Abstract

With the increasing prevalence of renewable energy supply contracts, utility suppliers are investing in new green sources and developing allocation policies of those to satisfy renewable targets required by customers. However, the variability of customer demand and the intermittency in supply complicates the supplier’s decision process. In this paper, we address these challenges by formulating the utility supplier’s problem as a two-stage stochastic program. For the second-stage problem, we propose an asymptotically optimal allocation policy (including the dispatching of on-hand resources and the purchasing of additional resources from the market) that satisfies renewable energy target constraints and service level constraints while minimizing the allocation costs. To determine investment levels in the first stage, we demonstrate that under mild assumptions about supply technology, the resource investment problem can be formulated as a minimax saddle point problem and solved efficiently. Our algorithms and main results are applicable for both stationary and periodic demand and supply distributions. We also investigate how the supply and demand distribution, renewable target, and unit investment cost affect the total cost and the optimal decisions. In our computational study, we conduct simulations to showcase the benefits associated with resource pooling and diversification, and highlight environments where aiming for slightly below 100\% may be the better strategy than 100\% renewable. Finally, taking our algorithms to a case study using field data, we provide novel insights into how utilities may invest in different types of green energy sources while optimizing the trade-off between affordability and reliability.

Note: Research papers posted on SSRN, including any findings, may differ from the final version chosen for publication in academic journals.  

---

• Keyword(s):
• operations 143
• sustainability 24

View Working Paper