th scenario, p.sub.H is a baseline hedge price shape, p.sub.H,S,τ is a hedge price shape perturbation in the season s, and the time block τ, p.sub.ESS is a rated power of the at least one physical hedge, E.sub.ESS is a rated energy of the at least one physical hedge, α.sub.s,96 is a multiplier of a hedge volume shape perturbation in the season s, and the time block τ, δ.sub.S,τ is a multiplier of a hedge price shape perturbation in the season s, and the time block τ, γ.sub.m is a factor biasing optimization towards specific generation forecast scenarios, λ is a risk tolerance factor, δ.sub.RT,m,k is a real time penalty, is a joint probability of (q.sub.RT=q.sub.RT,m; p.sub.RT=p.sub.RT,k), M is a number of possible renewable generation scenarios, indexed by 1≤m≤M, K is a number of possible RT price scenarios, indexed by 1≤k≤K, μ.sub.{tilde over (ϑ)} is defined as $\underset{\underset{0\le k\le K}{0\le m\le M}}{.Math.}{\gamma }_m,$ and $\stackrel{\mathrm{def}}{=}{\left(q_{\mathrm{RT},m}+d_{\mathrm{RT}}-c_{\mathrm{RT}}-q_{\mathrm{DA}}-q_X\right)}^T.Math.p_{\mathrm{RT},k}-{\delta }_{\mathrm{RT},m,k}-c_P{P}_{\mathrm{ESS}}-c_E{E}_{\mathrm{ESS}}-q_H^T.Math.\left(p_{\mathrm{RT},k}-p_H\right)-\underset{\underset{\tau }{s}}{.Math.}{\alpha }_{s,\tau }{q}_{H,s,\tau }^T.Math.\left(p_{\mathrm{RT},k}-p_H\right)+\underset{\underset{\tau }{s}}{.Math.}{\delta }_{s,\tau }{q}_H^T.Math.p_{H,s,\tau }.$

16. The method of claim 15, further comprising holding the optimal size of a physical hedge as fixed at a size of an existing physical hedge associated with the renewable generator asset.

17. A method to manage operation of a renewable power asset, comprising: operating one or more computing devices controlling or in communication with one or more control devices of a renewable power asset; identifying, by the computing device, one or more market products to be sold by the renewable power asset; generating, by the one or more computing devices, forward looking market price scenarios representative of the one or more market products; calculating, by the one or more computing devices, baseline revenues of the renewable power asset in the forward looking market price scenarios; determining, by the one or more computing devices, a first cash flow associated with the baseline revenues; determining, by the one or more computing devices, a second cash flow by adding at least one physical hedge and at least one financial hedge into the first cash flow; optimizing, by the one or more computing devices, a size of the at least one physical hedge, a size of the at least one financial hedge, and an operational strategy of the renewable power asset; sizing the at least one physical hedge at the optimized size, and charging and discharging the at least one physical hedge based on the optimized operational strategy; transacting, by the computing device, exchange of the at least one financial hedge at the optimized size through an energy market exchange interface; controlling, by the one or more computing devices or the one or more control devices in communication with the computing device, the renewable power asset to generate and deliver power based on the optimized operational strategy; and wherein optimizing further comprises: optimizing through a linear programming framework based on utility functions of risk and return metrics such that $={\left(q_{\mathrm{RT},m}+d_{\mathrm{RT}}-c_{\mathrm{RT}}-q_{\mathrm{DA}}-q_X\right)}^T.Math.p_{\mathrm{RT},k}-c_P{P}_{\mathrm{ESS}}-c_E{E}_{\mathrm{ESS}}-q_H^T.Math.