Nodal Capacity Expansion Planning with Flexible Large-Scale Load Siting

TL;DR

This work addresses capacity expansion planning under uncertainty by explicitly integrating nodal siting and on-site flexibility of large-scale loads, such as datacenters and direct air capture facilities, into a two-stage stochastic MILP that co-optimizes generation, transmission, and storage. Large-load flexibility is modeled through reliability tiers and expectation constraints, enabling strategic siting and operation, while the solution is made scalable via an augmented Progressive Hedging algorithm implemented in mpisppy on HPC. The paper demonstrates the approach on IEEE 24-bus San Diego-inspired and 500-bus South Carolina test systems, showing that proactive large-load planning can yield substantial cost savings and reliability benefits, though computational challenges persist at larger scales. Overall, the methodology advances CEP by combining nodal large-load siting with scenario-based decomposition to harness load flexibility for more cost-effective and reliable power system expansion.

Abstract

We propose explicitly incorporating large-scale load siting into a stochastic nodal power system capacity expansion planning model that concurrently co-optimizes generation, transmission and storage expansion. The potential operational flexibility of some of these large loads is also taken into account by considering them as consisting of a set of tranches with different reliability requirements, which are modeled as a constraint on expected served energy across operational scenarios. We implement our model as a two-stage stochastic mixed-integer optimization problem with cross-scenario expectation constraints. To overcome the challenge of scalability, we build upon existing work to implement this model on a high performance computing platform and exploit scenario parallelization using an augmented Progressive Hedging Algorithm. The algorithm is implemented using the bounding features of mpisppy, which have shown to provide satisfactory provable optimality gaps despite the absence of theoretical guarantees of convergence. We test our approach to assess the value of this proactive planning framework on total system cost and reliability metrics using realistic testcases geographically assigned to San Diego and South Carolina, with datacenter and direct air capture facilities as large loads.

Figures (3)

• Figure 1: Approach used to represent flexibility of large loads. Loads are assumed to be composed of tiers $k=1,\dots,\left \vert \mathcal{K} \right \vert$, with tier $k$ consisting of a fraction $u_k-u_{k-1}$ of total load, and requiring expected service of at least $\phi_k$.
• Figure 2: Left: Resource buildout for each test case. Case IIa has by construction the same buildout as Ic. Case IIb has the same buildout plus one natural gas unit co-sited with the datacenter; both are omitted from this figure. Transmission capacity is obtained by summing the capacities of all selected candidate lines or transformers. Right: Total cost and CO2 emissions for each test case.
• Figure 3: Left: Costs and emissions for cases IIa, IIb, IIc. Note that costs are in a log scale. Right: Achieved and required reliability for each tranche of each large load site built. Different shades of the same color show different tranches of the same site.