\left(p_{\mathrm{RT},k}-p_H\right)-\underset{\underset{\tau }{s}}{.Math.}{\alpha }_{s,\tau }{q}_{H,s,\tau }^T.Math.\left(p_{\mathrm{RT},k}-p_H\right)+\underset{\underset{\tau }{s}}{.Math.}{\delta }_{s,\tau }{q}_H^T.Math.p_{H,s,\tau }$ where: q.sub.X is a real time curtailment of renewable generation, q.sub.RT, mis a real time renewable generation forecast in an m.sup.th scenario, q.sub.DA is a day ahead commitments, q.sub.H is defined to be a baseline hedge volume shape, q.sub.H,S,τ is a hedge volume shape perturbation in a season s, and a time block τ, d.sub.RT is a real time optimal discharge schedule of the at least one physical hedge, c.sub.RT is a real time optimal charge schedule of the at least one physical hedge, c.sub.P is a storage unit power cost amortized in an optimization horizon, C.sub.E is a storage unit energy cost amortized in the optimization horizon, p.sub.RT,k is a real time price forecast in a k.sup.th scenario, p.sub.H is a baseline hedge price shape, p.sub.H,S,τ is a hedge price shape perturbation in the season s, and the time block τ, p.sub.ESS is a rated power of the at least one physical hedge, E.sub.ESS is a rated energy of the at least one physical hedge, α.sub.S,τ is a multiplier of a hedge volume shape perturbation in the season s, and the time block τ, δ.sub.S,τ is a multiplier of a hedge price shape perturbation in the season s, and the time block τ, M is a number of possible renewable generation scenarios, indexed by 1≤m≤M, and K is a number of possible RT price scenarios, indexed by 1≤k≤K; introducing an auxiliary optimization variable x and a unit returns r.sub.m,k such that for S seasons and N time blocks per day, $x\stackrel{\mathrm{def}}{=}\left[\begin{array}{c}{q}_X\\ d_{RT}\\ c_{RT}\\ {\alpha }_{1,1}\\ .Math.\\ {\alpha }_{1,N}\\ .Math.\\ {\alpha }_{S,1}\\ .Math.\\ {\alpha }_{S,N}\\ {\delta }_{1,1}\\ .Math.\\ {\delta }_{1,N}\\ .Math.\\ {\delta }_{S,1}\\ .Math.\\ {\delta }_{S,N}\\ P_{ESS}\\ E_{ESS}\\ x_{3J+3+\left(2*S*N\right)}\end{array}\right],$ $r_{m,k}\stackrel{\mathrm{def}}{=}\left[\begin{array}{c}-p_{\mathrm{RT},k}\\ p_{\mathrm{RT},k}\\ -p_{\mathrm{RT},k}\\ -q_{H1,1}^T.Math.\left(p_{\mathrm{RT},k}-p_H\right)\\ .Math.\\ -q_{H,1,N}^T.Math.\left(p_{\mathrm{RT},k}-p_H\right)\\ .Math.\\ -q_{H,S,1}^T.Math.\left(p_{\mathrm{RT},k}-p_H\right)\\ .Math.\\ -q_{H,S,N}^T.Math.\left(p_{\mathrm{RT},k}-p_H\right)\\ q_H^T.p_{H,1,1}\\ .Math.\\ q_H^T.p_{H,1,N}\\ .Math.\\ q_H^T.Math.p_{H,S,1}\\ .Math.\\ q_H^T.Math.p_{H,S,N}\\ -c_P\\ -c_E\\ {\left(q_{RT,m}-q_{DA}\right)}^T.Math.p_{\mathrm{RT},k}-q_H^T.Math.\left(p_{\mathrm{RT},k}-p_H\right)\end{array}\right],$ and
=x.sup.T.Math.r.sub.m,k;x.sub.3J+3+(2*S*N)=1 indexing scenarios to a variable t within a range of total scenario permutations, and expanding to components of x, r.sub.t, and equality contstraints such that {tilde over (ϑ)}.sub.t=x.sup.T.Math.r.sub.t=Σ.sub.1≤j≤3J+3+(2*S*N)x.sub.kr.sub.j,y; x.sub.3J+3+(2*S*N)=1; introducing an average return μ.sub.j at a given time step such that ${\mu }_j\stackrel{\mathrm{def}}{=}\underset{1\le t\le T}{.Math.}{w}_t{r}_{j,t};$ considering a hedge counter-party's cash flow to be
r.sub.H=q.sub.H.sup.T.Math.(p.sub.RT,k−p.sub.H)+Σ.sub.S.sub.τα.sub.S,τq.sub.H,S,96 .sup.T.Math.(p.sub.RT,k−p.sub.H)−Σ.sub.S.sub.τδ.sub.S,96 q.sub.H.sup.T.Math.p.sub.H,S,τ; and configuring a linear programming goal and applying one of: a mean absolute deviation (MAD), a minimax, a conditional value risk (CVaR), a Gini mean difference (GMD), or a weighted conditional value at risk (WCVaR).

18. The method of claim 17, wherein configuring a linear programming goal further comprises configuring the goal to optimize subject to a minimum return constraint as expressed in $\underset{x}{\max }\left\{\mu \left(x\right)-\rho \left(x\right):\mu \left(x\right)\ge {\mu }_0\right\}$ where ρ(x) is defined as Dispersion; and applying one of the MAD, the minimax, the CVaR, the GMD, or the WCVaR further comprises applying one of: the MAD as expressed in $\underset{x,d^{-}}{\max }\left\{\underset{j}{.Math.}{x}_j{\mu }_j-\underset{t}{.Math.}{d}_t^{-}{w}_t:d_t^{-}\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right)x_j;d_t^{-}\ge 0;\forall t=1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\right\},$ the minimax as expressed in $\underset{x,v}{\max }\left\{\underset{j}{.Math.}{x}_j{\mu }_j-v:v\ge \underset{j}{.Math.}\left({\mu }_j-r_{j,t}\right)x_j;\forall t=1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\right\},$ the CVaR as expressed in $\underset{x,d^{-},\eta }{\max }\left\{\eta -\frac{1}{\beta }\underset{t}{.Math.}{d}_t^{-}{w}_t:d_t^{-}\ge \eta -\underset{j}{.Math.}{r}_{j,t}{x}_j;d_t^{-}\ge 0;\forall t=1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\right\}$ where η is defined as VaR at an optimum, and β is defined as a probability of returns ≤VaR, with 0<β≤1, the GMD as expressed in $\underset{x,u}{\max }\left\{\begin{array}{c}\left(\underset{\underset{j}{t}}{.Math.}{r}_{j,t}{x}_j{w}_t^2+2\underset{\underset{t^{\prime }+1\le t^{\mathrm{\prime \prime }}\le T}{1\le t^{\prime }\le T-1}}{.Math.}{u}_{t^{\prime },t^{\mathrm{\prime \prime }}}{w}_{t^{\prime }}{w}_{t^{\mathrm{\prime \prime }}}\right):\\ u_{t^{\prime },t^{\mathrm{\prime \prime }}}\underset{j}{\le .Math.}{r}_{j,t^{\prime }}{x}_j;u_{t^{\prime },t^{\mathrm{\prime \prime }}}\le \underset{j}{.Math.}{r}_{j,t^{\mathrm{\prime \prime }}}{x}_j;\forall t^{\prime }=1.Math.T-1;t^{\mathrm{\prime \prime }}=t^{\prime }+1.Math.T;\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\end{array}\right\},$ or the WCVaR as expressed in $\underset{x,d^{-}}{\max }\left\{\begin{array}{c}\underset{h}{.Math.}{\omega }_h\left({\eta }_h-\frac{1}{{\beta }_h}\underset{t}{.Math.}{d}_{h,t}^{-}{w}_t\right):\\ d_{h,t}^{-}\ge {\eta }_h-\underset{j}{.Math.}{r}_{j,t}{x}_j;d_{h,t}^{-}\ge 0;\forall t=1.Math.T;{\omega }_h\in \mathrm{weights};\underset{j}{.Math.}{x}_j{\mu }_j\ge {\mu }_0\end{array}\right\}$ where: η is defined as VaR at an optimum, and β is defined as a probability of returns ≤VaR, with 0<β≤1.